Research On Power Harmonic Detection Algorithm And FPGA Implementation

Power system harmonic analysis has been studied for many years.With the large investment of non-linear loads and the rapid development of new energy technologies,voltage and current waveforms are generally distorted,and power quality problems such as harmonics and interharmonics are becoming more and more serious,and various types of faults and accidents caused by harmonics and interharmonics often occur,which directly affecting the safe and stable operation of the power system.Accurate and fast mastery of the harmonic content of the power system(the amplitude and phase of the fundamental frequency and harmonic frequency components)is the starting point for studying harmonic problems and the main basis for harmonic control.In order to improve the power quality,maintain the safe and stable operation of the power system,high-precision detection and analysis of harmonic and interharmonic components should be carried out to accurately grasp the actual situation of harmonic in the power system.This paper analyzes the typical power harmonic detection method,the principle of hardware acceleration of FPGA and the ARM+FPGA architecture of Xilinx All Programmable Zynq-7000 SoC device,the improved and expanded Prony harmonic analysis method and FFT harmonic analysis method are mainly studied.The windowed interpolation FFT algorithm can be implemented easily by embedded systems,so it is widely used in power system harmonic detection,it can restrain the spectrum leakage and fence effect caused by non-synchronous sampling and non-integer cycle truncation in order to improve analysis accuracy of FFT.Aiming at the shortcomings of common windowed interpolation algorithms,the spectral characteristics of 5-term MSD window are analyzed and a harmonic analysis algorithm based on 5-term MSD window and six-spectrum-line interpolation FFT is proposed.The algorithm uses six spectrum lines close to the peak spectrum line frequency point to perform weighted operations,the information contained in symmetric spectrum lines is consider.Through data fitting,the practical interpolation correction formula corresponding to the window function is obtained,which simplifies the operation process.MATLAB simulation results show that the algorithm can be designed and realized easily,it can restrain the spectrum leakage effectively and it also has strong anti-interference ability.Compared with other common four-spectrum-line interpolation FFT algorithms,the algorithm has higher detection accuracy.The technical indicators of the harmonic detection system and hardware overall design scheme are formulated.Considering the Zynq-7000 AP SoC chip's advantage of tightly combining the ARM processor with the Xilinx FPGA,it is determined that the proposed algorithm is transplanted to the Zynq-7000 chip for implementation.Choose an ideal development board Zedboard based on the Xilinx Zynq-7000 extended processing platform as a development platform.According to the basic process of Zynq design,Vivado is used to build hardware logic system,it mainly includes creating and calling windowed(multiplier)IP core and error correction IP core,calling and configuring Xilinx official XADC,BRAM,FFT,CORDIC,Floating-Point,FIR,DMA and other IP cores.And simulation of each module and the comprehensive implementation of the logic system.Using the SDK(Software Development Kit)as an ARM development tool,the ARM on the PS part is programmed to drive the transfer,and the data is stored and displayed after the processing of the PL part.At the end of the paper,the harmonic detection results(amplitude,frequency and initial phase)of the designed system are compared with the simulation results in MATLAB platform.The results show that the calculation accuracy of the system can fully meet the technical specifications set at the beginning of the system design.