Research And Application Aboutsite Selection And Constant Volume Of Distributed Generation Based On Intelligent Algorithm

With the growing demand for electricity and continuous expansion of the powergrid, the disadvantages of traditional large power grid has becameincreasinglyapparent,such as safety, reliability is low, serious environmental pollution, operationdifficulty and so on.So in twentieth Century80times, people put forward the concept ofdistributed generation,because of its low investment,flexible way of power plant,highenergy utilization ratio and no environmental pollution andso on, quickly attracted wideattentionand has been gradually applied. However,on voltage, power flow, short circuitcurrent, reliability, power quality will bring a lot of problems when distributedgeneration connection of the existing network, so the "micro-grid" which integrated ofdistributed power supply and loadingcan provide a good solution for this situation. Butif the DG position and the capacity is not appropriate in the micro-grid, it can not reducethe power loss and operate cost, on the contrary, the power loss will increased and thenode voltage will be affect, furthermore, the quality of power supply systembecame thelow. Therefore, for the realization of economic, environmental protection, power systemreliable, efficiently operation, it is very necessary for research the position and capacityof the distributed generation in micro-grid.This paper first introduces the characteristics of some kinds of distributedgenerations respectivelyand analysed the effects of the power network whichincludedistributed generation. On this basis, establish a single objective optimizationmodel that it consideringthe minimumabout investment cost and network loss basesonthe various constraints. By analyzing the deficiency of standard particle swarm algorithm, combined with differential evolution algorithm and other improved strategy，a hybrid differential evolutionary particle swarm algorithm has been proposed, at thesame time,the validity of the algorithm isverified using some standard test functions.The feasibility and effectiveness of improved algorithm and single optimization modelhas been verified.Then aiming at the environmental problems which more attention in today’s society,considering single objective models have been unable to completely characterize theactual problems, so establish a multi-objective optimization model. On the other hand,the current multi-objective algorithm has problems with premature convergence andlack of diversity, so one improved cooperative coevolutionary genetic algorithm hasbeen proposed and the algorithm with high accuracy approximation of the true Paretofront throughsome experiments. Using this improved algorithm, to realize theoptimization of location and capacity of distributed generation on global and the resultcan provides a variety of solutions for the decision maker.Finally, analyzedthe needs of the micro-grid preliminary construction, developed aprototype software with practical application value using this planning method whichstudied in this paper and the technolgy of mixed programming of VS2010andMATLAB, providing a solutions for decision makers in the micro-grid planning...