Research On Reactive Power Optimization Of Distribution Network With Distributed Generation

With the continuous consumption of fossil energy and the deterioration of the environment,the contradiction between resources and environment is becoming increasingly serious.Distributed generation is widely used in distribution network as a new energy generation mode.However,with the continuous improvement of the permeability of distributed generation,it has a significant impact on the voltage distribution and network loss of distribution network.Therefore,the research on reactive power optimization of distribution network with distributed generation has certain theoretical significance and application value for improving the voltage quality of distribution network and reducing network loss.Firstly,DG and its interface types are classified,and different DG interface types are modeled.Then,the power flow calculation of DG-containing distribution network is carried out by forward and backward substitution method,and the influence of DG on distribution network voltage distribution and network loss is preliminarily analyzed from three angles of access location,capacity and power factor.In order to solve the problem of reactive power optimal allocation in distribution networks with distributed generators,an improved loss sensitivity analysis method is proposed,which combines the proportion of each node load to the total loss and the power factor angle with the traditional loss sensitivity to construct a comprehensive loss sensitivity,thus realizing the rational allocation of distributed generators and capacitors.To improve the efficiency of reactive power optimization model and algorithm for distribution networks with distributed generations,an improved backbone particle swarm optimization(BBPSO)algorithm is presented,which improves the initial quality of BBPSO by utilizing the super-uniformity of Hammersley random sequence of quasi-Monte Carlo method.By introducing cloud model,the population is divided into three sub-groups to realize the stochastic weight in the population renewal formula.Adaptive adjustment improves the population updating strategy,so as to improve the convergence of the algorithm.Finally,in the simulation of standard IEEE33 nodes,the reasonable allocation of distributed power supply and capacitor group position is realized by using comprehensive sensitivity;the improved BBPSO and standard BBPSO are combined with reactive power optimization model with penalty function of minimum active power loss,voltage deviation and reactive power output deviation of distributed power supply respectively.Through simulation and comparative analysis,the improved algorithm in this thesis has better convergence efficiency than standard BBPSO.