Intelligence Control Based on Self-Correcting Fuzzy-Neural Networks for Brushless DC Motor Drive
-
摘要: 提出了一种自校正模糊神经网络控制器(SCFNNC)来实现无刷直流电动机起动、调速、制动等各运行阶段的性能指标.该SCFNNC是采用调整系统增益参数的方法完成较完善的控制规则的.重点研究了系统自校正增益参数的确定方法,模糊控制器的设计,人工神经网络实现模糊控制规则的方法等.自校正增益参数是根据系统对超调量、转速稳态误差、动态速降的期望值来确定的.设计模糊控制器时是根据系统的性能指标,确定出合适的模糊控制规则表,用于训练神经网络.为使系统的性能达到最佳,采用了自校正模糊神经控制、开关控制和比例控制相结合的复合控制方法,通过数学仿真证实配备SCFNNC的系统具有优良的动、静态特性,及较强的鲁棒性.Abstract: It presents a self-correcting fuzzy-neural networks controller (SCFNNC) to realize the requirement of the start,regulating the speed,resisting disturbance and the brake for brushless DC motor.The SCFNNC implements good controlling rules by the method of regulating self-conrrecting gain parameters for the system.The paper emphatically develops designing method of self-correcting gain parameters,the design of fuzzy controller and the method to realize the rules of fuzzy control by artificial neural networks etc.The self-correcting gain parameters depend on expected values of the overshoot,steady-state error and dynamic velocity-drop for the system.According to systematic requirement fuzzy controller is designed,suitable rule table of fuzzy control is also determined,which is used to training neural networks.In order to reaching optimal performances complex control of combining self-correcting fuzzy-neural networks with Bang-Bang and proportional control is applied.It is shown by digital simulation that the system with SCFNNC has excellent dynamic and static performances,good robustness and ability for resisting disturbance.
-
Key words:
- D.C. electric machines /
- fuzzy logic /
- neurons networks /
- controllers /
- self-correction
-
1. 余永权,曾 碧.单片机模糊逻辑控制.北京:北京航空航天大学出版社,1995 2. 张立明.人工神经网络的模型及其应用.上海:复旦大学出版社,1993 3. Buja G S,Todesco F.Network implementation of a fuzzy logic controller.IEEE Transactions on Industrial Electronics,1994(12):663~665
点击查看大图
计量
- 文章访问数: 2856
- HTML全文浏览量: 134
- PDF下载量: 245
- 被引次数: 0