Skoči na glavni sadržaj

Izvorni znanstveni članak

https://doi.org/10.1080/00051144.2023.2231713

A dynamic weight multi-objective model predictive controller for adaptive cruise control system

Shufeng Wang ; College of Transportation, Shandong University of Science and Technology, Qingdao, People’s Republic of China *
Baokang Zhang ; College of Transportation, Shandong University of Science and Technology, Qingdao, People’s Republic of China
Yadong Yuan ; Soto Automotive Intelligent Equipment Limited Company, Hangzhou, People’s Republic of China
Zhe Liu ; Soto Automotive Intelligent Equipment Limited Company, Hangzhou, People’s Republic of China

* Dopisni autor.


Puni tekst: engleski pdf 3.532 Kb

str. 919-932

preuzimanja: 92

citiraj


Sažetak

Adaptive cruise control (ACC) is recognized as an effective method to improve vehicle safety and reduce driver workload. This paper proposes a whole hierarchical multi-level state ACC system. According to the function of ACC system, the three-level state ACC system is designed and the conversion mechanism between different states is put forward. As for the complex car-following mode, considering the variable headway safety distance and the road adhesion coefficient, the expected safety distance model is established, using the distance error and the speed error as fuzzy input, based on the fuzzy control algorithm, the following mode is obtained; considering vehicle safety, tracking capability and ride comfort, the control objectives are formulated into the model predictive control algorithm. A dynamic weight strategy is proposed to solve time-varying multi-objective control problems, where the weight can be adjusted with respect to different following conditions. The simulation results demonstrate that the car following performance of ACC with the proposed dynamic weighted MPC can provide better performance than that using constant weight MPC, and the multi-level state ACC system can display the control mode more intuitively.

Ključne riječi

Adaptive cruise control; multi-level state; the following mode; model predictive control; dynamic weight

Hrčak ID:

315949

URI

https://hrcak.srce.hr/315949

Datum izdavanja:

10.7.2023.

Posjeta: 307 *