江浩斌++刘海++耿国庆++唐斌



存在操纵稳定性差的缺点,提出了一种旁通流量控制式电控液压转向系统((Electrical Controlled Hydraulic Power Steering,ECHPS)。建立了该转向系统核心部件电液比例阀数学模型,设计了ECHPS系统的助力控制策略和助力特性曲线,为了消除被控系统受到参数不确定性和外界干扰的影响,采用神经网络与自适应动态面技术相结合的算法设计了一种新型控制器。通过理论与仿真分析证明了所设计的自适应神经网络动态面控制器不仅响应快、跟踪效果好、控制精度高,而且能够实现汽车低速时的转向轻便性和高速时的良好路感要求。
关键词:电控液压转向系统;控制策略;助力特性曲线;神经网络;动态面控制;仿真
中图分类号:U463.4 文献标识码:A DIO:
Research on Control Strategy for ECHPS Based on
Adaptive Neural Network Dynamic Surface Algorithm
Jiang Haobin, Liu Hai, Geng Guoqing, Tang Bin
(School of Automobile and Traffic Engineering, Jiangsu University, Zhenjiang 212013,Jiangsu,China)
Abstract: A bypass flow-controlled electrical controlled hydraulic power steering system was introduced, for the problem of poor steering caused by HPS with fixed assist characteristics applied in heavy commercial vehicles. Electro-hydraulic proportional valve model of ECHPS core component was established. The control strategy and assistance characteristic curve of the steering system were designed, for the controlled system containing the parameter uncertainty and the impact of outside interference, a method combined neural network and adaptive dynamic surface control was used to design a new controller. .By theoretical analysis and simulation results indicated that adaptive neural network dynamic surface controller was not only fast response, good tracking performance, high control accuracy, but also could achieve the requirements of low-speed light and high-speed good road sense when the vehicle steering.
Key words: electrical controlled hydraulic power steering; control strategy; assistance characteristic curve; neural network(NN); dynamic surface control; simulation
流量控制式电控液压助力转向系统是在传统的液压助力转向系统基础上附加电液控制装置调节液压系统的流量实现助力可变,由于其结构简单,液压功率
大,技术风险小,助力可变,操纵稳定性好而被国内外众多学者研究[1-3]。在实际应用中,流量控制式ECHPS中电液比例控制系统的动态特性具有较强的非线性,且会受到参数不确定性,未建模动态以及各种外界干扰等影响[4-5]。因此,针对ECHPS系统的特点和实际要求研究其控制问题非常有必要。
后推法是非线性控制领域广泛应用的技术[6-7],然而该方法存在着计算膨胀问题,
为此提出了动态面控制法,该方法在后推法基础上通过引入一阶低通滤波器能够有效地解决后推法计算膨胀问题,简化了控制器的设计,并且近年来与神经网络控制或模糊控制相结合解决了一大类非线性不确定系统的控制问题[8-9]。……