Modelling of a polymer electrolyte membrane fuel cell system with anodic and cathodic internal recirculation
ABSTRACT. A control-oriented model about PEM fuel cell system with anodic and cathodic internal recirculation is introduced in this paper. It contains the water transport model, oxygen and nitrogen transport model. Then the model is verified by experiment. Results shows a good agreement between the simulated and measured cell voltages, cathode inlet molar water concentrations.
DOS-Robust Dynamic Speed Tracking Controller for an Integrated Motor-Gearbox Powertrain System of a Connected Car
ABSTRACT. This paper deals with the speed tracking control for a connected vehicle with an integrated motor transmission (IMT) powertrain system subject to a denial of service (DOS) attack. For a pure electric vehicle (PEV) with an IMT powertrain system, ensuring speed synchronization is a basic and important control requirement. Connected cars have considerable advantages. However, the in-vehicle controller area network (CAN) connected the external networks will bring network attack (e.g., DOS attack), which may greatly reduce the performance of the control system. To address these problems, a DOS-induced delay model and a dynamic output-feedback robust controller satisfying energy-to-peak performance is designed in this paper. The impact of a DOS attack on the control system is described by CAN message delays which are modeled by polytopic inclusions using Taylor series expansion. Then, a linear parameter varying (LPV) dynamic closed-loop control augmentation model in discrete-time frame considering time-varying delays is established. Furthermore, a dynamic output-feedback robust controller satisfying energy-to-peak performance is designed and the gains is calculated offline by solving a set of linear matrix inequalities (LMIs). Finally, the dynamic controller gains are used for the online step, and the speed tracking performance and oscillation damping capability of the proposed controller is demonstrated by some experiments and compared with a conventional PI controller which is the most widely used controller in engineering.
Experimental investigation of water-cooling based thermal management for cylindrical Li-ion batteries
ABSTRACT. With the increasing popularity of EVs and HEVs, the problem of battery thermal safety becomes more obvious. In this paper, a novel battery thermal management system (BTMS) for Li-ion batteries based on water cooling combined with spiral fins is designed. The performance of this system is verified by carrying out an experiment in a thermostat chamber, where flow rate and fin height are the primary factors to be considered. The temperature change of the battery pack is firstly observed by adjusting the flow rate, and then the influence of fin height on the temperature field is investigated. The results show that the optimized BTMS can effectively ensure the temperature field of the pack maintained in a proper range. Therefore, water-cooling system is an excellent method for the pack to prevent heat accumulation from leading to the sharp increase of temperature.
Optimization design and evaluation of shift point for electric bus based on multi-objective genetic algorithm
ABSTRACT. In this paper, an integrated simulation platform of electric bus based on AVL-Cruise and MATLAB is established to provide simulation basis for optimal design of shift point. By extracting the objective function, design variables and constraint conditions, a mathematical model for solving the shift point problem with multiple objectives is established. The NSGA-II algorithm was used to carry out multi-objective optimization design for the up-stop and down-stop points to obtain the pareto optimal solution. The optimal pareto solution was evaluated and analyzed by fuzzy comprehensive evaluation method, and the optimal results of shift MAP based on three cycle conditions were obtained. This paper evaluates and analyzes the optimal results of shift MAP based on working conditions from four aspects of power performance, economy, driving performance and braking energy recovery.
The electronic throttle controller based on the model reference adaptive nonlinear triple-step method
ABSTRACT. Electronic throttle control system is extremely important for engine intake control. Due to the nonlinear structure of the electronic throttle and the change of friction during the rotation of the throttle baffle, the angle of the electronic throttle controlled by the Electronic Control Unit (ECU) will cause a large control error. In order to control the electronic throttle more accurately and quickly, this article builds an electronic throttle control system based on dSPACE Rapid Control Prototyping (RCP). This paper first establishes a mathematical model based on the mechanism of the electronic throttle. Next the particle swarm optimization algorithm is applied to identify the parameters of the electronic throttle system. Then the controller based on discrete model reference adaptive control (MRAC) nonlinear triple-step method is designed. Finally, dSPACE RCP is applied for experimental verification. Experimental results demonstrate the efiectiveness of the MRAC nonlinear triple-step method controller with prescribed transient response.
Optimal Control of EGR-VVT Dual Loop in IC Engines: modeling and Validation
ABSTRACT. This study presents a model-based air-path dynamic optimal control problem for IC engine. The control-oriented models describes the intake and exhaust manifold pressure transient performance by two control input Exhaust Gas Recirculation (EGR) and Variable Valve Timing (VVT). Firstly, the recursive least square (RLS) algorithm and Polynomial fitting method are adopted to obtain an optimal estimation for the model parameters. Secondly, the nonlinear model is applied to the intake and exhaust manifold pressure tracking control problem with the optimal control algorithm. The control problem is formulated to ensure the fast pressure tracking performance, and meanwhile to guarantee the combustion stability. Finally, the model validation were implemented on a real gasoline engine, and the experiment results demonstrate the performance of the control-oriented models.
Resource Allocating Network-based Map Learning for Knock Constrained Combustion Control
ABSTRACT. This study presents a map calibration method for combustion engines. The map calibration is formulated as function approximation with basis functions. Firstly, Resource Allocating Network (RAN) is employed for determining the number and position of basis functions. Secondly, the map calibration method is applied in knock probability map learning and thermal efficiency learning using offline experimental data. RAN generates a group of bases offline. The resulting bases show that the calibration method allocates more bases at the area where the map changes dramatically. Finally, the resulting bases can be used for online map learning.
Accurate Torque Control of Hybrid Engine Based on Transient Air Intake Observation
ABSTRACT. The traditional torque control method has the imbalance of fuel and air in the transient state due to the filling effect and hysteresis of the intake pipeline, leading the output torque accuracy and the response speed to decrease. Aiming at this problem, this paper proposes an accurate torque control algorithm based on transient intake air quantity observation. A dynamic manifold model is built based on physical mechanism. Based on this model, a transient intake air quantity observer is built to obtain the actual amount of intake air in transient work condition. On this basis, a model-based torque control architecture is constructed, the control method of model feedforward combined with ADRC feedback is used to control the output torque of the engine. Through bench test, it proves that the method can accurately estimate the instantaneous intake air quantity of the engine and effectively improve the control accuracy and transient response speed of the engine output torque.
