Reactive Power Optimization For Ship Power System Based On Adaptive Particle Swarm Optimization

With the rapid development of the world economy, the demand of the production andlife, humans also set a higher request to the security, stability, economy of the ship powersystem. Ship power system optimization is an effective way for reasonable dispatch andconfiguration reactive power. Not only reactive power optimization can reduce networkpower loss of the ship power system and operating costs, but improve the power quality ofship power system significantly. Research for reactive ship power system optimization hasgreat significance in theory and application.After analyzing and discussing the characteristics, network structure and workrequirements of the ship power system, we construct a mathematical model applied to shipreactive power optimization refer to terrestrial power system. The objective function can becreated based on different requirements which can be economic indicators, voltage stabilityindex and power quality indicators. Single objective and multi-objective reactive poweroptimization are used to solver ship power system. After analyzing the characteristics of thenetwork structure of the ship power system, different types of flow calculation methods areestablished for different types of ship power system. Simulation model in the paper isannular ship power system which is widely used now, so node potential method is beingutilized for flow calculation.Reactive power optimization is a very complex multi-constrained and multi-variablelinear programming problem including multiple variables and multiple constraints. It is verydifficult to solve this problem with conventional mathematical methods. A suitable artificialintelligence algorithm must be searched to solve it.For multi-objective, multi-constraint, multi-variable and nonlinear characteristics ofreactive power optimization, in this paper, particle swarm optimization as a starting point,we introduce the basic principles and implementation process of particle swarm algorithmdetailed, summarize and discuss its improved form. A series of improvement measures havebeen created for the defect of standard particle swarm optimization. Based on the localversion of particle swarm optimization, increases the mutation learning of genetic algorithm,adaptive changes in size, the field of population size, particle acceleration factor to ensurepopulation diversity of the update process. Thus ensures the speed and accuracy of PSO.Finally, adaptive particle swarm optimization (APSO) is applied to solve reactive power optimization of the annular ship power system; simulation is being conducted inMATLAB, by comparing the simulation results, verifies the superior performance of thenew algorithm and the correctness of reactive power optimization model. Active networkloss minimization as single objective is used first, secondly, multi-objective is transformedinto single-target by taking the fuzzy theory, and reactive power optimization is solved withAPSO again. After the horizontal, vertical comparison of the results, verifies the superiorityof APSO. After the multi-objective reactive optimization, overall performance has beenmore superior, ensuring economic, stability and security of the ship system.