摘要
针对永磁同步电动机矢量控制系统,提出了一种将神经网络与模糊控制相结合的控制方法.通过对神经网络进行训练来记忆模糊控制规则,不需要存储模糊控制表,不依赖被控对象的精确数学模型,而且该方法具有很强的自学习能力,在模型参数发生变化时,可通过调整控制器在线自学习达到最佳效果.仿真结果表明此控制方案是十分有效的,具有响应快、鲁棒性强、较好的动、静态特性等优点,基于神经网络的模糊控制特别适用于结构复杂、干扰大、控制精度要求高的系统.
For a permanent magnet synchronous motor vector control system, a control method combining neural network with fuzzy controller is presented. The trained neural network can memorize fuzzy control rules. So the fuzzy control rules table needn't be stored in the memory. The controller is designed with independency of exact mathematical model and has the strong ability of self-learning. The performances of control system can be improved by the self-leaming on line when the model parameters change. The effectiveness of the proposed controller is verified by simulation. The fuzzy controller based on neural network has the characteristics of quick response, strong robustness and so on. The fuzzy controller based on neural network can work well for the system with complicated structure, strong disturbances and high accuracy.
出处
《沈阳工业大学学报》
EI
CAS
2006年第1期62-64,101,共4页
Journal of Shenyang University of Technology
关键词
永磁同步电动机
矢量控制
模糊神经网络
BP算法
PID控制
permanent magnet synchronous motor
vector control
fuzzy neural-network
BP algorithm
PID control