ADRC-Based Active Front Steering Strategy for Path Tracking of A Farm Vehicle
ABSTRACT. In view of the performance requirement for the trajectory tracking control of farming vehicles, an active front steering strategy is designed in this paper by active disturbance rejection control (ADRC) technique and particle swarm optimization (PSO) algorithm for path tracking of an agricultural machinery. Due to the capability of estimating the lumped disturbance including system unmodeled dynamics, parameter perturbations and external disturbance by the extended state observer (ESO), the ADRC with ESO is adopted to fully compensate the system uncertainties so as to improve the path tracking performance and guarantee the control performance robustness. Moreover, to further improve the control precision of farming, the PSO algorithm is used to optimize the adjustable parameters of the controller. Simulation results are presented to show the effectiveness of the proposed control strategy.
A Nonlinear Model Predictive Controller Design for Electronic Throttle
ABSTRACT. A well-controlled electronic throttle not only improves the fuel economy, but also reduces the emission of harmful gases from incomplete combustion. With the development of the semiconductor industry, the computing performance of the electronic control unit has also increased, and it is also feasible to control the electronic throttle using model predictive control. In this work, a control-oriented model for the electronic throttle is proposed first. Then, based on the model identification, a nonlinear model predictive controller is developed. The optimal solution is obtained by using the sequential quadratic programming combing the multiple shooting tool. Simulation validation results show that the designed control system has acceptable performance.
Performance Study of Direct Injection Gasoline Engine Based on Hydrogen Injection Strategy
ABSTRACT. By refitting a traditional gasoline engine into a hydrogen-blended gasoline engine with gasoline intake port injection and hydrogen direct injection, the influence of hydrogen injection strategy on the mixture distribution and performance of the engine is explored. Based on the optimized control parameters, the power performance and emission of the hydrogen-blended engine are optimized. The results show that hydrogen injection strategy has a significant effect on the performance of the hydrogen-addition engine. With reasonable hydrogen injection strategy and ignition time, stable lean combustion can be achieved, and engine power can be increased by 18% at low speed and low load, while decreasing NOx emissions by only 6%.
A Review on Source Seeking Control and Its Application to Wheeled Mobile Robots
ABSTRACT. The problem of localising the source of a scalar signal has attracted more and more attention. This paper attempts to present a literature review on research results in seeking the souce using wheeled mobile robots. First of all, this paper reviews the development history of source seeking. Then we enumerate the existing soucing seeking control methods. Finally, the research status and future research directions of source seeking are be summarized
Topology optimization design of 24GHz high gain microstrip antenna with metamaterial cover
ABSTRACT. Conventional microstrip antennas have inherent disadvantages such as low gain, large loss, and poor directivity，this paper taking the antenna gain maximization as the design goal, the presence or absence of square copper patch lattices as the design variables, and introduce a 0.05mm redundant design amount for the square lattice, established A topology optimization model of 24GHz microstrip antenna with metamaterial medium cover. Based on the genetic algorithm's solution strategy, the optimization model is solved, and a new type of super-material cover of microstrip antenna without the checkerboard format is obtained. The finite element software is used to simulate the optimized metamaterial cover microstrip antenna, compared with the conventional antenna, the result show antenna gain is improved by 4.24dB and the half-power beamwidth is shrinking by 48° in the E-plane and 52°in the H-plane, the gain and directivity of the antenna are significantly improved.
Connected Cruise Control Based on Adaptive Dynamic Programming Considering Inertial Delay
ABSTRACT. In this paper, we study the connected vehicular system, where the connected cruise control (CCC) vehicle is mixed into the flow of conventional vehicles, they are all capable of transmitting motional information by vehicle-to-vehicle communication. Considering the inertial delay of the longitudinal vehicle dynamics for the multi-vehicle system, an optimal control problem is formulated to ensure that all vehicles in the platooning drive at the same speed while maintaining the desired headway between adjacent vehicles. By employing the adaptive dynamic programming (ADP) technology, an optimal controller is obtained for the CCC vehicle only relying on online state and input data. Numerical results illustrate that the inertial delay is very important for the control of a platoon system.
Adaptive Tracking Control for Active Seat Suspension System with Time-Varying Full State Constraints
ABSTRACT. This study investigates a kind of active seat suspension systems with time-varying full state constraints (TFSCs), whose aim is to improve the riding comfort. In response to unknown functions in the system, the radial basis function neural
networks (RBFNNs) is adopted. Based on the Tangent Barrier Lyapunov Function (BLF-Tan), an adaptive control scheme is designed. It is demonstrated that all signals in the resulted system are semi-globally uniformly ultimately bounded (SGUUB) with TFSCs. Further, the effectiveness of presented control method is validated via the simulation example.
Modeling of Aluminum-ion Battery and Its Applications on Electric Vehicles
ABSTRACT. Technologies of secondary battery are always hotspots in energy storage society. The aluminum-ion battery that can pop out more electrons in a chemical cycle than lithium-ion battery is recently a good candidate. This study proposes the modeling of aluminum-ion battery in a linear circuit approach. Experimental evaluations are given to reveal its validations. Suggestions are made for its potential usage on electric vehicles.
Machine learning algorithm based battery modeling and management method: A Cyber-Physical System perspective
ABSTRACT. The battery is the bottleneck technologies of electric vehicles (EVs), a precise mathematical model is crucial for the battery management system to ensure the secure and stable operation of the battery. In recent years, in order to realize the accurate state monitoring and management of battery, the development of a flexible, self-reconfigurable and reliable model has become one of the most crucial technologies for electric vehicles. This paper mainly focuses on the battery management issues in new energy vehicles, in which the concept of artificial intelligence and grid-connected vehicle is introduced. Firstly, the concept of Cyber-Physical system (CPS) is applied in battery management issues in our work for a better use of battery data. To establish a precise battery model in cloud, the Support vector regression (SVR) algorithm, a classical artificial intelligence algorithm, is used in our work to model the battery. Finally, a rain-flow cycle counting algorithm-based battery degradation quantification method is proposed to deal with the influence of battery aging phenomenon during modeling the battery. Using the battery data extracted from electric buses, model effectiveness and accuracy are validated. The error of the estimated battery terminal voltage estimator is within 2.5%.
