Optimization Design Of Passive Filter Based On Global Optimization Algorithm

With the extensive use of power electronic devices and non-linear load, electricity gridharmonics and reactive power cause increasingly serious pollution, pose a potential threat tosecurity, stability, economic operation of power system, have great impact on the electricalenvironment around. Harmonic is considered to be one of the major power grid pollution.Governing harmonic issue has caused scholars’wide concern.Passive power filter with parallel operation of power grids, which is one of the currentapplying techniques of harmonic suppression, has been widely used due to its low cost, goodfiltering effect and many other advantages. When designing passive power filter installed ondistribution network for harmonic suppression and fundamental reactive compensationpurposes, its initial investment, reactive power compensation capacity, filtering harmoniccontent are the three principle objectives.Early passive filter design methods are generally based on work experience or a simpletechnical economic indicators to select parameters, which require repeated trial and error, arenot only inefficient, of limited precision, their combined effect is often not optimal. With thedevelopment and application of computer application technology and its optimizationalgorithm, mathematical modeling by using the target values and constraints of filterparameter, by conversing filter design problem to filter parameter optimization problem, bymeans of application software simulation, optimization algorithm for optimization design ofthe parameters, we can further improve the specification of passive filter. However,Gradient-based optimization which is the current conventional methods is often difficult toachieve the desired effect, very vulnerable to local optimization, of tremendous blindness andof weak optimization capability. To solve the problem of multi-objective optimal design ofpassive power filter, the algorithm must be improved.This article plans to solve filter LC parameter optimization problems by usingcatastrophic genetic algorithm, which not only can find global optimal solutions, but alsogreatly improve the speed of algorithm optimization. Finally by calculating specific projectexamples, comparing to the method of traditional engineering design, it demonstrates the superiority of genetic optimization algorithm filter design, and validity of improving geneticalgorithm.