基于开关状态函数计算的改进模型预测控制

Improved Model Predictive Control Based on Switching State Function Calculation

传统有限状态模型预测控制(FCS-MPC)利用枚举法选择最优电压矢量,数字控制器运算量大,有学者提出简化的MPC方法,在保持控制系统动稳态性能的同时降低数字计算复杂度,但这些方法均存在采样频率高、稳态功率误差大等缺点。在分析三相AC-DC变换器离散预测模型的基础上,提出一种基于开关状态函数直接计算的模型预测控制方法,以同步坐标系下功率的负共轭为控制变量,根据功率控制误差矢量的长度和所处的扇区,直接计算出各相开关状态函数,经过调制过程完成系统功率的精确跟踪控制。该方法不但保留了传统模型预测控制快速动态响应特性,而且实现了功率跟踪误差最小化。仿真和实验结果均证明了所提方法的有效性和正确性。

The traditional finite state model predictive control(FCS-MPC)uses the enumeration method to select the optimal voltage vector,resulting in a large amount of calculation of the digital controller.Therefore,a simplified MPC strategy is proposed to reduce the computational complexity while maintaining the dynamic and steady-state performance of the original algorithm.However,these methods have the disadvantages of high sampling frequency and large steady-state power errors.With the discrete prediction model of the three-phase AC-DC converter,a model predictive control method based on switching state function calculation directly is proposed in this paper,which takes negative conjugate of complex power in synchronous frame as the control variable.According to the length and the sector of the power control error vector,the switching state function of each phase is calculated directly.The precise tracking control of system power is also carried out through the modulation process.The proposed method can not only retain the advantage of fast dynamic response,but realize the minimization of power tracking error.Both simulation and experimental results verify the effectiveness and correctness of the proposed method.