Internal Short Circuit Fault Diagnosis for Lithium-ion Battery Based on Voltage and Temperature
ABSTRACT. The internal short circuit is one of the main causes of fire and explosion of electric vehicle power battery. It is of great importance to detect the internal short circuit fault of lithium battery early for the safe operation of electric vehicles. This paper presents an internal short circuit fault diagnosis method based on the voltage and temperature. The method uses directly measured cell voltages and temperatures without the need of additional hardware. The fault diagnosis of the internal short circuit is realized by analyzing the difference of the voltages and temperatures between the fault battery and the normal battery. In the experiment, three resistors with different resistance values are connected to the battery in parallel to simulate different levels of failure. Experimental results show that this method can effectively diagnosis the internal short circuit of the battery.
Research on Information Acquisition System for Battery compartment Operation of Energy Storage Power Station
ABSTRACT. Battery compartment is the core component of energy storage system, which stores a large number of batteries. Once a serious accident happens, it is easy to cause fire to destroy the battery compartment. It is impossible to obtain the operation data of batteries at the time of the accident, which brings difficulties to accident analysis. In this paper, the fault and abnormal situation of energy storage battery compartment are studied comprehensively, and a scheme of information collection system for battery compartment operation status is proposed. The system can record real-time state information of operation, and quickly start recording and save data when abnormal situation occurs, which provides support for accident analysis of energy storage system.
Research on Model Predictive Control Method for Multi-paralleled DC-DC Converters
ABSTRACT. With the continuous increase of battery capacity in the power battery charging and discharging system(PBCDS), the charging and discharging equipment of power battery also needs to output large power or energy, so the multi-DC-DC parallel-connected structure on the battery side becomes the first choice. In addition, considering the traditional PI control method have a complex parameter adjustment process and the complexity of parameter adjustment increases sharply with the parallel-connected of multi-DC-DC, but also it cannot meet the special requirements of high dynamic response speed and nonovershoot in the power battery charging and discharging process(PBCDP). Therefore, this paper innovatively proposed a model predictive control method for the parallel-connected system with n DC-DC converters, which does not require complicated parameter tuning, but also achieves the special requirements of high dynamic response speed and nonovershoot in the PBCDP. The method has strong universality and high flexibility, and can be applied to power battery charge and discharge conditions with any level of power. Finally, the simulation is conducted to validate the feasibility and correctness of the proposed MPC control method.
A travel classification method suitable for optimal control of electric vehicle operation
ABSTRACT. In order to optimize the operation interval of electric vehicles and hybrid electric vehicles, a travel classification method based on charging and operation location information is proposed according to the characteristics of daily use of vehicles. This method can automatically rank the running interval characteristics of electric vehicles and get the high frequency running range. And when the travel changes, the adaptive adjustment of high frequency travel order. Based on the obtained trip classification information, better interval optimization control can be achieved in high frequency running range. The proposed trip classification method is a novel and promising optimization idea, which provides a good way to further improve the efficiency of electric vehicles.
Real-Time Estimation of Inertial Parameter for Lightweight Electric Vehicle Using Dual Kalman Filter
ABSTRACT. Real-time and accurate knowledge of important vehicle states and inertia parameters in vehicle dynamics control systems is very essential for improving vehicle handling stability and active safety, especially for lightweight electric vehicle where effects of variation in inertial parameters for vehicle stability performance become much more pronounced due to the drastic reduction of vehicle weights and body size. This paper presents a parallel dual extended Kalman filter to simultaneously estimate vehicle states and inertia parameters including the vehicle mass and moment of inertia by utilizing real-time measurements in sensor-wheel motor and lateral acceleration. Simulation with sinusoidal steering-like manoeuvre is carried out using the platform of Matlab/Simulink-Carsim. It is found from the simulation results that the proposed estimation method can effectively estimate the vehicle states and inertia parameters.
Distributed Control Design based on Multi-Agent for Distributed Driving Electric Vehicle
ABSTRACT. This paper proposes a distributed control method for path-tracking of distributed driving electric vehicle (DDEV). We establish the architecture of DDEV based on multi-agent system (MAS), where subsystems of chassis are treated as agents. To ensure the performance of tracking, we adopt distributed model predictive control (DMPC) theory to design the controller. The main novelty of this paper lies in that the steering agent and driving/braking agent involved in path-tracking can compute respectively with less information exchanging and complete a common target for the operation of DDEV. This parallel computing architecture reduces the complexity of computation and provides scalability for controllers of DDEV. Comparison between DMPC and traditional MPC demonstrate that the efficiency of tracking has been improved greatly. This paper is also expected to motivate the exploration of control methods for DDEV.
Research on Mechanism and Key Technology of Intelligent Vehicles Brake By Wire system
ABSTRACT. Brake By Wire system is the development trend of the brake system and also one of the key technologies of Intelligent Vehicles. Firstly the paper introduces two types of Brake By Wire system, including electronic hydraulic brake system and electronic mechanical brake system, and compares its mechanism and energy state chain. Then it introduces the existing Brake By Wire systems and brake actuator of auto companies. Finally it predicts the research direction of the key technologies of Brake By Wire system.
A High Precision Vehicle Tracking Algorithm Based on Digital Image Correlation
ABSTRACT. In order to reduce traffic accidents and solve the problems, which refers to inaccurate tracking of the front vehicle and not-in-time collision pre-warning, existed in vision-based collision pre-warning systems, a high-precision vehicle tracking algorithm based on digital image correlation is proposed in this paper. This method is used in the reference sub-region, which means the detected vehicle area, to complete sub-pixel level search and match. Experiments show that this method performs well in long range tracking. The high-accuracy relative speed can be calculated by using the scaling value of tracking frame. Thus, the tracking accuracy and real-time performance can be improved effectively and the dependence on high-resolution lenses can be reduced.
Study on Dynamic Characteristics of Longitudinal Sloshing in Liquid Tank Semi-trailer
ABSTRACT. Based on the purpose of ensuring the safety of liquid tank semi-trailer, the dynamic characteristics of liquid in the tank truck have been studied. The liquid sloshing in the tank is simulated by the Volume of Fluid method of multiphase flow and k-epsilon turbulent model, and analyzes the effect of the initial velocity, deceleration and filling rate on the tank truck.The results show that, the initial velocity, deceleration and filling rate in the tank all produce different forces and torques to the tank, which affect the driving stability of the tank. If the deceleration and filling rate are constant, the initial velocity will have less influence on the longitudinal stability of the tank. However, when the deceleration increases, the longitudinal stability decreases. Meanwhile, if the filling rate is more than 0.8, the effective deceleration will be the lowest and the longitudinal stability is poor. What is the most important is that the method of this paper is feasible, which provides references for driving and supervising tank truck.
