Reactive Power Optimization Of Power System Based On Modified Particle Swarm Algorithm

It has been a tricky problem for power system to ensure the optimal configurationof the reactive power. Since the quality of the grid voltage is determined by thereactive power, only when the voltage stability is in a normal state could it be able tomaintain the normal operation of the power system, otherwise, it may affect thestability and security of the system. There are many resolutions to the optimalconfiguration of the reactive power, and the particle swarm optimization algorithm isan artificial intelligence algorithm which is innovative and burgeoning. Because it iseasy to realize and relatively easy to find the global optimal solution as well as its fastconvergence and good robustness, the particle swarm optimization algorithm has beensuccessfully applied in many fields though its short time development. This article isbased on the Particle Swarm Optimization, and a new algorithm is proposed byimproving the particle swarm algorithm. Moreover, it simulates the application in thenode system, and analyzes the simulation results. The contents of the article are asfollowing.Firstly, this article introduces the research background and significance of reactivepower optimization as well as current research status quo; meanwhile it analyzes thevarious optimization algorithms which are used in reactive power optimization.Secondly, this article illustrates the concept and significance of the reactive powerand voltage in the power system, and introduces and analyzes the mathematical modeof the reactive power optimization.Thirdly, this article elaborates the research background, objectives and basicprinciples of particle swarm optimization, and introduces algorithm parameters andseveral common Improved Particle Swarms. Moreover, it proposes the improvedparticle swarm algorithm aims at the possible precocious phenomena and slow rate ofalgorithm convergence. In order to control the diversity of particle, it adds a thresholdD into the algorithm, thus it could avoid the premature extreme of the algorithm. Andmore, it improves the velocity equation of particle swarm optimization by increasingthe contact between a particle and the individual optimal solution of other particles, tosome extent; it improves the convergence speed of the algorithm in the later stage.Finally, in order to verify the proposed algorithm has a stronger searchingcapability than the basic particle swarm algorithm, it simulates the two algorithms inIEEE-6and IEEE-30node system, and analyzes the simulation results, it ultimatelyproves that the improved particle swarm optimization is better than the basic particleswarm algorithm in the respect of solving the problems of premature phenomenonand convergence speed in the later stage.