Optimal Design Of Passive Power Filters Based On Genetic Algorithm

The problem of dealing with harmonics is regarded generally for the increasingly serious harmonics in power system. Passive power filters is adopt widely because it is simple configuration, convenient maintenance, low cost and etc. Otherwise, passive power filters is easy affect by change of system parameters, can' t remove sequential harmonics, and easy arose series and parallel resonance with system. So optimization of parameters of PPF is key of its design. Conventional approaches for PPF design only follow the experiences and simple criterions. These approaches have disadvantages that have too many assumptive conditions, low optimization ability, and don' t consider some questions such as reactive power compensation.On the base of making analysis of existing method, taking the advantage of parallel multi-objective optimization in Genetic Algorithm, this paper proposes a new strategy to optimal filter performance. Besides of basic resonance condition, each adaptable function is given of investment, the capacity of reactive power compensation and the harmonic distortion of PPF. A array composed by the value of capacitor is the origin chromosome in calculation. A chaos operator and Logistic Function are introduced to overcome the premature convergence in calculation process. In the exploitation of relative software, high-speed calculation abilityof computer and engineers experience are combined. Considering that the value of capacitor isn' t seriate, the software gives capacitors result with existing type in makert.This paper analyzes the process of new optimal design in detail and concludes a series of important mathematics formula. What' s more, it expatiate some means to key problems in software design at large. Finally, validity and feasibility of the new algorithm are proved by simulation and experiment results.