基于DRNN神经网络的挖掘机伺服系统参数辨识  

Parameter identification of servo system for excavator based on DRNN

作　　者：黎波[1] 严骏[1] 郭刚[1] 钱海波[1] 张梅军[1] 

机构地区：[1]解放军理工大学野战工程学院,江苏南京210007

出　　处：《解放军理工大学学报（自然科学版）》2013年第1期75-78,共4页Journal of PLA University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(51175511)

摘　　要：为有效分析挖掘机电液伺服系统,提高依据模型设计控制器的精度,建立了系统状态空间模型。针对模型中的不确定参数,提出了基于对角回归神经网络的系统辨识策略。通过神经网络在线学习得到系统Jacobian信息,将实测信息代入含Jacobian信息与待辨识参数的线性方程,利用最小二乘法求得未知参数。实验表明,辨识模型能从初始阶段的微小误差逐渐地逼近实际系统,所提出的方法能有效辨识系统参数。To analyze the servo system of the excavator effectively and design the controller precisely, the state space model was built. According to the characteristics of the parameter uncertainties, an identifica- tion method based on diagonal recurrent neural network （DRNN） was developed. The Jacobian informa- tion was achieved by on-line learning through neural networks, and least square method was proposed to i- dentify the unknown parameters by using the achieved information. The comparison experiment demon- strated that the identified model can approximate actual system gradually,and that the proposed method in this paper can meet the need for identifying the unknown parameters.

关 键 词：挖掘机 电液伺服系统 对角回归神经网络 Jacobian信息 参数辨识 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]