基于最优转矩分配的分布式驱动电动汽车车辆稳定性控制设计
需积分: 0 141 浏览量
2021-04-02
23:58:59
上传
评论 1
收藏 324KB PDF 举报
Design of Vehicle Stability Control of Distributed-Driven Electric Vehicle
based on Optimal Torque Allocation
HU Ying
1
, ZHANG Xi-zheng
*1,2
, WANG Yao-nan
2
1. Center of Wind Power Equipment and Energy Conversion (Hunan), Hunan Institute of Engineering, Hunan, Xiangtan 411104
* Correspoding Author, E-mail:
Z_X_Z2000@163.COM
2.School of Electrical and Information Engineering, Hunan University, Hunan, Changsha 410082
Abstract: This paper presented a hierarchical vehicle-stability-control design based on the longitudinal force distribution
optimization for the handling and stability control of the distributed-driven electric vehicle. The 8-DOF vehicle model and the
three-layer control system were developed. By selecting the sideslip angle and the yaw rate as the state variables and introducing
the virtual control to decouple two control variables, the upper controller adopted the integral 2-DOF vehicle model to calculate
the equivalent yaw moment for the vehicle stability. Under the restrictions of the vehicle actuators, the middle controller utilized
the linear quadratic optimization (LQR) method to optimize the distribution of the front and rear steering angles and the tire
longitudinal forces. The sliding-mode-based slip controller in the lower layer was also designed to reallocate the wheel torques.
A simulation test was carried out to verify the effectiveness of the proposed design. Results show that the control system can
make the vehicle follow the expectation effectively and enhance the vehicle handling and stability in extreme conditions with
high speed as well as its active safety under the actuator failures.
Key Words: distributed-driven electric vehicle; vehicle stability control; control allocationwheel slip control
1 Introduction
Stability control of the vehicle requires the vehicle to
travel along an ideal trajectory and remain the ideal state all
the time. In recent years, the new distributed-driven electric
vehicle (DEVs), which places the electric machinery into
the wheel and adopts independent electric machinery as the
driving source, has attracted widespread attention. Under
distributed-driven mode, the driving and brake torques of
each wheel are independently controlled, and precise torque
distribution to each wheel can be realized by directly
controlling the electric machinery according to the vehicle
running status [1, 2]. Therefore, in order to achieve the
stable operation of DEV, the physical longitudinal forces
and the torques must be obtained and then distributed
coordinately to the four wheels.
According to available research literature, the DEV
stability control often adopts hierarchical structure, of which
the upper level is the yaw moment controller, and the lower
level is the control distribution of yaw moment [3~9]. The
yaw moment controller will fulfill the decision-making and
computation of the yaw moment, and further calculate the
expected yaw moment according to the error between the
vehicle current state and the reference value. The control
policy can adopt such modern control methods as the sliding
mode control [4,5], the fuzzy and neural network control
[6,7], the model prediction [8], the optimization method [9]
and so on. At present, most DEV stability control researches
have focused on the yaw moment controller design.
Although this relatively mature design can realize fine
control effect, these control methods usually adopt such
control variables as yaw rate and centroid sideslip angle
when computing the expected yaw moment. Notably, there
Supported by the National Natural Science Foundation of China
(61203019), China Postdoctoral Science Foundation funded project
(2012M521518), the Key Project of Chinese Ministry of Education
(No.212122) and the Natural Science Foundation of Hunan Provincial
(No.13JJ9019).
are coupling relations between the yaw movement and the
sideslip movement, and the yaw moments controlling the
two variables may always be opposite. Thus, it is difficult to
effectively decouple the two variables only depending upon
the direct yaw-moment control (DYC). Especially under the
ultimate conditions, the conflict will be more prominent, and
vehicle stability will be affected greatly [10,11]. Therefore,
a simple and reliable method for decoupling the yaw
movement and the sideslip movement is needed so as to
enhance the vehicle stability under the ultimate conditions.
The expected yaw moment can be distributed precisely to
various actuators by the lower level controller to form the
wheel longitudinal force and the steering angle. Based on
the feed-forward control of front wheel steering angle and
the feedback control of vehicle state error, Ref. [12] adopted
the simple method of evenly distributing torque, but the
stability performance was not very good. Ref. [13] used the
indicator of tire utilization rate to distribute and optimize the
force, but this method required independent brakes and
independent sheerings of all the wheels. Ref. [14] suggested
treating the problem of evenly distributing longitudinal
force as a nonlinear programming problem with inequality
constraints, but the point by point quadratic programming
algorithm might be too complex. On the premise of the
feasible territory real-time estimation of longitudinal forces
and the yaw torque, Ref. [15] constructed vehicle stability
control system of the inner and the outer ring structure
through calculating and optimizing the resultant forces, but
neglected vehicle steering control ability. Besides, Ref. [16]
established a hierarchical modularized control framework
composed of movement controller for computing the yaw
moment and distribution controller for optimizing driving
torque of each wheel, which enhanced the vehicle stability
performance of lateral yaw and anti-rollover, but needed
many online optimizations with complex computation. In
essence, the distribution of force or torque is to solve an
optimization question with constraints. Therefore, how to
develop more effective stability control system through
Proceedings of the 33rd Chinese Control Conference
Jul
y
28-30, 2014, Nan
j
in
g
, China
195
评论0