基于前馈解耦控制的有源中点钳位五电平模型预测控制

Research on Model Predictive Control of ANPC-5L Permanent Magnet Synchronous Motor Based on Feedforward Decoupling

为解决五电平变换器电容电压不均衡且控制算法运算量过大的问题,该文结合前馈解耦控制方法提出了一种有限控制集模型预测控制优化算法。以变换器电容均压控制和永磁同步电机dq电流控制为目标选择最优电压矢量,采用前馈补偿对中点电压和悬浮电容电压进行解耦控制,并将解耦后的值代入模型预测控制算法中,以实现电流跟踪的目的。通过突加突卸负载和正反转切换仿真和实验,从响应时间和悬浮电容电压偏移度等多角度验证了该控制策略的有效性。

High voltage and large capacity power converter is an important part of power system.However,due to the current voltage level of power devices,the traditional three-level Neutral point clamped topology(NPC)is difficult to apply to high voltage converters above 6kV.In recent years,Active Neutral Point Clamped five-Level(ANPC-5L)has received extensive attention at home and abroad.It has the advantages of few power components,high system efficiency and the ability to achieve capacitive voltage balance control in the full power factor range.In order to solve the problem that the SVPWM algorithm of ANPC-5L topology is too complex and difficult to realize capacitor voltage decoupling control,an organic combination of feedforward decoupling control and model predictive control algorithm is presented in this paper to realize multi-objective control of motor.The effectiveness of the proposed algorithm is verified by simulation and experimental comparison.Firstly,the equivalent SVPWM algorithm of ANPC-5L topology is realized by using single vector MPC and carrier phase shift.To reduce the computational load of MPC,the optimization method of vector selection is designed.Some voltage vectors that do not meet the constraint conditions are eliminated to reduce the number of scrolling optimizations.Therefore,the constraint conditions for the selection of the output voltage vector of the converter are set as follows:the maximum single-phase level hop variable is 1,cross-level hop is not allowed.Thus,the voltage vectors involved in the calculation of a single control period are reduced from 125 to 27.Therefore,the calculation amount of the algorithm is reduced and the operation efficiency of the algorithm is improved.In order to achieve the balance between the bus capacitor voltage and the suspended capacitor voltage,the duty ratio of the switching function is controlled by double modulated signals respectively.The decoupling control of midpoint current and suspension capacitance current is realized by feedforward compensation.In each control period,27 possible discrete modulation signal combinations are substituted into the current prediction equation to calculate the variable predicted value of the next period.Simulation and experimental results show that the proposed model predictive control algorithm can realize the multi-objective optimization control of motor drive and multilevel voltage balancing.Therefore,the drive performance of the motor is ensured and the capacitor voltage is equalized effectively.