Optimization Research Of Power Multistage EPON Based On Adaptive Genetic Algorithm

Recently, with the development of national economy, the scale of power distribution network has gotten bigger and bigger. And each of the power company has gradually set up the core network that gives priority to optical fiber communications. The network has covered all the branch power plants and it already has formed a powerful electric power dispatching system which can fulfill from power generation to transmission and to substation integration. Because of the weakness in power distribution network communication, the construction of power distribution automation system has been under severe constraints. In the development of power distribution network communication system, it has been going through the stages like the power line carrier, wireless communication and the developing of optical fiber communication. But a variety of communication methods combined application still exist in power Grid Company’s distribution network automation communication system. Optical fiber communication, whether from the cost, capacity, or the power grid security point of view, will be a best communication mode of power distribution automation system in the future. Due to the situation that the power distribution of branch circuit is more complicated than the main, and in order to realize the distribution Network automation which must collect related information in multiple points, the backbone of the application of traditional point-to-point communication mode has been unable to meet the demand of the construction of the power distribution automation system, so it need a communication mode which can realize point to multipoint. The current widely used communication Passive Optical Network technology (Passive Optical Network, PON) can effectively meet the point to multipoint and to solve the problem. With the continuous development of our country grows and the urban planning and development started late, many grid are combined with the old and the new, network structure is complex, the access layer of the passive optical network topology optimization is especially important. The passive optical network topology optimization in the process of light (OBD) splitter and optical network unit (ONU) planning is the implementation of the optimization of the main path.On the basis of introduce the current development trend of the power distribution automation system in domestic and overseas,power distribution network architecture based on EPON technology,network mode,and the characteristics of multistage power EPON,this paper,firstly,use a city’s power supply distribution network communication system as the prototype,analysis the PON optimization problems in the construction of power distribution automation system,and propose the no source optical network optimization model.Secondly,studied the standard genetic algorithm (SAG) and the basic idea of the adaptive genetic algorithm (AGA),the mathematical principle, characteristic and application,and applies them in the network optimization model of the the no source optical network,through the algorithm to implement the code and then simulate it,use the method of contrast to research their minimum cost and convergence performance,the results showed that the adaptive genetic algorithm has a better optimization performance.Thirdly,We put up an adaptive genetic algorithm based on the similarity of the parent(AGABS),it based on the individual parent genetic similarity Sim(X,Y)to decide the crossover probability and the mutation probability, and adopted the double cross point of crossover to complete the crossover operationJBy writing an algorithm to implement the code, and simulate it,through comparing the optimization of the same model,AGABS algorithm’s cost optimization ratio has reached30%,17%higher than the general genetic algorithm,and6%higher than the standard adaptive genetic algorithm,the performance is better than the ordinary standard genetic algorithm and the adaptive genetic algorithm,fully proved that the improved algorithm in the power stage of EPON optimization has a very good application value.