Numerical Simulation on Rollover State of Heavy Semi-trailer Train Based on Trucksim
ABSTRACT. The lateral stability of semi-trailer trains is closely related to the recognition of rollover state. Based on the research purpose of handling stability of vehicles, the semi- trailer train model is constructed by using the multi- body dynamics software Trucksim. According to the national standard GB/ T6323-2014, the steering wheel angle step input has been selected to be tested. Through the method of controlling variables, the impact of a single parameter on the targeted vehicle, and the impact of vehicle speed, tractor, the mass of semi-trailer train, wheelbase and other parameters will be determined. The results will provide references for design and safe driving of semi-trailer.
Development and Application of Unmanned Virtual Simulation HIL Testing Platform
ABSTRACT. With the rapid development of unmanned technology, how to quickly and efficiently test all aspects of the safety and effectiveness of unmanned functions is an important part of ensuring its practical application. In view of the shortcomings of the existing unmanned simulation testing platform, this paper proposes a new unmanned virtual simulation hardware in the loop (HiL) platform, which uses the unmanned simulation software ScanerStudio to build realistic Simulation scenarios, vehicle model, multi-sensor model, and complex traffic flows, Jetson Xavier as a controller and Raspberry Pi as a communication analog converter, simulating the addition of sensor noise signals to enhance the authenticity of algorithmic testing. Based on the HiL testing platform, a hierarchical evaluation model is established, and three evaluation indexes are proposed. The weight vector distribution is performed for each evaluation index, and the algorithm is quantitatively evaluated. The experiment proves that the real vehicle controller can be connected to the platform to test and evaluate the adaptive cruise control algorithm, which greatly improves the test efficiency and accelerates the algorithm development progress.
Lateral Positioning Method for Unmanned Roller Compactor Based on Visual Feature Extraction
ABSTRACT. Accurate and real-time lateral positioning is the key to path tracking control of unmanned roller compactor. In order to ensure the accuracy and real-time of lateral positioning, a lateral positioning method based on visual feature extraction is proposed. Firstly, the camera calibration was carried out to obtain its internal parameters and external parameters. Then the image processing method was used to detect and fit the ground markings in the working area. Finally, in order to achieve the lateral positioning, the visual positioning model was used to calculate the distance between the look-ahead point in the moving direction of the roller compactor and the left and right ground markings. The experimental results show that the average absolute error of static positioning is 0.04 meters, the average relative error is 2.31%, the average absolute error of dynamic experiment is 0.06 meters, the average relative error is 3.51%, and the average processing speed of the algorithm reaches 70 frames/second. It fully meets the needs of positioning measurement accuracy and real-time in engineering applications. In addition, the method is less subject to the interference of electromagnetic signal and requires less hardware cost.
On Development of An Autonomous Ball Collecting Wheeled Mobile Robot
ABSTRACT. This paper presents the technological details of an autonomous ball collecting wheeled mobile robot for tennis and table tennis. A novel mechanical structure of ball collecting device is designed and fabricated. Two wheel differential driving mechanism with an omnidirectional wheel is implemented to actuate and steer the robot. Lidar and RGB-D camera are used to detect environmental information as well as scattered balls in the training field to provide robot with navigation information. Experimental results verify the effectiveness of each subsystem.
ABSTRACT. Robust lane information is critical for lateral control system. It brings possibilities for control system to operate in strict conditions such as no lane mark detected and fuzzy lane mark because of camera limit. Also, due to the time-consuming image processing, the sample rate of lane information is always so low for control systems to make precise control. In order to deal with the problems mentioned above, IMU dynamic information is adopted and fused with the lane information from camera as a new method. It provides instant and more robust lane information for the lateral control system to operate beyond the limits of camera.
A Model-in-the-loop(MiL) simulation based on PreScan-CarSim integration is built up to evaluate the performance of the proposed method. A comparison on Lane Following System(LFS) with/without sensor fusion when temporal failure of lane information occurs is made. It approves that the proposed method successfully provides accurate high sample-rated lane information and have LFS control the vehicle to travel inside the lane with less lateral deviation.
GA-based Velocity Planning Using Jerk as the Encoding Method for Autonomous Vehicles
ABSTRACT. A technique for the optimal velocity planning of autonomous vehicle, using the genetic algorithm, is proposed in this paper. A distance-time graph where the dynamic obstacles occupy the corresponding space will be established. Through gene coding method using jerk of genetic algorithm, the feasible distance-time curve is obtained. A fitness function comprehensively evaluates the safety, smoothness, economy, and speed performance of the curve. Through the reproduction and natural selection of several generations, the best individual is selected as the final velocity curve result. The simulation results based on Prescan show that the final velocity curve has good performance. This paper provides an intelligent, convenient and reliable solution for autonomous vehicle velocity planning.
ABSTRACT. The straight-ahead torque distribution strategies of four-wheel independently driving (4WID) electric vehicles greatly impact on the vehicle’s economy, traction and braking performance. Many adhesion-coefficient-dependent strategies have been put forward to improve the energy efficiency. To avoid the challenge to precisely estimate the road adhesion in practice, this paper proposes an energy-and-adhesion optimal torque distribution strategy with combination of online and offline methods. Both the workload utilization ratio and the energy efficiency are optimized. For the purpose of real-time application, the analytical solution of the optimal problem has been derived which is expressed by structure and experimental parameters. The simulation results reveal that the proposed strategy can obtain the equivalent economy performance under WLTP test cycle as the energy-oriented torque distribution strategy does. In addition, the traction capability and the brake safety can be better guaranteed.
Study on the Effect of Different AC Excitations on the Internal Heating for Low-Temperature Batteries
ABSTRACT. In cold climates, lithium-ion batteries have deteriorated charge-discharge performance so that the driving range of electric vehicles (EVs) will drop substantially. Thus, low-temperature lithium-ion batteries must be preheated to improve the life cycle and available capacity. Many internal AC heating methods have proposed to achieve a fast heating speed. However, the effect of different AC excitations including the current waveforms and RMS values on the internal AC heating has not been studied. Therefore, in this paper, heating experiments are conducted to study the different AC waveforms and RMS of AC excitation for the lithium-ion battery heating at low temperatures. Experimental results indicated that the square wave can achieve a good tradeoff between fast heating speed and less damage to the cycle life of lithium-ion batteries. Moreover, this excitation can heat the battery from -20℃ to 5℃ within 130s.
