摘要
根据模型参考自适应和直接转矩控制理论,针对常规速度辨识器中的基准模型易受积分初值和漂移问题的影响,造成辨识结果不准确的问题,设计了一种新型的基准和可调模型。并且以此为基础,采用基于神经网络的模型参考自适应辨识电机转速。M ATLAB仿真结果表明,系统具有较好的性能。
According to the theory of model reference adaptive system and the direct torque control, the normal model can always be affected because of the initial value and excursion of integral calculation in conventional speed-identifier, which leads to the wrong result of identifying. This paper has dealt with these problems by devising a kind of new normal model and adjustable model, and based on which, the motor speed is identified by adopting neural network. The MATLAB results show the system with neural network identification model has better performance.
出处
《电气传动》
北大核心
2006年第3期19-22,共4页
Electric Drive
关键词
直接转矩控制
定子磁链
励磁电流
神经网络
无速度传感器
direct torque control（DTC） stator flux excitation current neural network speed sensorless
