电感在线辨识的网侧变换器预测直接功率控制

摘要

An improved predictive direct power control ( SVM P-DPC) algorithm for grid-connected three-phase voltage source converters in doubly-fed wind power system was presented, which combines with the predictive control, the direct power control and space vector pulse width modulation technique. Because the prediction algorithm is sensitive to the filter inductor, a method for online identification inductance was put forward and the inductance identification in prediction model calculation used to improve the prediction precision. The algorithm has good dynamic characteristics. It operates with constant switching and provides sinusoidal-grid-current low harmonic distortion compared with direct power control based on hysteresis control. Experimental platform for doubly-fed wind power system was built. The experimental results showed that the control strategy was feasible and efficient.%以双馈风力发电系统网侧变换器为研究对象，将预测控制、直接功率控制和空间矢量脉宽调制技术相结合，提出一种无差拍预测直接功率控制( SVM P-DPC)策略。预测算法模型对滤波电感参数很敏感，采用一种简单有效的电感参数在线辨识方法，将辨识电感用于预测模型计算，防止当预测模型电感和实际电感参数不匹配时可能出现的预测误差和系统不稳定问题，提高了预测精度。算法具有良好的静态和动态特性。与基于滞环控制的直接功率控制相比，具有开关频率恒定、输出电流正弦度好，谐波含量低等优点。搭建了双馈风力发电系统试验平台，对提出的SVM P-DPC方案和基于滞环控制的直接功率控制方案进行对比试验。试验结果验证了SVM P-DPC策略的可行性和正确性。