| In power engineering,there are mainly active power and reactive power.The compensation amount and distribution of reactive power have a direct impact on the safe and economic operation of the power system.Therefore,the study of reactive power optimization is one of the important ways to adjust the running state of the power system.The reactive power optimization problem is a well-known nonlinear and multi-constraint optimization problem,which can not be effectively solved by current optimization algorithms.The traditional optimization methods have strict requirements for the accuracy of the model and have poor adaptability to different power models,thus resulting in large errors in the optimization results.However,Differential Evolution(DE)algorithm is widely used in reactive power optimization because it is not strictly limited to the model and has good adaptability to the power model.Existing DE algorithms are easy to fall into local optimum and often fail to find the global optimal solution when solving reactive power optimization problems.Therefore,in this paper,the Depth Information-Based Differential Evolution(DiDE)algorithm is used to solve the reactive power optimization problem.The specific task of this paper is mainly divided into the following four points:(1)As for the reactive power optimization,the advantages and disadvantages of GaussSeidel(G-S)method and Newton-Raphson(N-R)method in load flow calculation were studied and analyzed,and in this paper,the N-R method was chosen as the load flow calculation algorithm.Then the mathematical model of reactive power optimization was analyzed,and the minimum active power loss was selected as the optimization objective,and the objective function,constraint condition and penalty function were established.(2)As for the optimization algorithm,the DE algorithm was deeply studied.At present,the algorithms used in the research of reactive power optimization have an insufficient global searching ability and are easy to fall into local optimum.Therefore,the depth information in the population evolution process was incorporated into the mutation strategy to enhance the“exploration” and “exploitation” ability of the algorithm.The information could reflect the structural characteristics of the objective function,and then the parameter adaptive mechanism was set.Based on these two points,a Di-DE algorithm was proposed.(3)Numerical optimization verification of the Di-DE algorithm was carried out on international standard test suite CEC2013,and then the experiments were performed on Matlab2016 b.In the 30-dimension optimization,the convergence accuracy and convergence speed of the DE algorithm,iw PSO algorithm,QUATRE algorithm and Di-DE algorithm were compared,and the experimental results showed that the Di-DE algorithm was the best in these two aspects.(4)The Di-DE algorithm was verified for reactive power optimization on IEEE-14 bus,IEEE-30 bus and IEEE-57 bus systems.The simulation experiment was carried out on Matlab2016 b and compared with the DE algorithm,iw PSO algorithm and QUATRE algorithm.The experimental results indicated that the Di-DE algorithm had a better convergence accuracy and convergence speed,and could effectively reduce the active power loss of the system. |
