| The paper interpreted the coming into being of problems of theresearch on the optimal reactive power and researches the problem ofoptimal reactive power has the great significance in theory and practicalapplication, and simply introduce the development of kinds of theory.It also analyzed the advantages and shortcoming of all kinds of theory.The paper introduced the coming into being, principles and features.The optimal reactive power problem is a optimization problem witha great number of variables and uncertain parameters，its model has twokinds of variables: controlled variables and state variables. Thecontrolled variables include continuous variables and discrete variables,to simplify the calculation, looked the discrete variables as continuousvariables, after calculation, let the variables be the approximatediscrete value.In this paper, the objective function is to minimize the active powerloss. The PQ node voltage of system and reactive power of generators whichare step beyond the boundary are appended to it as punishment function.This paper used the genetic algorithm to calculate the optimal reactivepower, and improved shortcoming of the theory, included: encoded,selection operator, self-adaptively crossover rate and mutation rate,and so on. The theory also compounded some other algorithm, such as thefitness function with simulated annealing, commixing chaos search whichcooperate with genetic algorithm to achieve the best value, when theprocess back to the genetic algorithm, it would change the worsechromosome into the best ones.To prove the genetic algorithm after improved and hybrid is more effective, This paper used the algorithm to programmed with Matlab tocalculate the optimal reactive power of IEEE-14bus system and IEEE-30bus system, and compare the result with other simple algorithm, theresults presented that this hybrid genetic algorithm had a better abilityof global research and the higher convergence rate. |
PDF Full Text Request