基于BP神经网络PID控制的主动悬架仿真分析

为研究主动悬架的控制算法,采用基于BP神经网络的一种自适应PID控制算法来搭建主动悬架控制系统。以某轿车车型为例,在Simulink软件中建立了以随机路面不平度激励作为系统输入的七自由度整车主动悬架仿真模型。将车身垂向加速度均方根、俯仰角加速度均方根、侧倾角加速度均方根作为主动悬架性能的评价指标进行时域及频域分析。由仿真结果可知,相比于传统的被动悬架,运用该控制算法的主动悬架可显著提升汽车的行驶稳定性与舒适性。

Simulation and Analysis of Active Suspension Based on BP Neural Network PID Control

In order to study the control algorithm of active suspension, an adaptive PID control algorithm based on BP neural network is used to build the active suspension control system. Taking a car as an example, a 7-freedom active suspension simulation model with random road roughness excitation as system input is established in Simulink. The car body vertical acceleration rootmean-square, pitching angular acceleration root-mean-square and rolling angular acceleration root mean square are taken as the evaluation indexes of active suspension performance in time-domain and frequency-domain. The results of simulation show that compared with the traditional passive suspension, the active suspension with this control algorithm can significantly improve the stability and comfort of the car.