论文部分内容阅读
本文使用有序神经网络和改进的模糊控制器构成了一种新型的神经模糊预测控制方法。有序网络学习速度快 ,所需神经元数目少。用事先训练好的有序网络代替传统的预测模型 ,以期增强输出预测的准确性 ;同时 ,用一种改进的模糊控制器代替原有的PID控制器 ,增强系统的鲁棒性。仿真结果表明 ,所提出的神经模糊预测控制方法可以获得理想的控制效果
In this paper, an orderly neural network and an improved fuzzy controller form a new neural fuzzy predictive control method. Orderly network learning speed, the required number of neurons less. In order to enhance the accuracy of output prediction, a pre-trained ordered network is substituted for the traditional prediction model. At the same time, an improved fuzzy controller replaces the original PID controller and enhances the system robustness. The simulation results show that the proposed neural fuzzy predictive control method can achieve the desired control effect