【摘要】 针对观光电动汽车驱动系统中异步电机因参数时变导致控制性能恶化问题，提出一种基于模型参考自适应系统(model reference adaptive system, MRAS)与遗传算法(genetic algorithm, GA)的交互在线辨识方法。首先，建立转子磁场定向矢量控制数学模型，通过构造包含Popov超稳定理论自适应机构、转子磁场电流可调模型和电压参考模型的MRAS辨识系统。其次，为解决定子电阻对参考模型输出的影响，设计转子时间常数与定子电阻交互辨识策略，并采用GA优化自适应机构参数。在MATLAB/Simulink平台和基于TMS320F28335 DSP平台实现的控制系统实验结果表明，该方法能有效提升转子时间常数辨识精度，误差从常规4.7%降至2.1%,使电机系统获得优异的动静态性能，为观光电动汽车驱动控制提供了新的解决方案。更多还原

【Abstract】 To address the control performance degradation caused by time-varying parameters in induction motor drive systems for sightseeing electric vehicles, an interactive online identification method based on a model reference adaptive system(MRAS) and a genetic algorithm(GA) is proposed. First, a mathematical model for rotor field-oriented vector control is established. An MRAS identification system is constructed, comprising an adaptive mechanism based on Popov’s hyperstability theory, an adjustable current model for the rotor flux, and a voltage reference model. Second, to mitigate the influence of stator resistance on the reference model output, an interactive identification strategy for the rotor time constant and stator resistance is designed, with GA used to optimize the adaptive mechanism’s parameters. Experimental results from both the MATLAB/Simulink simulation platform and a control system implemented on a TMS320F28335 DSP platform demonstrate that the proposed method effectively improves the identification accuracy of the rotor time constant, reducing the error from 4.7%(conventional method) to 2.1%. This enables the motor system to achieve excellent dynamic and static performance, providing a novel solution for the drive control of sightseeing electric vehicles.更多还原