The FPGA Implementation of Wavelet-Neural Network Based On DSP Builder

According to the wavelet-neural network algorithm of relay protection device has the characteristics of requirement of high sampling frequency and large amount of data calculation, we use FPGA to realize the algorithm in hardware. A system model is built and converted into VHDL code through the system-level digital signal processing development tool named DSP Builder from Altera Company. We verified the correctness and feasibility of the method by the timing simulation diagram get from Modelsim SE software. This paper implemented protection method of the wavelet transform and neural network with hardware, and completed a series of research work in the field of transformer differential protection.