Active Distribution Network Planning Based On Invasive Weed Optimization Algorithm

With the increasing penetration of renewable distributed generation(DG)and electric vehicle(EV),traditional distribution network is under severe challenges.Active distribution network(ADN)has been developed rapidly,which can solve DG and EV access problems with active control measures.ADN is a complex system with high-order,uncertainty,nonlinear and other characteristics,and ADN planning is a multi-constrained nonlinear combinatorial optimization problem.It is urgent to establish a comprehensive planning model and design an efficient optimization algorithm.Battery swap station(BSS)can achieve fast EV battery swapping utilizing the decoupling between EV and batteries.Besides,BSS can be used as a kind of flexible load(FL)resource to participate in ADN dispatching plan.This paper considers BSS in ADN planning model with multi-objective including feeder cost,operation and maintenance cost and carbon tax.To verify the superiority of BSS,ADN planning results between EV battery swapping mode and EV charging mode are compared.This paper also proposes improved binary invasive weed optimization(IBIWO)algorithm to optimize ADN planning.IBIWO introduces function sig to discrete continuous space,so that it can solve combinatorial ADN planning.It sets a diversity threshold to limit Hamming distance and ensure population difference.Besides,IBIWO designs evolutionary parameters targeted according to the characteristics of penalty function.Finally,carry out planning simulation on 25-node system.The comparisons of planning results between four planning schemes prove the economic and environmental benefits brought by battery swapping mode.The comparisons between IBIWO and particle swarm optimization algorithm(PSO)prove IBIWO behaves better than PSO in global and stable search.