Shared steering control for human–machine co-driving system with multiple factors

Abstract

To guarantee the best driving experience and status of the driver, further improve the driving safety of human–machine co-driving vehicles, reduce the conflict between the driver and controller, and weaken the impact of the driver's uncertain behaviours, a human–machine co-driving system with a dynamic weight allocation model is designed. First, a human–machine co-driving system is built with a fixed allocation coefficient. Evaluation indexes based on the degree of participation of the driver and driving safety are proposed. Subsequently, several important influencing factors affecting weight allocation are analysed, including driving characteristics, driving states, and changes in the controller parameters. The results show that the impact of these factors can be weakened by the designed system. However, a good driving experience of the driver cannot be guaranteed. In addition, a conflict between the driver and controller still exists. Next, a model of dynamic weight allocation considering the volatility of the driving characteristics and states of the driver is proposed. Further, the human–machine co-driving system is modified by considering the influence of changes in controller parameters and external interference. Finally, the validity of the designed model of dynamic weight allocation and the modified system were verified by simulation. The results show that the modified system could improve the driving experience and safety better than a system with a fixed allocation coefficient. In addition, the modified system has a better anti-interference ability and lower sensitivity to interference.