【摘要】 针对传统高频电压注入法中解调过程复杂、带宽局限、估计精度易受磁路饱和及交叉耦合效应影响等缺陷,提出一种基于离散电流特征分析和电感误差在线辨识的内置式永磁同步电机(IPMSM)无传感器优化方法。首先,建立考虑交叉饱和效应的高频电压模型,将正交的高频方波电压信号注入静止坐标系,设计梳状滤波器,分离出高频响应电流,并利用相邻采样时刻的高频电流解调出转子位置的初定值。在分离和解调过程中消除了滤波器带来的劣势,并且计算量小。然后,分析电感参数与估计误差的关系,提出针对正交方波注入法的电感辨识方法和转子位置估计误差补偿策略。最后,将估计误差辨识结果补偿到转子位置的初定值中,提高了估计精度。在15kW电机控制平台上的实验结果证明了该策略的可行性和有效性,而且该方法简单可靠,不受带载时电感参数变化的影响。更多还原

【Abstract】 Aiming at the defects of traditional high-frequency voltage injection method, such as complex demodulation process, bandwidth limitation and estimation accuracy susceptible to magnetic circuit saturation and cross-coupling effect, an interior permanent magnet synchronous machine(IPMSM) sensorless optimization algorithm based on discrete current characteristic analysis and on-line identification of inductance error is proposed. Firstly, a high-frequency voltage model considering crosssaturation effect is established, and two orthogonal high-frequency square-wave voltage signals are injected into the stationary coordinate system, and the high-frequency responsecurrentsare separated by the designed comb filter. A comb filter is designed to separate the high-frequency response current, and the initial value of the rotor position is demodulated by using the high-frequency current at the adjacent sampling time. In the process of separation and demodulation, the disadvantages of the filter are eliminated, and the amount of calculation is small. Then, the relationship between inductance parameters and estimation error is analyzed. The inductance identification method and rotor position estimation error compensation strategy for orthogonal square-wave injection method are proposed. Finally, the estimated error identification result is compensated to the initial value of the rotor position, which improves the estimation accuracy. The experiment on the 15 kW motor control platform proves the feasibility and effectiveness of the strategy. The experimental results show that the method is simple and reliable, and is not affected by the change of inductance parameters under load.更多还原