车辆半主动悬架系统模糊神经网络控制研究  

On Fuzzy Neural Network Control of Vehicle Semi-active Suspension System

作　　者：贝绍轶[1] 赵景波[1] 张兰春[1] 陈龙[2] 

机构地区：[1]江苏技术师范学院机械与汽车工程学院,常州213001 [2]江苏大学汽车与交通工程学院,镇江212013

出　　处：《系统仿真学报》2010年第12期2952-2956,3007,共6页Journal of System Simulation

基　　金：国家自然科学基金(50475121);高等学校博士学科点专项科研基金(20050299005)

摘　　要：悬架系统对车辆平顺性具有重要的影响,通过预瞄控制在后轮处提前预测路面不平度,用于解决半主动悬架模糊神经网络控制存在的时滞问题。建立了1/2车辆模型和路面输入模型,设计了基于预瞄控制的半主动悬架模糊神经网络控制结构,并进行了白噪声输入仿真分析。结果表明:预瞄控制后的车身加速度峰值和标准差比被动系统分别减少了61.61%和44.28%,比模糊控制的悬架系统分别减少了21.23%和21.20%;预瞄控制后的质心加速度峰值和标准差比被动系统分别减少了35.21%和57.81%,比模糊控制的悬架系统分别减少了7.83%和20.10%。后轮处车身加速度和质心垂直加速度均有明显减小,较好改善了悬架系统适应道路的性能,有效缓和了车辆的振动和提高了汽车的行驶平顺性。Suspension system has important effect on vehicle ride comfort.Wheelbase preview control method could be used to forecast the road surface roughness at the front wheel,and to be used to solve the delay problem in fuzzy neural network controlled semi-active suspension system.The 1/2 vehicle model and road input model was established,and the fuzzy neural network control structure based on wheelbase preview control theory was designed.The white noise input simulation was carried out and the results show that the peak and standard deviation of body acceleration are separately decreased by 61.61% and 44.28% compared with passive suspension system,and are 21.23% and 21.20% compared with fuzzy controlled suspension system;on the other hand,the peak and standard deviation of vertical acceleration at CG are separately decreased by 35.21% and 57.81% compared with passive suspension system,and are 7.83% and 20.10% compared with fuzzy controlled suspension system.The body vertical acceleration at rear wheel and the vertical acceleration at CG are significantly decreased to adapt to the suspension performance,which helps to effectively easy the vehicle vibration and to improve the vehicle ride comfort.

关 键 词：车辆 半主动悬架 预瞄控制 模糊神经网络控制 平顺性 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置] U463.4[自动化与计算机技术—控制科学与工程]