Based On The Improved Ant Colony Optimization Algorithm Research To Reconstruct The Distribution Network Fault Recovery

With the rapid development of distribution automation technology,people pay more attention to distribution network reliability. Quickly fault location, isolation and restoration of reconstruction has become an important aspect of distribution automation and distribution system optimization. In this paper, a mathematical model is built based on an improved ant colony optimal algorithm and forward and backward substitution method, the main work are as follows.1. Firstly, the basic concept of distribution network fault recovery is introduced, the distribution network is presented with graph theory and matrix, and a new load calculation method is used to calculate the load.The basic principles of distribution network fault location, isolation and restoration is studied and the artificial intelligence combined with matrix expression is used to fault location, and the switch set operating is used to deal with fault isolation and the basic failure recovery process is determined.2. Secondly,the traditional basic ant colony algorithm was improved in this paper, single footwork method is used as the setup model, the original search way and the pheromone release method is modified, which to get the existing improved ant colony optimal reconstruction algorithm.When distribution network failure occurs, the improved ant colony optimal reconstruction algorithm can be used to transform the operational the state of the switch and then using fuzzy logic analysis for load balance.According to the network loss and voltage fluctuation values to find an optimal recovery program which can satisfy the constraint conditions.3. Finally,take the specific distribution network fault as the object, and use the above method in the distribution network fault recovery reconstruction problem, according to the fault data collected, use the improved ant colony optimization algorithm to conduct real-time simulation, and by comparing with the branch exchange method and genetic algorithm,the effectiveness and high convergence and fast real-time advantages of the algorithm can be verified.