基于神经网络的电涡流缓速器制动力矩特性研究  

Braking Torque Analysis of Eddy Current Retarder Based on Neural Network

作　　者：王生发[1] 潘敏[1] 马力均[2] 

机构地区：[1]广东五邑大学机电工程系,广东江门529020 [2]浙江大学现代制造工程研究所,杭州310027

出　　处：《农机化研究》2009年第3期24-26,共3页Journal of Agricultural Mechanization Research

基　　金：国家863计划项目(2005AA411910)

摘　　要：针对电涡流缓速器在高速时计算力矩同电涡流缓速器的实际输出力矩存在较大偏差的缺点,提出了一种基于神经网络的电涡流缓速器制动力矩模型,研究了现有的电涡流缓速器控制系统,提出了基于神经网络的电涡流缓速器的PWM控制系统,通过实验说明了基于神经网络模型对实际制动力矩曲线的逼近效果非常有效。Real output torque of eddy current retarder is compared at high velocity,the error of the mathematical model＇s computation torque is big. A braking torque model based on neural network model is proposed. The neural network model of eddy current retarder is established in terms of the braking torque＇s characteristics. The existing control system of eddy current retarder is studied and a novel PWM control system based on neural network model is proposed. The capability of neural network model is shown in terms of the experiment and the braking torque model based on neural network model is more valid.

关 键 词：电涡流缓速器 制动力矩 神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]