Fast Segmentation of 3-D Point Clouds Based on Ground Plane State Tracking
ABSTRACT. 3D point cloud segmentation is the first and essential step for LIDAR-based perception, and its result has a great impact on subsequent tasks such as classification and tracking. This paper proposes a fast and precise two-step 3D point cloud segmentation algorithm based on ground plane state tracking. The algorithm first extracts the points belonging to the ground. To avoid processing the whole point cloud in every frame and improve computational efficiency of the algorithm, we introduce tracking idea into the ground point extraction step and estimate the ground plane state by the fusion of prior information and measurement information. the second step is to cluster the remaining non-ground points. We introduce an adaptive threshold-based RBNN (Radially Bounded Nearest Neighbor strategy) clustering algorithm which reduces the number of mis-segmentation by determining the adaptive threshold function according to the characteristics of Lidar point cloud. Experimental results demonstrate that the algorithm consumes less time and achieves high segmentation accuracy than previous works.
Trajectory Planning and Tracking for Autonomous Vehicle Considering Human Driver Personality
ABSTRACT. The rapid development of autonomous driving technology has made it possible for autonomous vehicles to enter a wide range of people’s daily lives. On the basis of realizing automatic driving, it is necessary to improve the comfort of autonomous driving in the future to meet the requirement of different people. In this paper, we integrated trajectory planner
and tracking controller for autonomous vehicle to implement trace planning and tracking. Then, based on the trajectory characteristics of the human driver’s actual driving, the driver’s personalized trajectory is planned. The simulation shows that the personalized autonomy can be achieved.
Path tracking control based on Deep reinforcement learning in Autonomous driving
ABSTRACT. Lane keep assist (LKA) and Adaptive Cruise Control (ACC) are two fundamental yet critical functions for autonomous driving, and conventional methods using PID controllers may not perform well in certain extreme driving conditions. In this paper, we propose a reinforcement-learning-based approach to train the vehicle agent to learn LKA and ACC and hence adapt to diverse scenarios. Particularly, we employ deep deterministic policy gradient (DDPG) algorithm to train the agent and consider both state space and action space as continuous, and designed two neural network critic-network and actor-network to simulate the strategy function and Q-function. Then, we train the two neural networks by deep learning method. Finally, Simulations are conducted with both reinforcement learning and traditional PID controller, and the results of reinforcement learning is more adaptive to extreme road conditions in comparison with a traditional PID controller.
Frictional-electro Composite Braking System Based On Dual Brake Pedal
ABSTRACT. At present, most of the battery electric vehicles in the market use composite braking system with a single pedal input，which requires great changes to the original brake system and cost much.Therefore,This paper aims at designing composite brake system of double brake pedal, Only add a brake pedal with travel switch on the basis of the original frictional braking system,Can we realize multi-breaking mode including pure regenerate break and don’t need to change the original brake system.Tested by real vehicle shows that the system greatly improves the cruising range and effectively improve energy efficiency.
Intelligent Vehicle Collision Risk Modeling and Comprehensive Evaluation Method
ABSTRACT. Collision risk assessment is one of the key technologies of intelligent driving. At present, there are many studies on vehicle collision risk and its evaluation methods, but there are still some problems, such as multiple evaluation methods, inconsistent standards, and huge numerical range. Various evaluation methods make the quantitative description of the collision risk in the same scene different, and there are great differences in the description of the collision risk in different scenarios. It is not convenient to compare the collision risk degree in different dangerous scenarios horizontally. Based on the analysis of the mechanism and kinematics model of vehicle longitudinal and lateral collision, and the idea of time-distance model and motion model, this paper puts forward an evaluation method of vehicle longitudinal and lateral collision risk, and then puts forward a comprehensive evaluation method of vehicle collision risk during driving. This method takes into account both the simplicity of time-distance model calculation and the accuracy of motion model modeling. It can describe the collision risk in dangerous scenarios without dimension normalization. It is of great significance for the development and testing of intelligent vehicles.
Personalized Controller Design for an Electro-mechanical Booster Braking System
ABSTRACT. Electro-mechanical booster(Ebooster) braking system is becoming the mainstream of braking system because its independence from vacuum source, rapid response and high security. In particular, the Ebooster braking system can provide personalized assist characteristics. In fact, Drivers with various driving style have different braking demands, and single assist characteristic can no longer meet the individual demands in intelligent vehicles. So, based on the driver's personalized braking demands, a personalized controller for an Ebooster braking system is designed in this paper. Firstly, personalized braking demands are analyzed. Then a self-designed Ebooster braking system is introduced and modeled. Based on this, the mathematical equations of the assist characteristics is obtained and a personalized controller is designed for different braking styles. Finally, the overall performance of the controller designed in this paper is proven effective by simulation tests.
Study on IMU/Wheel speed sensor integrated zero-speed detection
ABSTRACT. Zero speed detection is a key technology in vehicle integrated navigation and is the basis for zero speed error correction and initial quasi-static alignment. Zero-speed detectors that rely solely on IMU (Inertial Measurement Unit) have zero-speed misjudgment in dynamics conditions, which leads to problems such as erroneous updates. In order to solve the mentioned problems in vehicle integrated navigation, a novel zero-speed detection method based on generalized likelihood ratio test and IMU/wheel speed fusion is proposed and verified by experiments. The test results show that compared to the zero-speed detector relying only on IMU or only on the wheel speed, the zero-speed detector based on the IMU/wheel speed sensor fusion could reduce the misjudgment in dynamic conditions, and improves the accuracy of zero-speed detection.
Direct Torque Control of A Segmented Switched Reluctance Motor for BSG in HEVs
ABSTRACT. This paper proposes a novel direct torque control (DTC) of a segmented-rotor switched reluctance motor (SSRM) for belt-driven starter/generator (BSG) in hybrid electric vehicles (HEVs). First, the initial design of this SSRM is given according to the specifications of BSG. The stator of the SSRM consists of two types of poles, i.e., exciting and auxiliary poles, and the rotor contains a series of discrete segments. Then, a novel nonlinear modeling method for DTC is presented. The torque-balanced method, Fourier model, and cubic spline interpolation method are employed to describe the nonlinear characteristics of the SSRM. Finally, simulation and experiments of DTC and current chopping control (CCC) methods are carried out and compared based on the modeling method. It is found that DTC can greatly reduce the torque ripple and the amplitude of flux linkage under DTC can be well controlled around the given value. Besides, the torque curves obtained from the DTC with and without the rotor clamping device basically agree well with each other.
