Improved Differential Evolutionary Algorithm Based On DG Distribution Network With Reactive Power Optimization

As an important part of power system analysis,the problem of reactive power optimization has been studied by many experts and scholars for many years and has achieved fruitful results.Today,in the distributed generation grid capacity of the increasing environment,but also brought new challenges,after the grid connection caused by voltage fluctuations,harmonic pollution,network loss increase,cost increase and a series of other problems,for reactive power optimization put forward a new problem,of which photovoltaic,wind power output uncertainty caused by voltage fluctuation problems are particularly prominent,so the Distributed Generation distribution network reactive power optimization is particularly important.Firstly,the significance and background of the study are described,the distribution network,the reactive power optimisation model,the algorithm and the current status of reactive power optimisation with distributed power sources are presented.The impact of distributed power supply access on the voltage magnitude of each node in the distribution network and the total active network loss of the system is analysed in terms of the location of the grid-connected nodes and the size of the grid-connected capacity,and further corroborates it in combination with simulation experiments.Secondly,the DG output model is analyzed,and the mathematical model of DG output and load is established,and the probability flow is processed by the three-point estimation method,and the probability density function is obtained by processing the correlation and Gram-Charlier series expansion by the Nataf transform.It was verified by simulation compared with 20,000 samples of Monte Carlo and the original three-point estimation method.Then,in order to solve the shortcomings of the traditional differential evolutionary algorithm,which tends to converge too early and leads to poor local search ability,this paper improves it by means of an elite strategy to avoid falling into a local optimum,adaptive control is used to adjust the parameters to improve the search ability of the algorithm,and in order to verify the effectiveness of the improved algorithm,the simulation method is compared with the traditional algorithmFinally,the pareto optimal solution theory is combined to construct a dual-objective optimization model with minimum active network loss and minimum voltage deviation,and the improved differential evolution algorithm based on the pareto optimal solution is tested to see the effect of reactive power optimization.The pareto-optimal solutions are obtained through the simulation verification of the IEEE33 standard node system,with ten sets of solutions.The corresponding ten reactive power optimisation strategies are also proposed on this basis,and the objective of ensuring proper application in different scenarios is achieved with the validation of three typical application scenarios.