Research On Adaptive Boundary Limiting Particle Swarm Optimization And Its Application To Unit Commitment Problem

Swarm intelligence algorithm provides great convenience for solving many engineering optimization problems.Particle Swarm Optimization(PSO),inspired by the bird flock foraging behavior,is a kind of swarm intelligence algorithm widely used in complex optimization problems in engineering,science and management.However,designing new strategies to improve the performance and efficiency of particle swarm optimization is a bottleneck in this field.Boundary limiting strategy has a great relationship with the performance of particle swarm algorithm,and embedding boundary limiting strategy into particle swarm algorithm can improve the performance of the algorithm.However,the current limiting strategy are still relatively single,and the limiting rules are relatively fixed,resulting in poor performance.Velocity boundary limiting is one of the important methods of limiting strategy.This paper studies a particle swarm algorithm with adaptive limiting strategy,and uses adaptive velocity limiting strategy to improve the performance of particle swarm algorithm.The unit commitment problem is an important optimization problem in the power system,but with the development of the scale of the system and the impact of the actual constraints,the traditional optimization methods have been difficult to meet the requirements.As an intelligent algorithm,the particle swarm optimization algorithm has certain advantages in solving the unit commitment problem.In this paper,the improved particle swarm optimization algorithm is applied to solve the unit commitment problem of power system,and related constraint conditions are processed to improve the optimization effect.The main contributions for this paper are listed below:(1)By analyzing the relationship of the evolutionary state evaluation to iterations and the dimension of problem for particle swarm optimization,a formula is designed to calculate the evolutionary state evaluation(ESE)which is influenced by the iterations and problem dimension,and calculates the velocity limit on the basis of the ESE,so a Particle swarm optimization with velocity limit combining iteration and problem dimension is obtained.Finally,the algorithm is affected by iteration and problem dimensions,which is adaptive.(2)A self-adaptive boundary limiting binary particle swarm optimization algorithm is proposed.Aiming at the uniqueness of binary particle swarm algorithm position in binary form,a formula for calculating the average distance of the population is proposed,and its population evolution state evaluation value formula is constructed.In view of the problems existing in the current binary particle swarm optimization algorithm transformation function,a U-shape function is adopted.Finally,the improved algorithm is applied to solve the feature selection problem,and the effectiveness of the algorithm is verified.(3)The above two improved particle swarm optimization algorithm is applied to solve the unit commitment optimization problem at the same time,and the constraint conditions are processed according to the characteristics of the unit commitment problem,the problem is solved using the phased optimization idea.Through the unit test case experiment,the solution results are compared and analyzed with those of other intelligent algorithms and traditional algorithms to verify the good performance of the improved algorithm.