Several methods have been proposed in Refs. [1][2][3][4] for optimizing the locations and sizes of distribution substations. But they are proposed either on the basis of giving candidate substations' locations in advance or on the assumption that the load density is identical for the whole area. Then, an approach reported in Ref. [5], which divides the whole problem into two sub-problems-substation locating and combination optimization, has been utilized and proved to be practical. It also has the advantage of being able to obtain the optimal size, location and service areas of each of the high voltage (above 35 kV) substations automatically, while having no need of getting the candidate substations' locations in advance [6]. However, the algorithm used in solving the substation locating sub-problem, with the name of substation locating, using the ALA algorithm, is found to be significantly affected by the substation's initial location. The ploughing-around combination algorithm, which is used in solving the subproblem of combination optimization, always results in a prohibitively long computation time when the scale of the problem is large, while the optimization of the solution may not be sure.To avoid the problems listed above, some modifications have been made in the ALA algorithm from Ref. [5], and the branch-and-bound technique is adopted to solve the subproblem of combination optimization, where some heuristic rules are utilized to cut the branches with low probability in generating the global optimal solution. Examples show that the approach proposed in this paper can obtain a better solution than that from Ref. [5]. With the computation significantly reduced, it becomes more practical in searching for an optimal planning solution of high voltage distribution substations.
Mathematical modelsThe optimization problem of the HV distribution substation is to make the decision on the location and size of each substation, which properly supplies the particular region at a given load level with less investment and operational cost.In this paper, according to the geographic distribution of actual consumers, the service area of the power distribution system is divided into lots of small irregular areas, which are called "sectors". Each has a load point in its center and acts Abstract Aimed at solving the problem of optimal planning for high voltage distribution substations, an efficient method is put forward. The method divides the problem into two sub-problems: source locating and combina tional optimization. The algorithm of allocating and locating alternatively (ALA) is widely used to deal with the source lo cating problem, but it is dependent on the initial location to a large degree. Thus, some modifications were made to the ALA algorithm, which could greatly improve the quality of solutions. In addition, considering the non-convex and nonconcave nature of the sub-problem of combinational optimization, the branch-and-bound technique was adopted to obtain or approximate a global optimal solution. To improve th...