English

Journal of Chinese Agricultural Mechanization

Journal of Chinese Agricultural Mechanization ›› 2023, Vol. 44 ›› Issue (10): 129-136.DOI: 10.13733/j.jcam.issn.2095-5553.2023.10.019

Previous Articles     Next Articles

Parameter optimization of electric tractor dual motor coupling drive based on DDPG algorithm

Zhang Yonggang1, Liu Mengnan2, Xu Wenxiang1, Xu Liyou1   

  • Online:2023-10-15 Published:2023-11-09

基于DDPG算法的电动拖拉机双电机耦合传动参数优化

张勇刚1,刘孟楠2,胥文翔1,徐立友1   

  1. 1. 河南科技大学车辆与交通工程学院,河南洛阳,471003; 2. 洛阳拖拉机研究所有限公司,河南洛阳,471039
  • 基金资助:
    河南省高等学校重点科研项目(22B416001);河南省高校科技创新团队支持计划项目(24IRTSTHN029)

Abstract: In view of the optimization of transmission ratio between drive motor and PTO motor in the dual motor coupling drive system of electric tractor, a dynamic optimization method for speed ratio of transmission system considering the coupling law of motor was proposed. The objective functions of minimum transmission system speed ratio, highest motor coupling efficiency and shortest time to reach the design working condition were designed, Simscape was used to build the tractor driveline model, the driver model, motor model, controller model and tractor body model created in Simulink to form a simulation environment. DDPG algorithm was used to iteratively calculate  the optimal value of the driveline in the continuous motion space. Compared with the static calculation, the optimization results showed that in the transportation mode, the transmission ratio decreased by 0.7, the acceleration time of the tractor from 0 to 20 km/h advanced by 0.3 s, and the overall efficiency of the motor increased by 10%. In the rotary tillage mode, the transmission ratio decreased by 1.6, the acceleration time of the tractor from 0 to 15 km/h advanced by 1 s, and the overall motor efficiency increased by 5%.

Key words: electric tractor, double motor coupling transmission, parameter matching, DDPG algorithm, physical modeling

摘要: 针对电动拖拉机双电机耦合传动系统中,驱动电机与PTO电机耦合驱动时传动系统各档之间传动比优化的问题,提出一种考虑电机耦合规律的传动系统速比动态优化方法。通过设计传动系统速比最小,电机耦合效率最高和达到设计工况时间最短的目标函数,借助Simscape搭建拖拉机传动系统模型与Simulink中创建的驾驶员模型、电机模型、控制器模型和拖拉机车身模型组成仿真环境;采用DDPG(强化学习)算法在连续动作空间内对传动系统进行迭代求出最优值。与静态优化相比结果表明:在运输模式下,传动比下降0.7,拖拉机从0~20km/h加速时间提前0.3s,电机综合效率提升10%。在旋耕模式下,传动比下降1.6,拖拉机从0~15km/h加速时间提前1s,电机综合效率提升5%。

关键词: 电动拖拉机, 双电机耦合传动, 参数匹配, DDPG算法, 物理建模

CLC Number: