SynRM在线电感解耦辨识算法

Online inductance decoupling identification algorithm for SynRM

关键词:  SynRM  饱和及耦合特性  模型解耦  电感  在线辨识

Key words:  synchronous reluctance motor  saturation and coupling characteristics  model decoupling  inductance  online identification

摘要:

为分析同步磁阻电机(synchronous reluctance motor,SynRM)饱和及耦合特性并实现扰动小且误差低的SynRM电感辨识,提出一种SynRM在线电感解耦辨识算法。首先,描述了电机饱和及耦合特性对SynRM电压方程和磁链方程的影响,分析了SynRM磁场饱和及耦合效应以解释电感的饱和及耦合特性,同时引入耦合角设计SynRM的解耦模型,并从解耦角度分析电机饱和及耦合特性;其次,设计基于虚拟轴系等效阻抗模型的耦合角及电感在线辨识策略;最后,将所提算法应用于3 kW的SynRM实验及测试平台,通过不同工况下辨识及对比实验验证所提算法的有效性。实验结果表明:所提算法可有效实现SynRM在线电感解耦辨识,且耦合角及电感辨识结果误差在允许范围之内；同时,电感随着电流的增大而减小,耦合角随着电流的增大而增大,耦合角及电感辨识结果的变化趋势也验证了电机饱和及耦合特性分析的准确性。相比于其他电感辨识算法,所提算法对芯片算力要求不高,在简化电感运算的同时可实时跟随电机控制并输出准确的电感辨识值。

Abstract:

To analyze the saturation and coupling characteristics of SynRM and realize inductance identification with small disturbance and low error, an online inductance decoupling identification algorithm for SynRM is proposed. The influence of magnetic saturation and coupling on the voltage and flux linkage equations is first described to interpret the saturation and coupling characteristics of the inductance, and a decoupling motor model is developed by introducing a coupling angle. This model enables the analysis of saturation and coupling effects from a decoupling perspective. Then, an online identification strategy based on a virtual-axis equivalent impedance model is designed to identify both the coupling angle and inductance in real time. The proposed method is validated on a 3 kW SynRM experimental platform under various operating conditions. Experimental results demonstrate that the proposed algorithm effectively realizes online inductance decoupling identification, with identification errors for both the coupling angle and inductance within acceptable limits. Moreover, the inductance decreases with the increase of current, and the coupling angle increases with the increase of current. The changing trends of coupling angle and inductance identification results also verify the accuracy of motor saturation and coupling characteristic analysis. Compared to other inductance identification algorithms, the proposed algorithm does not require high chip computing power. While simplifying inductance calculations, it can also follow motor control in real time and output accurate values.