基于加速进化规则的模糊神经网络理论在汽门控制器中的应用

THE FUZZY NEURAL NETWORKS VALVE CONTROLLER BASED ON ACCELERATED EVOLUTIONARY PROGRAMMING

本文提出了一种训练模糊神经网络的加速进化规则算法。该算法是进化规划算法的一种改进 ,适合于多维高精度的数值优化问题。并用该算法优化模糊神经网络汽门控制器。经仿真验证 ,其寻优速度有了明显的改善 ,得到了较好的控制效果 ,提高了电力系统的稳定性。

In this article,a kind of accelerated evolutionary programming is presented to training the fuzzy neural networks. Accelerated evolutionary programming algorithms is a kind of improved way of evolutionary programming which suits real digital optimize problems of multi dimension highaccuracy. The algorithms is proved by simulation examples, which presents its optimum speed has been improved obviously, and acquire a better control effect.