Search Results

Durability performance prediction is a critical issue in fuel cell research. During the demonstration operation of fuel cell commercial vehicles in China, this issue has attracted more attention. In this article, the long short-term memory neural network (LSTMNN), which is an improved recurrent neural network (RNN), and the demonstration operation data are used to establish the prediction model to predict the durability performance of the fuel cell stack. Then, a model based on a back-propagation neural network (BPNN) is established to be a control group. The demonstration operation data is divided into training group and validation group. The former is used to train the prediction model, and the latter is used to verify the validity and accuracy of the prediction model. The outputs of the prediction model, as the durability performance evaluation indexes of the fuel cell, are the polarization curve (current-voltage curve) and the voltage decay curve (time-voltage curve).

With rapid development of the automobile industry and the growing maturity of the automotive electronic technologies, the distributed-traction electric vehicle with four-wheel-distributed steering/braking/traction systems is regarded as an important development direction. With its unique chassis structure, it is the ideal benchmark platform used to evaluate active safety systems. The distributed-traction electric vehicle with four-wheel-distributed steering system is essentially full drive-by-wire vehicle. With its flexible chassis layout and high control degrees-of-freedom, the full drive-by-wire electric vehicle acted as a kind of redundant system is an ideal platform for the research of integrated control. In this treatise, the longitudinal dynamics of the electric vehicle as well as its lateral and yaw motions are controlled simultaneously.

Fuel cell vehicle commercialization and mass production are challenged by the durability of fuel cells. In order to research the durability of fuel cell stack, it is necessary to carry out the related durability test. The performance prediction of fuel cell stack can be based on a short time durability test result to accurately predict the performance of the fuel cell stack, so it can ensure the timeliness of the test results and reduce the cost of test. In this paper, genetic algorithm-BP neural network (GA-BPNN) is proposed to modeling automotive fuel cell stack to predict the performance of it. Based on the strong global searching ability of genetic algorithm, the initial weights and threshold selection of neural networks are optimized to solve the shortcoming that the random selection of the initial weights and thresholds of BP neural network which can easily lead to the local optimal value.

The tire mechanics characteristics are essential for analysis and control of vehicle dynamics. Basically, the effects of sideslip, longitudinal slip, camber angle and vertical load are able to be represented accurately by current existing tire models. However, the research of velocity effects for tire forces and moments are still insufficient. Some experiments have demonstrated that the tire properties actually vary with the traveling velocity especially when the force and moment are nearly saturated. This paper develops an enhanced brush tire model and the UniTire semi-physical model for tire forces and moments under different traveling velocities for raising need of advanced tire model. The primary effects of velocity on tire performances are the rubber friction distribution characteristics at the tire-road interface.

The vehicular fuel cell stack is unavoidably impacted by the vibration in the real-world usage due to the road unevenness. However, effects of vibration on stacks have yet to be completely understood. In this work, the mechanical integrity and gas-tightness of the stack were investigated through a strengthen road vibration test with a duration of 200 h. The excitation signals applied in the vibration test were simulated by the acceleration of the stack, which were previously measured in a vehicle vibration test. The load signals of the vehicle vibration test were iterated through a road simulator from vehicle acceleration signals which were originally sampled in the proving ground. Frequency sweep test was conducted before and after the vibration test. During the vibration test, mechanical structure inspection and pressure maintaining test of the stack were conducted at regular intervals.

A new electric power steering system (EPS) dynamic friction model based on normalized Bouc-Wen model is given, as well as its structure form and model features. In addition, experimental method is used to identify corresponding parameters. In order to improve road feel feedback, this paper analyzes the shortcoming of traditional constant friction compensation control method and proposes a variable friction compensation control method which the friction compensation current changes according to the assist characteristic gain. Through simulation and real vehicle test verification, variable friction compensation control method eliminates the effect of basic assist characteristic, and improves the driver’s road feel under high speed.

Electronically controlled air suspension (ECAS) systems have been widely used in commercial vehicles to improve the ride comfort and handling stability of vehicles, as it can adjust vehicle height according to the driving conditions and the driver's intent. In this paper, the vehicle height adjustment process of ECAS system is studied. A mathematical model of vehicle height adjustment is derived by combining vehicle dynamics theory and thermodynamics theory of variable mass system. Reasons lead to the problems of “over-charging”, “over-discharging” and oscillation during the process of height adjustment are analyzed. In order to solve these problems, a single neuron proportional-integral-derivative (PID) controller is proposed to realize the accurate control of vehicle height. By simulation and semi-physical rig test, the effectiveness and performance of the proposed control algorithm are verified.

