Applications of Goal Programming in Receding-Horizon Optimization and Online Identification of Predictive Control
-
摘要: 针对有约束多目标多自由度预测控制问题,应用目标规划方法,提出了一种既适合于 参数模型又适合于非参数模型的在线滚动优化策略,并且通过计算饥仿真研究,验证了该方法 的有效性.然后,对于参数模型预测控制问题,提出了一种抗扰动的最小绝对值辨识算法.由于 该辨识算法可用目标规划快速求解,因此可作为慢时变工业过程控制的在线辨识算法.
-
关键词:
- 有约束多目标多自由度预测控制 /
- 在线滚动优化 /
- 目标规划法 /
- 抗扰动最小绝对值辨识算法
Abstract: An online receding-horizon optimal strategy for constrained predictive control with multi-goals and multi-degrees of freedom is proposed in terms of goal programming principles, which is adaptable to parameteric predictive control as well as non-parameteric predictive control. The effectiveness and performance of the approach are demonstrated by a computer simulation example. Then, based on the least absolute errors between referential outputs and predictive ones, a disturbance-rejection identification algorithm is presented for parameterized predictive control. Since the identification algorithm can be efficiently solved by employing a goal programming, it is suitable for online identification of slow time-varying industrial process.
计量
- 文章访问数: 2966
- HTML全文浏览量: 78
- PDF下载量: 1165
- 被引次数: 0