摘要
针对平衡重式叉车底盘各子系统间的干涉和耦合特性,利用非线性系统的神经网络逆系统方法进行叉车主动后轮转向(ARS)与直接横摆力矩控制(DYC)的解耦控制.在分析底盘系统可逆性的基础上,确定解耦变量配对关系,建立BP(Back Propagation)神经网络逆系统模型并串联到原底盘系统前,使叉车底盘系统解耦成两个独立的伪线性系统;设计PD(Proportion Differentiation)闭环控制器并与神经网络逆系统组成复合控制器,并进行仿真验证.仿真结果表明:神经网络逆系统解耦控制策略能够消除底盘各子系统间的干涉和耦合,提升叉车的状态跟踪和操纵稳定性.
With regard to the interference and coupling properties among counterbalanced forklift chassis subsystems,the decoupling control on active rear steering(ARS)and direct yaw moment control(DYC)is first proposed using the neutral network inverse system method for nonlinear systems.With analysis on the chassis system invertibility,the matching relationship of decoupling variables is then determined.By establishing a BP(Back Propagation)neural network inverse system model to connect in series to chassis system,the forklift chassis system is next decoupled into two independent pseudo-linear systems.Finally,a close-loop PD(Proportion Differentiation)controller is designed and combined with inverse system into a compound controller.Therein,it is verified from simulation results that the decoupling control strategy for neural network inverse system can eliminate the interference and coupling among chassis subsystems so as to promote forklift conditional tracking and operational stability.
出处
《中国工程机械学报》
2014年第5期384-390,共7页
Chinese Journal of Construction Machinery
基金
国家自然科学基金资助项目(51205101)
安徽省"十二五"科技攻关资助项目(12010202032)
关键词
叉车底盘
主动后轮转向
直接横摆力矩控制
解耦控制
神经网络逆系统
forklift chassis
active rear steering
direct yaw moment control
decupling control
neutral network inverse system