Design of a real-time nonlinear predictive controller for stability enhancement and experimental results for a production vehicle with four in-wheel motors
ABSTRACT. Vehicle stability control under critical conditions is influenced by the coupled nonlinear characteristics of vehicle dynamics, corresponding safety constraints and rapid response requirements. To address these problems, this paper proposes a real-time nonlinear predictive controller for a distributed-drive electric vehicle on low friction coefficient roads. Two measures are taken to prevent the vehicle from destabilizing due to excessive side slip angles. First, the ideal yaw rate and side slip angle are simultaneously determined. Second, a side slip angle constraint is added to the cost function. The results of the tests show that the overall passing speed can be improved from 50-55 km/h to 60-70 km/h with the proposed control.
Multiple Missiles Cooperative Guidance Based on Leader-follower Strategy
ABSTRACT. According to the characteristics of multiple missiles cooperative guidance, a time-cooperative guidance architecture based on leader-follower strategy is proposed. This architecture is composed of individual guidance for each missile and coordinating strategy of the whole system. The central process unit situated at the leader coordinates the system, and the cooperative information is broadcasted from leader to followers. The coordinate information includes expected impact time and relative motion information of leader and target. The time-cooperative guidance architecture has the characteristics of brief structure, short coordinate time and little information to hand on. For the features of structure and the follower have no seeker, a specific and feasible algorithm of line of sight rate is presented, the requirement of follower computation and detection equipment is decreased. Simulation results show the proposed architecture can realize coordinated attack effectively.
Submit type — Position Paper.
Multi-Object Tracking with Aggregated Local Correlation and Global Flow
ABSTRACT. In this paper, we propose a multi-object tracking algorithm which exploits both aggregated local correlation and global flow. Firstly, we get object position in 2D image and 3D point cloud. The recurrent convolutional network is used for image detection, then lidar point cloud is detected with the image priori information. Secondly, we model the local affinity using aggregated local correlation descriptor, which contains object information of appearance, shape and position. Finally, we model the global multi-object tracking problem as the minimum cost flow problem, which is solved the using the shortest path algorithm.
The proposed approach is tested on the KITTI benchmark datasets. The results show that our algorithm achieves state-of-the-art performance.
An Optimal Trajectory Planning for Autonomous Vehicles Based on Vehicle Dynamics Constraints
ABSTRACT. In order to improve the safety and comfort of the unmanned vehicle, an optimal local trajectory planning algorithm based on the constraints of vehicle dynamics equation is proposed in this paper, which can also conveniently satisfy the constraints at the starting and end point of the trajectory. In order to prove the superiority of this algorithm, this paper compares it with another two commonly used trajectories described by trigonometric and polynomial function between the same starting and end point, and contrastive analysis of the tracking deviation under a same control strategy. The tracking results show that zero error tracking can be achieved for the optimal trajectory because the optimal trajectory and corresponding steering angle are obtained at the same time in our algorithm and the vehicle dynamic constraints also have been taken into consideration. The polynomial path tracking error is the largest and the trigonometric path is moderate comparatively. In the whole tracking process, the optimal path proposed in this paper can obtain the minimum input absolute value and energy value of the steering angle while the vehicle's side angle, heading angle and yaw rate are relatively small, so that the safety and comfort of the vehicle are ensured.
Vehicle Sideslip Angle Estimation based on Switched Fuzzy Model
ABSTRACT. The vehicle sideslip angle is one of the most important state variables for vehicle motion control. Its accurate estimation is necessary for many vehicle control systems. In this paper, a novel switched Takagi–Sugeno (T-S) fuzzy observer is proposed to estimates the vehicle sideslip angle. Firstly, to approximate the nonlinear tire model well, a new piecewise affine (PWA) tire model is established. Different from the classic two-segment PWA tire model, the lateral tire force is divided into three regions, i.e. linear region, nonlinear region and saturated region in the proposed model. Substituting it into the bicycle model, the switched lateral dynamic model is obtained whose switching signal is the tire slip angle. Then, the T-S fuzzy modelling technique is applied to represent the vehicle lateral dynamics with a varying speed. Combined with the aforementioned model, a switched T-S fuzzy model is obtained to describe the lateral dynamics. Based on the measured yaw rate, a switched T-S observer is designed. Simulation results of different maneuvers based on the high- fidelity simulation software veDYNA show that the designed observer provides a fast response and has a good estimation result.
Research on Battery Heating System of Electric Vehicle in Low Temperature Environment
ABSTRACT. At low temperature, the electrolyte viscosity of lithium-ion batteries will increase due to the physical characteristics of lithium-ion batteries, resulting in the decrease of power density and energy density of lithium-ion batteries. Aiming at the problem that the charging and discharging performance of lithium-ion batteries decreases at low temperature, a method of heating battery pack by self-discharging current limiting circuit is proposed, which enables the batteries to spontaneously heat in a short time to achieve the normal working temperature of the batteries under low temperature conditions, so as to achieve the purpose of reducing battery loss, prolonging battery life and reducing the probability of safety accidents.
Study on sensitivity of internal states to operating conditions within PEM fuel cell
ABSTRACT. In order to improve the output power of automobile fuel cell under dynamic load and prolong its working life, it is necessary to understand the spatial inconsistency and sensitivity of its internal state. A simplified two-dimensional model of direct-channel PEM fuel cell is established. The migration of water and gas in membranes and channels is mainly introduced. The sensitivity analysis of the operating conditions of the internal state is carried out in detail. Results show that sensitivity of different variables to operating conditions is different. The water vapor pressure is more sensitive. The temperature and voltage are the key factor of sensitivity analysis.
