Distribution Network Reconfiguration Based On Improved Harmony Search Algorithm

Distribution network reconfiguration refers to a kind of structure which network loss, voltage quality, load balancing and other targets reach the optimal value while meeting the constraints consist of ensuring the maintenance of radial structure, meeting the line current, voltage amplitude, transformer capacity and so on. It does not require additional investment and it is an important means to optimize the operation of distribution system. Distribution network reconfiguration is a large-scale nonlinear mixed integer programming problem which many scholars both at home and abroad have much deep analysis and research on it and draw many possible methods. However, most of the methods are subject to various limitations, thus lacking good applicability. Based on results of predecessors'research, this paper proposed an improved harmony search algorithm and new branch group set off principle which used on the simulation analysis of distribution network reconfiguration.This article briefly described the existing distribution network reconfiguration algorithms and their details at first and then analyzed and compared the advantages and disadvantages of these methods.Analyzed topology of distribution network after brief introduction about the structure of distribution network. Described various methods about power flow calculation in distribution system meanwhile highlighted the forward and backward substitution method.This paper researched and analyzed deeply into the basic principles of harmony search algorithm, analyzed the influence which fixed setting of parameters have on the algorithm optimization speed and overall performance, such as harmony memory size, harmony memory considering rate, pitch adjusting rate and bandwidth of tone adjustment. Proposed particle swarm algorithm incorporated with harmony search algorithm, take the parameters, especially harmony memory considering rate and pitch adjusting rate as the original individual, that is harmony vector's symbiosis individual. Used harmony search algorithm to guide the original population's evolutionary while particle swarm algorithm to the symbiosis population in order to overcome constraint which fixed setting of parameters have on harmony search algorithm's search capabilities and speed.Then, on the basis of brief depiction over the methods dealing infeasible solutions, T-point of the new definition, classification of T-point and related concepts were proposed as well as provided corresponding branch group set off principle so as to reduce the generation of infeasible solutions effectively. Used the combination of two above methods, that is improved harmony search algorithm with particle swarm algorithm guiding parameters' evolution and branch group set off principle to reconstruct also verified the feasibility and effectiveness of the proposed method by examples and compare to the effect of other reconstruction algorithms.