Aiming at estimating the vehicle mass and the position of center of gravity, an on-line two-stage estimator, based on the recursive least square method, is proposed for buses in this paper. Accurate information of the center of gravity position is crucial to vehicle control, especially for buses whose center of gravity position can be varied substantially because of the payload onboard. Considering that the buses start and stop frequently, the first stage of the estimator determines the bus total mass during acceleration, and the second stage utilizes the recursive least-square methods to estimate the position of the center of gravity during braking. The proposed estimator can be validated by the co-simulation with MATLAB/Simulink and TruckSim software, simulation results exhibit good convergence and stability, so the center of gravity position can be estimated through the proposed method in a certain accuracy range.

As a new form of electric vehicle, Four-wheel-independent electric vehicle with X-By-Wire (XBW) inherits all the advantages of in-wheel motor drive electric vehicles. The vehicle steering system is liberated from traditional mechanical steering mechanism and forms an advanced vehicle with all- wheel independent driving, braking and steering. Compared with conventional vehicles, it has more controllable degrees of freedom. The design of the integrated vehicle dynamics control systems helps to achieve the steering, driving and braking coordinated control and improves the vehicle's handling stability. In order to solve the problem of lacking of vehicle state information in the integrated control, some methods are used to estimate the vehicle state of four-wheel-independent electric vehicles with XBW. In order to improve the estimation accuracy, unscented Kalman filter (UKF) is used to estimate the vehicle state variables in this paper.

A full drive-by-wire electric vehicle, named Urban Future Electric Vehicle (UFEV) is developed, where the four wheels' traction and braking torques, four wheels' steering angles, and four active suspensions (in the future) are controlled independently. It is an ideal platform to realize the optimal vehicle dynamics, the marginal-stability and the energy-efficient control, it is also a platform for studying the advanced chassis control methods and their applications. A centralized control system of hierarchical structure for UFEV is proposed, which consist of Sensor Layer, Identification and Estimation Layer, Objective Control Layer, Forces and Motion Distribution Layer, Executive Layer. In the Identification and Estimation Layer, identification model is established by utilizing neural network algorithms to identify the driver characteristics. Vehicle state estimation and road identification of UFEV based on EKF and Fuzzy Logic Control methods is also conducted in this layer.

Pneumatic Electric Braking System (EBS) is getting widely spread for commercial vehicles. Pneumatic EBS improves the problem of slow response of traditional pneumatic braking system by implementing brake-by-wire. However, the time-delay response and hysteresis of some electro-pneumatic components and some other issues decrease the response and control accuracy of the pneumatic EBS.

This paper first presents an algorithm to detect tire blowout based on wheel speed sensor signals, which either reduces the cost for a TPMS or provides a backup in case it fails, and a tire blowout model considering different tire pressure is also built based on the UniTire model. The vehicle dynamic model uses commercial software CarSim. After detecting tire blowout, the active braking control, based on a 2DOF reference model, determines an optimal correcting yaw moment and the braking forces that slow down and stop the vehicle, based on a linear quadratic regulator. Then the braking force commands are further translated into target pressure command for each wheel cylinder to ensure the target braking forces are generated. Some simulations are conducted to verify the active control strategy.

A fault detection and diagnosis (FDD) algorithm of 4WID/4WIS Electric Vehicles has been proposed in this study aiming to find the actuator faults. The 4WID/4WIS EV is one of the promising architectures for electric vehicle designs which is driven independently by four in-wheel motors and steered independently by four steering motors. The 4WID/4WIS EVs have many potential abilities in advanced vehicle control technologies, but diagnosis and accommodation of the actuator faults becomes a significant issue. The proposed FDD approach is an important part of the active fault tolerant control (AFTC) algorithm. The main objective of the FDD approach is to monitor vehicle states, find the faulty driving motor and then feedback fault information to the controller which would adopt appropriate control laws to accommodate the post-fault vehicle control system.

