Island Partitioning Method Of Distribution Networks Based On Genetic Algorithm

For the sake of economy,safety and environmental considerations,the widespread access of distributed power is a major trend in the development of distribution networks.The access of distributed generations has brought profound changes to the traditional distribution networks and further researches of power quality,relay protection,power recovery still remain to be done.The construction of self-healing function of distribution networks is an important part of smart grid construction.Island partitioning is one of the main technologies of self-healing function construction,and it is also an urgent problem to be solved.In order to solve this issue,based on the concept of weighted connected graph in graph theory,the topological abstraction of distribution network with distributed generations is carried out,and the islanding partitioning model is established with the goal of maximizing the power supply restoration rate.A cloud adaptive genetic algorithm and improved quantum genetic algorithm are raised to solve the model.In the cloud adaptive genetic algorithm,the cloud theory and genetic algorithm are combined to improve the crossover and mutation operator.Replacing the conventional membership curves with a series of points with stable tendency can not only reduce the probability of crossover and mutation with the increase of fitness value,but also keep the uncertainty in the evolutionary process.In the improved quantum genetic algorithm,the concepts of quantum bit,probability amplitude and revolving door are used to design the algorithm.The probability amplitude is used to encode the chromosome,and the probability amplitude is evolved by the revolving door operation towards the currently searched optimal value.Through the improvement of the algorithm,the adverse effect of the observation operation on the individual continuity between the upper and lower generations of the population is avoided,which makes the algorithm more suitable for the multi-tree backpack problem under the constrains of topology.The two improved methods are proved by the adapted PG&E69 node system with multiple distributed generations.The results show the effectiveness of the two algorithms.Through the comparative analysis,it is pointed out that the cloud adaptive genetic algorithm and the improved quantum genetic algorithm represent two different ideas of genetic algorithm:The essence of cloud adaptive genetic algorithm is the information exchange between individuals.The activity and diversity of the population are influenced by controlling the crossover and mutation operator.The crossover and mutation operation have the trait of blindness.In contrast,the improved quantum genetic algorithm has a clear direction.Through the revolving door operation,individuals can rotate to the maximum found so far,in which process more excellent solutions might be found gradually.The selection of the initial population has a great impact on the search results in improved quantum genetic algorithm.Therefore,it is necessary to intervene the population initialization and evolutionary process,and good artificial experience can significantly improve the search results.Both the two algorithms can give full play to tie switches,more feasible solutions may be included into the search range.Compared to the current methods that do not consider the impact of tie switches,the methods proposed in this paper can increase the flexibility of power recovery and obtain better solutions.