Adaptive Output Voltage Tracking Control for a Fuel Cell-Boost Converter Power Supply
ABSTRACT. A DC/DC boost converter is an essential subassembly for power/energy transfer from a power source to electrical facilities. Polymer electrolyte membrane fuel cell (PEMFC) is currently drawing great attention as the power source for electric vehicles, and with the coordination of a boost converter tackling with widely changing external load, an electric vehicle can obtain the required voltage supply for its powertrain system. This paper focuses on PEMFC power system voltage control in face of varied tracking references and disturbance injections. An equivalent electrical circuit model for the PEMFC and the boost converter averaging state model are employed to structure the entire power system. Then, a sliding mode control integrated with backstepping methodology is used to provide the control signal in the form of the PWM duty ratio, wherein the control stability is ensured via suitable configuration of Lyapunov functions. Moreover, an adaptive law for dealing with the parameter uncertainties is also adopted. Finally, simulation and experimental tests are carried out to verify the voltage tracking effectiveness of the proposed control method, with external load variation and parametric uncertainties considered in the operation conditions.
A Zero-Current-Switching Heater Based on Four-Resonant-State LC Converter for Low-Temperature Lithium-Ion Batteries of Electric Vehicles
ABSTRACT. A zero-current-switching (ZCS) heater based on four-resonant-state LC converter is proposed to heat low-temperature lithium-ion batteries of electric vehicles (EVs). Under the control of two complementary PWM signals, the proposed AC heater can generate a high-frequency sinusoidal current to internally heat batteries with zero switching loss. The proposed onboard heating topology has the features of compact size, low cost, simple control, zero switching loss, fast heating speed, and high reliability. Experimental results show that by motivating a sinusoidal current among cells with the average amplitude of 4.2 C and the frequency of 17.9 kHz, the proposed ZCS AC heater can warm lithium-ion batteries from −20 oC to 0 oC within 7.8 min, which consumes about 6% of battery energy. Therefore, the proposed heater is beneficial for realizing “all-climate” battery packs without changing battery structures and electrolytes.
Degree of Hybridization Design for a Fuel Cell/Battery Hybrid Electric Vehicle Based on Multi-objective Particle Swarm Optimization
ABSTRACT. Currently, traditional hybrid electric vehicles are not true clean, and more efforts are taken to develop fuel cell/battery hybrid vehicles (FCHVs). However, there are power matching problems on multi power supply system, including the degree of hybridization (DOH) design. In last decade, many research works of DOH design were devoted to optimizing the vehicle’s fuel economy, but the vehicle cost and dynamic performance were mostly neglected. In this paper, a multi-objective optimization design method based on particle swarm optimization (PSO) algorithm is proposed to optimize the vehicle cost and fuel economy on the premise of meeting vehicle dynamic performance. Combined with a simulation platform, the algorithm can automatically search the optimal solution in the entire reasonable range of the DOH. As a result, an optimal hybrid scheme of the FCHV was obtained. The research results of this paper provide basis for the powertrain design of FCHVs.
Dynamic Programming Algorithm for Energy Management Strategy of Fuel Cell Electric Vehicle
ABSTRACT. Due to the basic dynamic programming algorithm applying to power allocation of fuel cell vehicle is not accurate and its calculation is too large, this study proposes a dynamically efficient energy management strategy for the purpose of achieving an optimal power allocation between the energy sources. The simulation results show that the: the algorithm is more accurate than the basic dynamic programming algorithm, and the time of calculation is about only 20 minutes. What’s more, low hydrogen consumption is realized for different initial State of Charge(SOC) of battery.
Speed Control of Four-Wheel Independently Actuated Vehicle based on MPC Algorithm
ABSTRACT. This paper presents a speed tracking control method based on the model predictive control (MPC) algorithm for the four-wheel independently actuated (FWIA) electric vehicle. In the upper level controller, a switching control law is designed to reduce the impact when the car from acceleration to deceleration or from deceleration to acceleration. To avoid high frequency oscillations in the speed control, acceleration increment input law is designed in the lower level controller. The lower level controller calculates the desired torque using Newton's second law. The validity of the proposed control algorithm is verified using the CarSim/Simulink simulation, and the simulation results of the proposed method under extreme conditions are compared with PI controller, which shows the superiority of the proposed method. Finally, by the vehicle test, the experimental results verify the effectiveness and robustness of the proposed control approach.
Keywords—speed tracking control, model predictive control, switching control law
Simulation of energy control strategy for hybrid electric vehicle based on modified dynamic programming
ABSTRACT. Hybrid electric vehicle (HEV) achieves good performance of power, economy and emissions through acceleration assistance, regenerative braking and energy management. The dynamics model of each power component and transmission system is established. In order to reduce the running time of the program, a modified dynamic programming algorithm is proposed to optimize the control strategy globally. The simulation results show that the vehicle dynamics performance can be satisfied, and the power battery’s state of charge(SOC) can be improved significantly. It shows that the proposed energy management strategy is effective and can provide a theoretical basis for formulating the control strategy of complex hybrid system.
Sliding Mode Control Method For Four-wheel Steering Vehicle Stability
ABSTRACT. In the steering process of vehicles, the active steering system mainly controls the lateral force of tires. When the lateral force of tires is close to saturation, the yaw torque generated by the lateral force of tires is not enough to maintain the stability of vehicle steering. To solve this problem, this paper designs a combined four-wheel steering (4WS) and differential braking controller, which compensates for the limitation of lateral tire force saturation in the steering control system by generating an appropriate additional yaw moment. Firstly, the linear quadratic optimal following controller for active steering of rear wheel is designed to output the steering angle of rear wheel required for stable control and control the lateral force of tire. The braking force is distributed by the single wheel differential braking strategy, and an appropriate additional yaw moment is generated by controlling the longitudinal force of the tire to compensate the deficiency of the additional yaw moment generated by the lateral tire force control. The control effect of single control and joint control is verified in the CarSim and MATLAB joint simulation environment, and the simulation results show that the control method proposed in this paper can better improve the vehicle handling stability.
MPC-based fault tolerant control system for yaw stability of distributed drive electric vehicle
ABSTRACT. Usually, the yaw stability of distributed drive electric vehicle (DDEV) is hard to be guaranteed effectively, because the yaw performance is affected by the actuator and sensor faults, as well as system modelling error. To attenuate the effect of these faults, a novel model predictive controller-based fault tolerant control system (MPC-FTCS) is proposed in this work. The MPC-FTCS consists of two MPCs. In one MPC, the sensor fault and system modelling error can be tolerated by the linear quadratic regulation (LQR) design. Another MPC is designed as an observer to estimate and compensate for the actuator fault. The proposed MPC-FTCS is evaluated on the Matlab simulation platform and simulation results show the benefits of the proposed control system.
