Considering that the driver's response was not timely under extreme conditions, and there was a time delay between the driver's behavior and the actuator, which would destabilize the vehicle. A coordinated control strategy based on driving state prediction was designed for improving 4WID-EVS lateral stability and handling performance through direct yaw moment control system (DYC) and active front steering (AFS). Firstly, the coordinated control strategy of AFS and DYC was designed, which consisted of two parts: (1) A stability judgeing method based on sideslip angle and sideslip angular velocity phase plane was proposed. The phase plane stability boundary coefficient was adjusted by the road adhesion coefficient, which divided the coordination control region reasonably, so as to judge whether the vehicle was unstable and switched the appropriate subsystem. (2) A driving state prediction algorithm based on data stream mining technology and Markov theory was proposed.
The fuzzy control theory which had good robustness was applied, the driving state in the future were made as the inputs, which could determine the weight coefficients of AFS and DYC in advance and effectively avoid the potential danger of entering the unstable state. Secondly, the nonlinear 3-dof vehicle model was used as the reference model, and AFS control strategy was designed based on model predictive control (MPC) theory. At the same time, in order to avoid no solution in the calculation process and improve the convergence speed of the solution process, the relaxation factor was introduced. The quadratic programming algorithm was adopted to find the optimal solution, so as to adjust the front wheel angle in real time to follow the desired path. In order to make up for the deficiency of AFS control strategy, the feedforward - feedback DYC control strategy was designed.
A variable weight coefficient LQR feedback control strategy was proposed, and the matrix weight coefficient could be adjusted in real time according to the vehicle steering state in the future and the road adhesion coefficient at the present. Finally, the accuracy and real-time of the prediction algorithm were verified according to different and actual driving cycle data. The simulation experiments of typical working conditions were carried out based on commercial dynamics simulation software to verify the effectiveness of the coordinated control strategy. The simulation results showed that the coordinated control method of AFS and DYC proposed had better effect of driving stability control, which provided a new method for the design of vehicle active safety.
KEY WORDS: distributed drive, coordination control, data stream mining, Markov, MPC
Dr. Cong Liu, Beijing Institute of Technology, CHINA; Prof. Dr. Hui Liu, Beijing Institute of Technology, CHINA