The Study On Solving Unit Commitment Problem Based On Improved Binary Particle Swarm Optimization

The unit combination(UC)problem belongs to the economic dispatch problem of the power system.Its goal is to reasonably arrange the start and stop status of the unit and the load distribution of the unit in a dispatch cycle,so that the operating cost of the entire unit system reaches the minimum under the constraints In recent years,many intelligent heuristic algorithms have been applied to solve the UC problem.Binary particle swarm optimization(BPSO)is an intelligent optimization algorithm that is very suitable for solving UC problems.This paper studies the improvement of BPSO algorithm and applies it to the solution of UC problems.This paper studies the improvement of BPSO algorithm and applies it to solving UC problem.Binary Particle Swarm Optimization(BPSO)is developed by extending the continuous particle swarm optimization(PSO),which is mainly used to solve discrete combinatorial optimization problems.But BPSO parameter settings and transfer function are not reasonable.By analyzing the variation of the distance between each particle and the globally optimal particle,this paper concludes that the original BPSO algorithm has the defect of too strong global search ability,improve the BPSO algorithm by changing its transfer function and improving its own particle diversity.The improved BPSO algorithm is used to solve the UC problem.Through special treatment of the constraints of the UC problem,different methods for solving the UC problem are designed,and achieved good results.The main work of this paper is as follows:1.Define the distance formula between the BPSO algorithm population particles and the global optimal particle,used to analyze the particle population changes with iteration,and find that the original BPSO algorithm S-shape transfer function will affect the algorithm converge,and concluded that the global search ability in the later BPSO is too strong to getting the optimal solution.2.Aiming at the problem that the original BPSO algorithm's S-shaped transfer function is too strong in global search ability,a new type of V-shaped transfer function is proposed.The particle speed transfer function and position update formula are modified and applied to the feature selection problem.The new V-shaped transfer function can move the particles to the direction of the global optimal particles,and improve the search ability of the algorithm.3.From the perspective of changing the particle diversity of the BPSO algorithm,an adaptive mutation operation is proposed for the BPSO algorithm,which causes each particle to dynamically change with a probability of changing from large tosmall after updating the position.Meanwhile,a linear increasing strategy is adopted and the weight setting interval is experimentally determined,and the experimental verification improves the performance of the algorithm.4.The BPSO algorithm is applied to the UC problem,and a solution method for the UC constraint condition is proposed: two strategies of non-segmentation and segmentation are used to solve the scheduling cycle problem,and then the improved BPSO algorithm combined with the segmentation idea is used to solve the UC problem.Through experimental analysis,it is concluded that the solution proposed in this paper can get lower consumption cost and fewer startup numbers,and the algorithm has higher stability.