Research On Internal Model Control Of Micro Grid Energy Storage Inverter Based On Support Vector Machine

With the lack of the non renewable energy which main body is coal and oil, the energy revolution for low carbon energy has already started. The micro-grid has become an essential way to solve the seamless connection between the distributed renewable energy and the large power grid, and the energy storage inverter plays an important role in it. The inverter performance determines the effect of the inverter, exploring the advanced control strategy can effectively improve the performance of the inverter. In this paper, the control strategy of the micro-grid energy storage inverter is studied, and the method of support vector machine is used to improve the performance of the inverter.First, the structure characteristics of the micro-grid energy storage inverter is explored. Through the topology analysis of the typical single phase full bridge inverter circuit, the frequency characteristics and reversibility of the system is studied. Based on the research above, the simulation model is built on the Simulink platform.Then, the support vector machine theory and the regression principle is introduced,and according to the characteristic that the selection of the kernel function parameters has an important influence on the regression results, using the glowworm swarm optimization algorithm to achieve the optimization. Based on the analysis of the structure and the characteristics of the control system, the internal model control method based on support vector machine is studied, and the construction process of the control system is described.In the end, the internal model and inverse controller model of the energy storage inverter is built by the support vector machine method, and the model is optimized by using the glowworm swarm optimization algorithm, the internal model control system for the energy storage inverter based on support vector machine is constituted. Aiming at the problems in the simulation results, the improved scheme is proposed, the PI controller and the repetitive controller is introduced, and the improved internal model control system is built, after simulation in a variety of operating conditions, comparing the control effect with the neural network compound inverse control and the PI control. The results show that the improved control strategy is effective and it can achieve the improvement of the inverter efficiency.