摘要
通过模仿专家整定控制参数的启发式过程,开发了一种在线实时的专家式控制参数自学习算法。它是一个实时专家系统。阐述了基本思想,介绍了各部分的具体设计,还在计算机上进行了性能验证。
By imitating the heuristic process where an expert settles the regulation gains, an expert-like real-time on-line self-learning algorithm for the regulation gains was developed, which is a real-time expert system. The fundamentals of this algorithm were presented with detailed description of its designing procedure. Finally, the performance demonstration of the system was carried out on computer.
关键词
专家系统
学习系统
最佳控制
self-adjusting systems
learning systems
learning control
pattern recognition
optimal control. real-time expert systems
frame representation
evolving model of performance evaluation
