摘要
解决船舶发电机在负荷突变尤其是大负荷突变时能够稳定发电机端电压是控制系统需要解决的基本问题.为此提出和使用一种新型智能PID控制系统,其中PID控制器由三层前馈神经网络组成.利用神经网络的自学习能力,PID控制器的参数能够根据系统动态特性通过神经网络权系数进行自行调整,其特点是结构简单、工作稳定.仿真结果表明,该智能PID控制器比传统的PID控制器具有对扰动响应速度快和具有更好的控制效果和更好的鲁棒性,能更好地稳定船舶发电机端电压.该智能PID控制器已用于船舶发电机控制系统中,并取得满意的效果.
It is a basic problem in the control system to stabilize marine generator voltage while the load changes, especially, while there is a drastic load change. The generator voltage was stabilized by its excitation control system. Generator and excitation control system were quite complex nonlinear dynamic systems. To solve the problem, a new intelligent PID controller was proposed and applied. The PID controller was composed of three-layer neural network. The neural network has self-learning ability. According to the system dynamic characteristic, the PID parameters can be self-adjusted by automatically tuning neural networks weights. The PID controller structure was simple but worked steadily. The simulating results indicated that the intelligent PID controller possessed faster response speed and better control effects than the traditional PID controller. The marine generator terminal voltage can be better stabilized. The PID controller has been successfully used in marine generator control system and a satisfactory result has been achieved.
出处
《大连海事大学学报》
CAS
CSCD
北大核心
2006年第2期5-8,16,共5页
Journal of Dalian Maritime University
基金
交通部交通应用基础研究资助项目(200332922505)
关键词
船舶工程
发电机电压
励磁控制系统
PID控制器
神经网络
ship engineering
marine generator voltage
excitation control system
PID controller
neural networks
作者简介
程木军（1965-），男，辽宁大连人，工程师
