| In recent years,China’s social electricity consumption has continued to grow.Connected with the users directly as the end of the power system,the distribution network is not only an important infrastructure related to the national economy and people’s livelihood,but also the focus of power grid construction.Reasonable distribution network planning can ensure reliable power supply performance,so as to meet people’s growing energy demands for a better life.Heuristic intelligent optimization algorithm has the advantages of intuitive,flexible and fast data processing,and is widely used in various fields.When intelligent optimization algorithm is used for distribution network planning,better planning effect can be obtained than the traditional method.The ant colony algorithm has many applications in path planning problems because of its good robustness and reliability in solving such problems.However,there are also some disadvantages,for example,the ant colony algorithm is blind in the early stage of search,and the accumulation of pheromones on certain paths is too much,and the algorithm tends to fall into local optimum leading to a decrease in search efficiency.Physarum polycephalum,a single-celled multinucleated protoplasmic slime mold,was firstly found to solve the shortest path in the maze.Physarum polycephalum model shows its good optimization ability in the real path planning with its research continued.In this paper,the researcher combine ant colony algorithm and Physarum polycephalum algorithm,proposing an ant colony algorithm based on the optimization of Physarum polycephalum model and applying it to the planning of distribution networks.The main work and completion results are as follows:(1)The research background,significance and current research status of distribution network planning were expounded.Several common power flow calculation methods were introduced,and their advantages and disadvantages were also analyzed.Aiming at the special radial structure of the distribution network,the forward push-back method was selected for power flow calculation in the distribution network,and an improved forward push-back method was adopted as the power flow calculation method in this paper.(2)An improved Physarum polycephalum algorithm was proposed,which was combined with ant colony algorithm to form a hybrid algorithm.The basic hybrid algorithm improved the basic Physarum polycephalum algorithm to the Physarum polycephalum algorithm of multiple terminal nodes,which realizes the simultaneous optimization of multiple target points.The improved algorithm could be used to solve the minimum spanning tree and minimum Steiner tree problems.The improved Physarum polycephalum algorithm was used to optimize the basic ant colony algorithm and the max-min ant colony algorithm.Firstly,the prior knowledge of Physarum polycephalum pretreatment was added to the initial pheromone matrix of the ant colony algorithm to guide the ant colony to search purposefully.Secondly,the solution set of the ant colony algorithm was synthesized and optimized again by the Physarum polycephalum algorithm,and the optimized information was introduced into the pheromone update of the ant colony.The hybrid algorithm could make the ant colony have a better convergence speed.What’s more,the hybrid algorithm could ensure that it was not easy for the ant colony to fall into local optimum.(3)The hybrid algorithm of Physarum polycephalum and ant colony was applied to distribution network structure planning to guide the ant colony to optimize purposefully in the initial stage of search.With the gradual evaporation of initial pheromones added by the algorithm iteration,the search scope was expanded and then Physarum polycephalum algorithm was used to continue optimization.Simulation results showed that the hybrid algorithm had fast convergence speed and good optimization ability. |