The tire mechanics characteristics are essential for analysis, simulation and control of vehicle dynamics. This paper develops the UniTire model for tire forces and moments under combined slip conditions with anisotropic tire slip stiffness. The anisotropy of tire slip stiffness, which means the difference of tire longitudinal slip stiffness and cornering stiffness, will cause that the direction of tire resultant shear stress in adhesion region is different from that in sliding region. Eventually the tire forces and moments under combined slip conditions will be influenced obviously. The author has proposed a “direction factor” before to modify the direction of resultant force in the tire-road contact patch, which can describe tire forces at cornering/braking combination accurately. However, the aligning moments which are very complicated under combined slip conditions are not considered in previous analysis.

The passive fault-tolerant performance of the integrated vehicle controller (IVC) applied on 4WID/4WIS Electric Vehicles has been investigated in this study. The 4WID/4WIS EV is driven independently by four in-wheel motors and steered independently by four steering motors. Thanks to increased control flexibility of the over-actuated architecture, Control Allocation (CA) can be applied to control the 4WID/4WIS EVs so as to improve the handling and stability. Another benefit of the over-actuated architecture is that the 4WID/4WIS Electric Vehicle has sufficient redundant actuators to fight against the safety critical situation when one or more actuators fail.

The properties of contact patch are key factors for tire modeling. Researchers have paid more attention to the contact patch shape and vertical pressure distribution. Some innovative concepts, such as Local Carcass Camber, have been presented to explain special tire modeling phenomena. For a pragmatic tire model, a concise model structure and fewer parameters are considered as the primary tasks for the modeling. Many empirical tire models, such as the well-known Magic Formula model, would become more complex to achieve satisfactory modeling accuracy, due to increasing number of input variables, so the semi-empirical or semi-physical modeling method becomes more attractive. In this paper, the concept of Tire Carcass Camber is introduced first. Different from Local Carcass Camber, Tire Carcass Camber is an imaginary camber angle caused only by lateral force on the unloaded tire.

Steering-By-Wire (SBW) system has advantages of advanced vehicle control system, which has no mechanical linkage to control the steering wheel and front wheels. It is possible to control the steering wheel actuator and front wheels actuator steering independently. The goal of this paper is to use a joystick to substitute the conventional steering wheel with typical vehicle SBW system and to study a variable steering ratio design method. A 2-DOF vehicle dynamic reference model is built and focused on the vehicle steering performance of drivers control joystick. By verifying the results with a hardware-in-the-loop simulation test bench, it shows this proposed strategy can improve vehicle maneuverability and comfort.

A novel predictable tire model has been proposed for combined braking and cornering forces, which is based on only a few pure baking and pure cornering tests. It avoids elaborate testing of all kinds of combinations of braking and side forces, which are always expensive and time consuming. It is especially important for truck or other large size tires due to the capability constraints of tire testing facilities for combined shear forces tests. In this paper, the predictive model is based on the concept of slip circle and state stiffness method. The slip circle concept has been used in the COMBINATOR model to obtain the magnitude of the resultant force under combined slip conditions; however the direction assumption used in the COMBINATOR is not suitable for anisotropic tire slip stiffness.

Steer-by-Wire system is capable of improving the performance of vehicle handling and stability, and assisting driving. It becomes a key technique to control front wheel angle and simulate the steering resistance delivered to the driver because of removing mechanical linkages between the steering wheel and the front wheels. This paper proposes a bilateral control method of steering wheel torque drive/pinion angle feedback, which is disaccustomed of controlling steering wheel block and steering actuator as master-slave plants. The pinion angle, steering wheel angle and its torque signals are used in the control logic without estimating or measuring the tire/road force. Simulations and vehicle experiments proceeded with this proposed method and the results confirmed that it achieves the bilateral control of the position and torque between the two plants.

Omni-directional electric vehicle built by our research group is an advanced electric vehicle whose four wheels can drive, steer and brake independently. The vehicle chassis system is composed of four in-wheel motors, four independent steer motors and electromagnetic brake system, and its control system is divided into logical control layer and underlying execution layer. The information exchange between these two layers is implemented by CAN bus. In this paper, the traction control logic for Omni-directional electric vehicle is developed. The study mainly involves two aspects: the vehicle states estimation and the traction control logic design. The vehicle states, including vehicle longitudinal velocity, lateral speed, side slip angle and yaw rate, etc, are estimated based on Extended Kalman Estimation and multiple degrees of freedom vehicle model.