摘要
针对实际工业过程中普遍存在的非线性特性,采用T-S模糊模型来描述复杂的非线性系统.利用模糊聚类算法在线竞争学习模糊规则的输入区域中心,并按预先规定的规则之间的重叠度在线确定每条规则的输入区域半径,而模糊规则结论中的参数则由递推最小二乘(RLS)算法得到,得到的模糊局部线性化模型作为每个时刻的受控自回归积分滑动平均(CARIMA)模型.结合广义预测控制(GPC)的思想和有限脉冲响应滤波器,提出了一类新型的模糊预测PID控制器的实现方法,解决了一般预测PID控制器设计当中对系统模型阶次的限制问题.仿真算例说明了该方法的有效性和实用性.
As most industrial processes are nonlinear, the proper T-S fuzzy model was obtained by online recursive fuzzy identification method and then applied to describe the complex nonlinear systems. Fuzzy cluster algorithm was used to determine the center input area of the fuzzy rules, and the predefined degree of over lapping between rules was used to determine the input-zone radius for the fuzzy rules. After the rules' consequent parameters were obtained by the recursive least squares(RLS) algorithm, the fuzzy local linearization model was used as a CARIMA-like model. Finally, by combining the principles of the generalized predictive control(GPC) and adaptive FIR filter, a novel predictive PID controller algorithm was developed. The method solves the problem of restriction on the system's order. Simulation results show the effectiveness and practicability of the proposed method.
出处
《浙江大学学报(工学版)》
EI
CAS
CSCD
北大核心
2004年第7期826-830,共5页
Journal of Zhejiang University:Engineering Science
基金
国家杰出青年基金资助项目(60025308)
高等学校优秀青年教师教学和科研奖励基金资助项目.
关键词
输出误差预测
模糊预测PID
广义预测控制
有限脉冲响应滤波器
output-error prediction
fuzzy predictive PID
generalized predictive control (GPC)
finite impulse response (FIR) filter