摘要
ABS系统中主缸压力和轮缸压力具有强非线性关系,且传统的控制算法中控制参数不能根据路面的变化及时进行调整,很难在多变的路面工况下取得较好的控制效果。为提高ABS的控制效果,提出一种神经网络模糊PID控制策略。将理想滑移率与实际滑移率之差作为系统输入,经模糊化和归一化后输入神经网络,应用神经网络的自学习和误差反向传播,采用遗传算法优化网络的初始权重,利用BP算法完成网络训练实现模糊规则,使模糊规则的生成转变为加权系数初值的确定和调节,从而表示出模糊规则,寻找一个最佳的PID非线性组合控制率,调整车轮制动力矩,以适应多变的路面。仿真结果表明:该控制策略兼备了神经网络、模糊控制、PID控制的优点,能适应复杂多变的路面。
In the ABS system,the master cylinder and wheel cylinder pressure has strongly nonlinear relationship,and the parameters of traditional ABS algorithm can ＇t adjust in time according to the actual situation of the road,so it is difficult to obtain better braking performance in complex and changeable road conditions. Based on the two-wheel vehicle model,a neuro-fuzzy PID control strategy was put forward in order to improve the ABS control performance. The error between the ideal slipratio and the actual slip ratio was as the system input which was imported to neural network after fuzzification and normalization by fuzzy module. Then the initial weights of the BP neural network were optimized by using genetic algorithm,and the BP algorithm was used to complete the network training to achieve fuzzy rules,so that the generation of fuzzy rules was transformed into the determination and adjustment of the initial value of the weighting coefficient. Thus,the fuzzy rules were expressed in the search for an optimal PID parameter combination. And then the wheel brake torque will be adjusted in order to adapt to the changing pavement. The simulation results indicate that the control strategy has the advantages of neural network,fuzzy control and PID control.
作者
马忠武
倪兰青
陈宇珂
张会琪
林棻
MA Zhongwu;NI Lanqing;CHEN Yuke;ZHANG Huiqi;LIN Fen(Jiangsu Jintan Changdang Lake Amperex Technology Limited,Changzhou 213200;College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2018年第9期14-22,共9页
Journal of Chongqing University of Technology：Natural Science
基金
中国博士后科学基金特别资助项目(2017T100365)
中国博士后科学基金面上资助项目(2016M601799)
中央高校基本科研业务费专项资金资助项目(NT2018002)
南京航空航天大学研究生创新基地(实验室)开放基金资助项目(kfjj20180207)
