Optimal Parameters Design For Passive Power Filter Based On Chaos Algorithm

There are many ways to control power harmonic pollution. From Chinese practical conditions and economical considerations, this thesis chose passive power filter. Because of its simple structure, low cost and easy maintenance, passive power filter still works as the most common filtering method in practical engineering application today.First, the causes, danger and elimination approaches for power harmonic distortion were discussed in this thesis. It gave details on various types of passive power filter, their filtering principles and their characteristics. Furthermore, physical models of the passive power filters were worked out.Secondly, commonly used parameter-optimization design methods for passive power filter were elaborated respectively. Their merits and disadvantages were also generalized.Thirdly, the thesis introduced the concept of chaos optimization. The development, definition and characteristics of chaos were also presented. The ergodicity nature of Logistic map in chaos state was studied through mathematical deduction and the current use of chaos algorithm in optimization application was summarized.Then, by synthesizing the initial investment, the compensation capacity and the harmonic distortion into general consideration, a multi-objective parameter-optimization design method based on chaos optimization algorithm was proposed. It utilized the ergodicity of chaos to optimally design the parameters for passive power filter. By simulation research and comparison with the traditional design methods, it can be concluded that this optimization algorithm has better overall performances. Finally, to solve the long searching period causing by the only use of pure chaos optimization, the thesis proposed a new chaos optimization algorithm combined with annealing strategy to design the parameters for passive power filter. The design process for the new improved algorithm was analyzed and generalized. Once again, by computer software simulation, the better feasibility and high efficiency of this parameter-optimization design method was proved.