The Application Of Genetic Simulated Annealing Algorithm In The Urban Network Reconfiguration

As modern urban distribution networks have been developing fast, the network structure is increasingly complex and the demand of power load is big, the power consumer's requirement for improving power quality and supply reliability is becoming more and stronger. Network reconfiguration is an important way to the operation and control of the distribution system, and it is the important means to reducing network losses, eliminating overload, balancing load and improving voltage quality.In theory, distribution network reconfiguration is a large-scale, mixed integer, non-linear combination optimization problem. According to the characteristics of distribution network, the distribution network reconfiguration based on genetic annealing algorithm aiming at the minimization of power losses is researched, combining with no current and voltage limits violation, network maintaining radial, and other constrained conditions.In distribution network reconfiguration, the simulated annealing algorithm is adopted to avoid the problem of falling into local convergence when using genetic algorithm alone. In the mixed algorithm, some improvements are made combined with the characteristics of urban distribution network, such as using the based on"tree"chromosome coding method to coding the section switches in the trees and connection switches respectively, executing the operation of crossover and mutation only in the range of trees with the adjustment by section switches to avoiding islands and loops, using both the best saving tactics and the roulette selection to ensure the unity and variety of population.The examples show that the mixed genetic annealing algorithm can solve the problem of distribution network reconfiguration effectively with satisfied search speed and convergence.