基于RBF网络的高速电磁开关阀模型辨识  

Model Identification of High Speed on/off Solenoid Valve Based on RBF Network

作　　者：泮健[1] 施光林[1] 

机构地区：[1]上海交通大学机电控制与物流装备研究所,上海200240

出　　处：《机床与液压》2011年第17期92-94,共3页Machine Tool & Hydraulics

基　　金：国家自然科学基金资助项目(50775137)

摘　　要：简介高速电磁开关阀常用模型,分析其特点;给出RBF神经网络用于非线性系统辨识的一般结构与步骤;分析了使用该网络进行高速电磁开关阀模型辨识的结果,同时使用一些未作为训练样本的数据进行验证,并与理想公式的计算结果进行对比。结果表明:即使在高速开关阀的电气与结构参数未知的情况下,仍然可以使用RBF网络辨识出适合实际应用的高速电磁开关阀模型。The frequently-used models of high speed on/off solenoid valve and their characteristics were introduced and analyzed. The general structure and steps of RBF neural networks employed for non-linear system identification were depicted. As an example, the RBF network was used for the model identification of high speed on/off solenoid valve. The experiments were made and the results were classified into two sets, one set for identification and the other for validation. The identified model was testified through the comparison to the calculation results from ideal flow rate equation. The results show that the identified model using RBF network can reflect the non-linear character of high speed on/off solenoid valve even if the electrical and mechanical parameters are not known.

关 键 词：RBF神经网络 高速电磁开关阀 模型辨识 

分 类 号：TH137[机械工程—机械制造及自动化]