Direct Power Predictive Optimal Control For Reduced Matrix Converter

The reduced matrix converter(RMC)uses the high frequency transformer to couple the front and back stages,which not only has excellent network side performance and stable DC voltage,but also realizes the electrical isolation between the power supply and the load side.However,this coupling form also makes RMC topology anti-interference performance weak,and the conventional modulation strategy is unable to solve this problem.In this paper,direct power predictive control(DPPC)method is introduced into RMC to improve the anti-interference performance.However,the conventional DPPC strategy has a large amount of calculation,which will increase the calculation burden of the digital processing chip and deteriorate the RMC input and output performance.Therefore,it is of great theoretical significance and practical value to study the high-efficiency and low-computation DPPC method.The continuous and discrete mathematical model of RMC topology are established,and the cost function with input power as the core is constructed,so as to realize the DPPC strategy of RMC.The RMC two-way flow of energy under DPPC strategy is realized by improving the strategy.Although the DPPC strategy has a good control effect and can effectively suppress external disturbances,the algorithm is more complex,and the computational burden is heavy,which requires higher requirements on the controller.For this problem,this paper studied a fast vector selection-power predictive control(FVS-PPC)method.Based on the complex power theory,the RMC complex power differential continuous mathematical model and complex power differential discrete mathematical model are constructed,and the optimal non-zero vector is selected according to the negative complex power conjugate error prediction value corresponding to the zero vector.Using the negative power of complex conjugate error prediction corresponds to the optimal non-zero vector and zero vector can get the final optimal vector,to achieve fast selection of the optimal vector.This method can reduce the calculation amount of predictive control,shorten the calculation time,saved the computing resources of the chip,and saved the system cost.The DPPC strategy and FVS-PPC strategy of the RMC mentioned in this paper are verified by simulation and experiment.The results show that:Based on the DPPC strategy,in the energy forward mode,RMC has excellent steady-state and dynamic performance,and can effectively suppress load interference.In the energy feedback mode,RMC can still maintain excellent steady-state and dynamic performance,and can achieve automatically switching between the two modes.The FVS-PPC strategy can effectively reduce the average time spent on each selecting the optimal vector of predictive control and the computational burden of the chip,while still maintain the excellent input and output performance of RMC and effectively suppress load interference.