基于多目标模型预测控制的ANPC并网逆变器

ANPC Grid-connected Inverter Based on Multi-objective Optimal Model Predictive Control

三电平有源中点箝位(ANPC)并网逆变器能够平衡器件功率损耗,有效提高输出电能质量和效率。以三电平ANPC并网逆变器为研究对象,提出了一种多目标优化模型预测控制(MOO-MPC)算法,通过消除三电平输出可能产生多电平跳变的开关状态,可以减少模型的滚动优化次数。通过器件的开关动作损耗线性拟合,实时计算出各开关器件的功率损耗,并在目标成本函数中令不平衡损耗最小化。最后,通过算例分析验证了该控制方法的正确性和有效性。

The three-level active neutral point clamped(ANPC)grid-connected inverter can balance the device power loss and effectively improve the output power quality and efficiency.A multi-objective optimal model predictive control(MOO-MPC)algorithm is proposed for three-level ANPC grid-connected inverter.By eliminating the switching state of three-level output which may produce multi-level jump,the rolling optimization times of the model can be reduced.Through the linear fitting of the turn-on and turn-off losses of the switches,the power loss of each switch device is calculated in real time and the unbalanced loss is minimized in the target cost function.Finally,the correctness and effectiveness of the control method are verified by an example.