摘要
感应电机多标量模型具有状态变量是标量且物理意义明确和不需旋转坐标变换等优点;神经网络逆系统适合解决不确定性因素(参数变化和外在扰动等)存在的情况下,感应电机高性能的控制问题。为此,提出基于多标量模型的感应电机神经网络逆控制结构,实现感应电机系统的自适应解耦线性化,进而提高系统控制性能。最后对系统进行了仿真研究和软硬件实现方案讨论,理论分析和仿真表明所提控制结构是有效的。
Abstract：The multiscalar model of induction motor （IM） own the advantages such as the state variables are scalar, the physical meaning of state variables is clear and the rotating coordinate transformation is not nec- essary. The neural network inverse system（NNIS） is suitable to solve control problem of IM with the uncer- tain factors （the parameters variation and external disturbance etc. ）. For that, the NNIC structure based on the muhiscalar model of IM was given,the adaptive decoupling and linearization （D￡L） of IM was realized and the system control performance was improved. At last, the simulation study was done and the software hardware implementations of the proposed system were discussed, theory analysis and the simulation show that the proposed method is valid.
出处
《电气传动》
北大核心
2010年第12期43-47,共5页
Electric Drive
关键词
多标量模型
神经网络逆系统
自适应解耦线性化
感应电机
multisealar model
neural network inverse system
adaptive decoupling and linearization
induc-tion motor
