永磁同步电机转动惯量在线辨识关键技术研究

摘要

针对工程实践中电机转动惯量在线辨识方法中输入参数的异步性和噪声干扰,采用基于模型参考自适应的转动惯量辨识方法,并使用跟踪微分器对输入参数进行预处理来解决以上问题.对模型参考自适应和跟踪微分器这两种关键技术进行研究,为工程应用中普遍得不到理想转动惯量辨识结果的在线辨识方法提供了一个比较好的解决思路.实验表明:经过跟踪微分器预处理后,基于模型参考自适应的转动惯量在线辨识方法能获得比较理想的辨识效果,结果精确、辨识速度较快,具有良好的抗干扰性能.%To deal with the asynchronism and noise interference of inputs of online identification of motor inertia in engineering practice,motor inertia identification based on Model Reference Adaptive Algorithm (MRAA) and Tracing Differentiator (TD) were proposed to preprocess the inputs mentioned above.While the results of inertia identifwation online were not good enough in engineering application,researches on MRAA and TD these key technologies of online identification of inertia of PMSM could offer a better solution.The experiments have shown that the MRAA could get more ideal results after the preprocessing of TD.Result is accurate,identification of fast,and good anti-interference performance.