摘要
分析了一类神经网络智能ＰＩＤ控制算法的机理及该类算法存在的一些不足之处，提出一种基于输出多步预测的改进算法。仿真及应用表明，该方法对非线性和时变对象有较强的适应能力。
A kind of adaptive PID control algorithm is analysed, and the drawbacks of the existing algorithms are commented. As an improvement, a neural network intelligent control algorithm based on multi-step prediction is developed. Simulation and application show that the new control method is more adaptable to the control of time-varying and nonlinear control systems.
出处
《控制与决策》
EI
CSCD
北大核心
1998年第4期311-316,共6页
Control and Decision
基金
东南大学科学基金
关键词
神经网络
智能PID控制
工业控制
neural networks, intelligent PID control, multi-step prediction
