摘要
常规PID控制器以其算法简单、可靠性高等优点,在工业生产得到了广泛应用。但是,PID控制器存在控制参数不易在线实时整定、难于对复杂对象进行有效控制等不足。利用神经网络自学习、自适应和非线性映射等特点,将神经网络和PID控制相结合,形成一种PID神经网终挖制系统,可对工业中使用的具有大时滞、慢时变、非线性特点的电炉系统进行有效辨识与控制。
General PID controller, because its algorithm is simple and high reliabihty,so has been widely used in industrial production. However, PID controller, there is not easy to line real-time control parameter tuning, is difficult for complex objects such as lack of effective control. Using neural network self-learning, adaptive and nonlinear mapping characteristics of neural network and PID control combmed to form a PID neural network control system can be used in industry with a large time lag, slow time-varying, nonlinear characteristics of electric systems for effective identification and control.
作者
任慧
王伟智
REN Hui, WANG Wei-zhi(Institute of Automation, Fuzhou University, Fuzhou 350002, China)
出处
《电脑知识与技术》
2009年第10期8028-8030,共3页
Computer Knowledge and Technology
基金
国家自然科学基金项目(60675058),福建省自然科学基金项目(2009J01248)
关键词
PID神经网络
智能控制器
滞后系统
时变系统
电炉控制系统
PID neural network
intelligent controller
time lag system
time-varying system
electric cooker systems