论文部分内容阅读
根据神经网络理论,提出了一种具有学习功能的参数整定算法。当通过一定数量的样本学习得到成功的 BP 神经网时,调用启发式控制算法闭环控制对象,将获得的响应数据送到 BP 网,即可得到 PID 的整定参数。并以二阶、三阶系统为例,经仿真得到了合适的整定参数。整定方法在分散控制系统中,对大量在线参数辩识计算的处理结果是令人满意的。
According to neural network theory, a parameter tuning algorithm with learning function is proposed. When a successful BP neural network is learned through a certain number of samples, the heuristic control algorithm is called to close the control object and the obtained response data is sent to the BP network to obtain the PID tuning parameters. Taking the second-order and third-order systems as an example, the appropriate tuning parameters are obtained through simulation. Setting method In the decentralized control system, the processing result of a large number of on-line parameter identification computations is satisfactory.