Research On Power Distribution Network Structure Optimization Of City

Distribution network optimization is not only an important part of powerdistribution network planning, but also the solid basis of ensuring the stability of gridoperation and power grid safety and reliability. Therefore, there is an importantguiding significance to optimize the distribution network scientifically andreasonably.Firstly, the distribution network optimization problem is described in detail, andthe model based on the minimum cost is selected as the optimization objectivefunction. Secondly, the ant colony optimization algorithm and genetic optimizationalgorithm improvements are both made and applied in the distribution networkoptimization work to verify the feasibility and effectiveness of the algorithms. Finally,a hybrid model including both of the algorithms is established to optimize thedistribution network work. A specific example is employed to demonstrate theEffectiveness of this model.The main contents and results are as follows:An in-depth study is done for the urban distribution network optimizationproblem. According to the features of distribution network planning, the investmentcosts, maintenance costs and depreciation expense line losses are considered as threemain aspects to establish the power grid optimization objective function, and set theappropriate constraints conditions.The ant colony algorithm is briefly described and improved based on itscharacteristics. This paper not only achieve the linkage among the parameters, and the maximum number of iterations in the improvements of transition probabilities,but also propose an adaptive volatile factor in the improvements of the pheromoneevaporation factor. Then the optimized procedures of distribution network aredetermined and applied to optimize the distribution network in Hohhot to verify thefeasibility and effectiveness of this method.As the aspect of improvements in genetic algorithms, this paper make a basicintroduction about the genetic algorithm and make an improvement among encoding,crossover probability and mutation probability according to the characteristics ofgenetic algorithm. Then the improved genetic algorithm is applied in the distributionnetwork optimization in Hohhot to verify the effectiveness and feasibility of themethod, and describe the guiding significance and practical value in distributionnetwork.A series of questions are caused easily in solving optimization problems byusing a single algorithm because of its own defects, such as genetic algorithm appears redundant iteration and ant colony algorithm appears initial solution scarcity. Tomake the distribution network optimization more accurately, this paper mix theimproved ant colony algorithm and the improved genetic algorithm to absorb theadvantages and overcome the disadvantages from each other. Finally this hybridmodel is applied in the distribution network optimization problem in Hohhot tovalidate the superiority and Practicality in distribution network.The hybrid algorithm, the improved ant colony algorithm and improved geneticalgorithms are respectively applied in optimizing the distribution network. Thecomparison conclusion shows that the hybrid algorithm has a better effect and higheraccuracy in optimization performance compared with a single algorithm. Thereforethis method provide a new kind of idea domain in the urban distribution networkoptimization in the future.