Optimal Allocation Algorithm Of Smart Grid Distribution Network Protection

In recent years, with the continuous development of social economy there is a growing demand for electricity. In this context, the development of the distribution network has become more complex. For the contemporary network protection, the conventional protection methods have been unable to play its due role. This is mainly reflected in the following aspects:first, when changes occur in the power system and if the relay protection mode is off-line confirmed, then its setting values cannot be at their best; Second, at the minimum operating mode, once the power system have serious failure, the performance of relay protection systems will be seriously impaired; in serious cases, such failures may also lead to tripping phenomenon. To solve the above problems, a "adaptive protection" concept has been raised and this concept has been widely used in power protection.At first this paper outlines the origin and Development conditions of the "adaptive protection"; Then discusses the startup mode, implementation and structural form of the online subscription system; and finally briefly depicts the deficiencies of distribution network protection, the use of adaptive protection and the current study situation of this research area in China and other countries.Secondly, this paper presents a systematical exposition of the theory of genetic algorithms. Improvements relating to the inadequacies of the algorithm are proposed, such as a lack of local search and algorithm premature convergence, which means the problem constraints cannot be reflected and coding standard is not strong, and so on. Further improved are the genetic crossover, mutation operator and selection operator of genetic algorithm. And the following improvement measures are proposed:first, such two alternatives are used as a two-way competition and optimum retention policies. Second, for an increase in the rate of genetic cross, the way of best individual reservations algebra hyperbolic is applied. Third, For increase in mutation rate, methods of the best individual preservation algebraic exponential function are used.Finally, based on genetic optimization algorithm and integrated with the actual situation of distribution network protection, a distribution network protection model is built based on genetic algorithms. In this model, the primary optimization objective function is the overall time of minimum protection actions; This model synchronously determines the switching and operation mode of algorithm protection and a detailed analysis of the genetic algorithm constraints are given. Eventually, based on Matlab simulation, mathematical modules are built to carry on the simulation experiment combined with mathematical models and algorithms proposed in this paper, which show that the genetic algorithm in the distribution network protection is effective and feasible.