An Improved Particle Swarm Optimization and Its Application to Power System Transfer Capability Calculation

In this paper, an algorithm based on particle swarm optimization (PSO) for power system transfer capability calculation is presented. A dual fitness scheme that takes both objective and constraint into account is adopted to evaluate the survival chance of any particle, thus avoid the drawbacks of traditional penalty method. In the evolution process, if the population best particle has no update during a prescribed number of consecutive generations, it is regarded as a local optimum solution and the searching space around this particle is locked to prevent other particles flying into it. And this particle is saved as one of the candidate solution. In the end, by comparing the fitness of all saved particles and the current population best particle, the optimum value can be obtained. This improved particle swarm algorithm is then successfully applied to IEEE118 bus system optimization problem. Compared with a traditional well-known method, sequential quadratic programming (SQP), our proposal obtains better solutions for this problem.