Research On Harmonic Detection Method Based On Improved Shuffled Frog Leaping Algorithm

With the development of power electronic technology, as well as the applicationof a large number of nonlinear loads in power system, there exists more and moreharmonics and inter-harmonics in power grid. Harmonic pollution affects the safeoperation of power system, and makes the power quality to drop, therefore the researchon harmonic problems has important significance. And harmonic detection as thestarting point of harmonic analysis, high efficient and accurate harmonic detectionmethod has a certain practical significance.The article took harmonic of power system as the research object, on the basis ofintroducing the causes and the damage of harmonic in detail, compared the research ofthe harmonic detection algorithm at home and abroad, and put forward a researchthought of harmonic detection method based on Shuffled Frog Leaping Algorithm(SFLA).Based on chaotic mapping, introducing chaotic operators into SFLA global searchand using the global optimal frog guides the evolution, It proposed a chaotic ShuffledFrog Leaping Algorithm (CSFLA), and fused Least Square (LS), a harmonic detectionfusion Algorithm based on CSFLA-LS has been proposed. It discussed the influenceof the sampling frequency, the data window length, and the dc component, thesimulation under noise showed the feasibility and effectiveness of the fusion algorithm.Analysis of the latest distribution estimation method. Based on the thought ofGaussian distribution estimation, It introduced the concept of Gaussian modeling,established modeling of excellent Frog from on macroscopic in mix process of SFLA,then put forward a new harmonic detection algorithm based on Gaussian Shuffled FrogLeaping Algorithm(GSFLA). Compared with PSO algorithm, experimental simulationdata showed that average estimation accuracy of harmonics amplitude was enhancedby5.3%, and the average estimation accuracy of harmonics phase was enhanced by4.7°.The research shows that the proposed algorithm (GSFLA) has faster convergencespeed and higher estimation precision in power system harmonics estimation.Finally, this paper summarizes the research contents and make prospect forfurther research aimed at the limitation of the paper research. Intelligent detection of harmonic will be the research trends of harmonic detection method, and will get rapiddevelopment.