Simulated Annealing Based Multi-Object Optimal Planning of Passive Power Filters

Because of the wide application of power electronics equipments, harmonic pollution to power system caused by these nonlinear loads is increasingly severe. With the advantages such as simple structure, lower cost and so on, passive power filters (PPF) is the most wide used harmonic suppression device at present. To plan PPF optimally, optimizing targets including effect of harmonics suppression, original investment and reactive power compensation must be met, but traditional design of PPF is based on experience or consideration of part of economical and technical targets, so the optimal solution can’t be acquired. Simulated annealing algorithm (SAA) is a random optimal solution searching algorithm based on Monte Carlo iteration solution strategy. Began from a initial temperature, random search is progressed among solution space by adopting the Metropolis sampling strategy with probability kick character, the sampling process is repeated with the temperature falling, at last, the global optimal solution is acquired. In this paper, four optimal sub-target functions are constructed, then the general optimal target function is constructed as the linear weighted sum of the four sub-functions and is solved by SAA. The validity of the method is tested by experiment result, which shows that a better general character of PPF in accordance with the optimizing targets can be presented.