Research On The Control Algorithm Of Energy Storage Inverter Based On FPGA

Micro grid is a hot research topic in the field of smart grid in the world. It is generally provided by the energy storage inverter under the condition of off grid operation mode. Because of the smaller capacity, the waveform quality of the system voltage will be affect when the nonlinear load is connected or the load changes, and even lead to instability of the system. With the continuous improvement of the quality requirements of the user, the problem is to be solved urgently.Inverse system method is an important breakthrough in the control theory of nonlinear system. However, due to the complexity of the controlled object in the practical engineering, it is often difficult or impossible to obtain the accurate modeling, and the solution of the inverse system is very difficult. Therefore, the application of this method is limited. Artifical Neural Network(ANN) is a typical intelligent method, which has excellent approximation ability for complex nonlinear systems. Therefore, we can constitute a novel control method with the inverse system method and the ANN.In order to improve the energy storage inverter output voltage waveform quality of micro grid. In this paper, an inverse control scheme based on ANN is proposed. Because of the Large computation of ANN and the control period of energy storage inverter is usually only a few dozen us, the conventional single chip microcomputer controller is difficult to complete the operation of neural network in one control cycle. Compared with the traditional single chip microcomputer and PC machine, FPGA has the advantages of fast computation speed, high precision and parallel computing.Therefore, this paper will use FPGA to design a ANN special operation device,and in order to realize the inverter voltage inverse control scheme a composite controller is designed by FPGA and single chip microcomputer. In the design of the neural network, the implementation scheme of the excitation function and the floating point number is the focus. Experimental results show that the proposed control scheme is feasible and effective.