Frequency Resolution Research Of Adaptive Detection Algorithm For Inter-harmonics In Power System

With the rapid development of China’s economy and the interconnection of large-scale power grid,the load with nonlinear characteristics is used more and more frequently.In this trend,in addition to linear stable power signals,inter-harmonic signals often appear in power grid.At present,there are two kinds of algorithms for power inter harmonic analysis,which are based on integral operation and differential operation based on frequency estimation.The method based on differential operation has the advantages of short data period,small amount of calculation and good real-time performance.However,multiple parallel frequency estimation algorithms have the problem of frequency resolutionFirstly,for the normalized adaptive comb filter algorithm,a conclusion about its frequency resolution is derived by using the slow integral manifold method,the average method and the Lyapunov stability theorem.In order to satisfy the asymptotic stability of the algorithm,firstly,the frequency adaptive gain must be set to a small enough value,And this value decreases with the minimum value of frequency difference of sinusoidal component;Second,the offset of the estimated frequency relative to the initial value is limited in the region where the minimum difference of the frequencies of each sinusoidal component of the inter-harmonic signal is divided by the number of parallel estimators in the normalized adaptive comb filter.Then,Matlab is used to simulate the two aspects of the above conclusion,and the correctness of the above conclusion is verified.Secondly,the frequency resolution analysis of this algorithm is verified by experiments.The experimental platform includes algorithm implementation based on STM32,signal generation based on TMS230f28335 and FPGA,and waveform display based on LCD.The schematic diagrams of signal generation circuit,power conversion circuit,DA conversion circuit and AD conversion circuit are given.The flow charts of FPGA configuration program,algorithm iteration subroutine and waveform display subroutine are given.The experimental results are consistent with the MATLAB simulation results,which verifies the effectiveness of the conclusion.Finally,the dissertation is summarized and the future research direction of frequency resolution analysis of adaptive comb filter is discussed.