Fuzzy-Sliding Mode Control based Yaw Stability Control Algorithm for Four-in-wheel-motor Drive Electric Vehicle
ABSTRACT. Four-in-wheel-motor drive electric vehicle bas become one of the new leading research fields for modern electric vehicles. To make full use of the benefits of the in-wheel-motor drive electric vehicle, this paper designs a yaw stability controller based on fuzzy theory and sliding mode control algorithm for a four in-wheel motors driven electric vehicle. Co-simulation dynamical model with speed control of driver for four-in-wheel-motor driven electric vehicle is built firstly. Then, a yaw stability controller based on fuzzy-sliding mode control algorithm is designed for the four in-wheel motors driven electric vehicle and the driving torques are distributed to each wheel. Finally, the performance of the model and the yaw stability control strategy are tested using double lane change test.
Takagi-Sugeno fuzzy control for the semi-active seat suspension with an electromagnetic damper
ABSTRACT. In this paper, a state observer-based Takagi-Sugeno (TS) fuzzy controller is proposed for a semi-active seat suspension installed with an electromagnetic damper (EMD) system. As the nonlinear and hysteresis characteristics of the seat suspension friction, the control system design for seat suspension is difficult. Therefore, the Bouc-Wen friction model is applied to represent the seat suspension friction force and this nonlinear model is linearized by using TS fuzzy approach at first. Then a H_∞ controller that considers the seat suspension performance requirements is designed. It is noticed that not all the state variables are measurable in practice, a state observer is designed based on the measurement variables. Finally, two typical excitation are applied to validate the effectiveness of the proposed controller and a tuned passive suspension is also simulated for comparison.
Power shift control strategy based on the power shift module of high-powered tractor
ABSTRACT. The power-shift module of wet multiplate clutch is selected as the research object. Aiming at the power-shift process that one pair of clutches join in,the power-shift control strategy is proposed with considering the limitation of governing characteristics of diesel engine and loss of slippery work. The power-shift process is divided into two periods -- power transfer process and oncoming clutch speed synchronization process. Then, the determination principle and method of pressure increment relation and upper limitation of actuating pressure of engaging and separating clutches are proposed. Subsequently, the hydraulic pressure of engaging clutch and separating clutch is controlled during each period to guarantee the transmission of large torque. In the end, the simulation of power-shift model based on Matlab/Simulink is carried out. The analyzed results show that the proposed practical power-shifting control strategy of high-powered tractor is able to shift under load condition, while to retain large power output with smooth transition of speed in shift process.
A real-time dynamic trajectory planning for autonomous driving vehicles
ABSTRACT. This paper presents a real-time dynamic trajectory planning algorithm that provides an optimal trajectory for autonomous driving. The proposed trajectory planning method determines not only an optimal path, but also the appropriate acceleration and speed for a vehicle. In this method, a center line is first constructed based on a set of predefined waypoints that are usually used as a reference line. Then, a series of path candidates are generated by the arc length and offset to the center line in the Frenet coordinate system, and the speed space is discretized using polynomial functions. The optimal path trajectory is determined considering the objectives of safety, smoothness, comfortability and consistency, together with the associated acceleration and speed also identified. The simulation results verifies the effectiveness of the proposed method.
Lateral Stability Control of Four-wheel Steering Vehicles
ABSTRACT. This study aims to reduce the side-slip angle and the error of tracking the ideal yaw rate of four-wheel-steering (4WS) vehicles. A fuzzy PID (Proportion Integration Different-iation ) controller is used to control the rear wheel angle. We choose yaw rate deviation as PID controller input; the fuzzy controller sets the yaw rate deviation and deviation change rate as inputs, and the output of the fuzzy controller is used to correct the PID parameters. The output of the 4WS system controller is the rear wheel angle that reduces the yaw rate deviation. We adopte the CarSim/Simulink co-simulation platform. The front wheel steering two-degree-of-freedom vehicle model is selected to be the reference model, and the CarSim vehicle model is used as the controlled object to perform simulation experiments. Simulation results show that the controller can reduce the error of yaw rate tracking and improve the stability of vehicles.
Hardware-in-the-loop simulation of electronic differential moment power steering control strategy for multi-axle vehicle
ABSTRACT. With advantages like high efficiency, low pollution, and convenience of structure layout, the electric vehicles have been developing to become the main research direction of the automobile industry. Compared with the traditional centralized drive mode, the distributed electric drive can improve the power transmission efficiency and the fault tolerance. As the distributed electric drive technology developing in the field of multi-axle vehicles, research for mature dynamic coordinated control strategy has been an important task of the researchers in the automobile industry. This paper was aimed at the dynamic coordinated control strategy of multi-axle vehicles. Due to the independent driving characteristics of distributed drive multi-axle vehicles, a lateral dynamic control method using electronic differential moment power steering was proposed and verified, which can improve driving reliability, steering performance and vehicle mobility. A control model of six-axle-distributed-electric-drive vehicle was set up. With the MATLAB/Simulink software, the C code of the control model was generated and eventually burnt into the vehicle control unit (VCU). Then the hardware-in-the-loop experiment was successfully implemented on the motor bench, verifying the feasibility of the control algorithm.
Research on Dynamic Characteristics of Lateral Sloshing in Liquid Tank Semi-trailer
ABSTRACT. Abstract— In order to ensure the safety of liquid tanker transportation and aiming at the liquid sloshing of the lateral movement of the semi-trailer tanker. The VOF(Volume of Fluid) multi-phase flow model and the standard k-ɛ turbulence model are used to numerically simulate the shaking process of liquid in the tank. The simulation results show that different lateral acceleration, the number of baffles, different liquid filling ratio can pose different forces and moments on the tank. Thus affecting the driving stability of the tanker. When the liquid filling ratio is fixed the greater the lateral acceleration, the greater the amplitude of the liquid sloshing, the greater the impact on the tank. With the increase of liquid filling ratio, the peak value of the impact on the tank decreases and gradually becomes steady over time. Within one cycle, the more baffles the smaller peak value of the lateral force impact on the tank and the less the number of impacts, the faster the rate of lateral force decay . The results show that the research method in this paper is feasible and provides a design reference for improving the lateral stability of liquid tank truck .
