Reactive Power Planning In Distribution Network With DG Based On Hybrid Intelligent Algorithm By AFSVM-PSO

As a new and potentially energy comprehensive utilization mode, distributed generation(DG) has a diverse energy types, flexible operation, good environment benefits, obviousinvestment effect and other advantages. As the representatives of intermittent renewable energy, photovoltaic plant and wind turbines developed rapidly in recent years. But due to therole of climatic and environmental factors, the randomness of DG power output will make the power flow volatile. This may lead to the node voltage level is not qualified, so the reactive power planning in distribution network with DG is crucial. This paper focuses on the above questions to research as the following work.Firstly, this paper introduces the distribution operation challenges after DG access todistribution network and summarizes the domestic and foreign information in the present research of reactive power planning in distribution network with DG. Then the probability models of photovoltaic plant and wind turbines are established. Taking into account the power output of DG and the nodes load fluctuation both have some correlativity, the random scenes are generated by correlational Latin hypercube sampling in this paper.Secondly, the mathematical model of reactive power planning in distribution network with DG based on chance constrained programming is established. The objective function takes into account the operation economy of distribution network, power quality, and staticvoltage stability margin according to the weight coefficient. And the lower and upper bound constraints of nodes voltage are chance constraints by the given confidence level.Thirdly, the optimal algorithm for the reactive power planning in distribution network with DG is explored. Based on support vector machine(SVM), artificial fish swarm algorithm(AFSA) and particle swarm optimization(PSO), this paper presents the AFSVM-PSO hybrid intelligent algorithm and gives the detail algorithm steps and flowcharts for solving the mathematical model of reactive power planning. The AFSVM-PSO hybrid intelligent algorithm can achieve coordination equilibrium between calculation efficiency and accuracy in the whole process of optimization.Finally, the paper takes the IEEE33 nodes distribution network and an actual 64 nodes distribution network in Guangdong as examples, uses the AFSVM-PSO hybrid intelligent algorithm to solve the reactive power planning for these two distribution network. And the optimization effects are analyzed. The results of optimization for these two cases show that the reactive power planning scheme can reduce the operation cost of distribution system, and improve the nodes voltage qualities and static voltage stability margin. The simulation results verify that the proposed model and method in this paper are reasonable and effective.