摘要
为提高静变电源输出电压的质量,采用了优化模糊神经网络PID控制器代替模糊PID控制,所采用的优化模糊神经网络充分融合了模糊逻辑和神经网络两者的优点,使推理速度加快,并通过在系统运行时神经网络不断地增加和完善模糊控制规则,单神经元通过自学习调整控制因子,提高了系统控制的精度。将该方法和PID稳态控制性能的优势相结合,实时地对系统控制量进行调整。在MATLAB/SIMULINK环境下,对于优化模糊神经网路PID和模糊PID在静变电源控制中的应用分别进行了仿真。仿真分析结果表明,经过BP神经网络和单神经元网络学习后,控制器具有良好的控制性能和自适应能力,很好地满足了系统的鲁棒性、快速性的要求。
The optimized Fuzzy Neural PID controller plan is proposed to increase the quality of the Static Inverter output voltage,which has advantages of fuzzy logic and neural network,and the reasoning is speeded up,and the fuzzy control rules are also increased continually.The control factor is adjusted from the single neuron auto-study,and the accuracy of the system control is improved when the system is running.This plan combines with the steady performance advantages of PID control,so it can improve the performance of the system,making the real-time concurrent adjustment to the system control measures.Based on MATLAB/SIMULI NK,the system was simulated under the circumstance of the Static Inverter.Compare with the fuzzy PID control,the analysis of the optimized fuzzy neural PID controller simulation result shows that the control system has good performance of adaptive capacity,which meets the high robust and the requirements of rapid in the system.
出处
《机械设计与制造》
北大核心
2012年第11期46-48,共3页
Machinery Design & Manufacture
关键词
静变电源
优化模糊神经网络
单神经元
PID控制
仿真
Static Inverter
Optimized Fuzzy Neural Network
Single Neuron
PID Control
Simulation
