| Substation location and feeder planning is an important part of electric power system planning. Substation location influence network configuration, power supply quality and running cost of future electric power system. Good feeder configuration can improve on power supply quality and decrease running cost of electric power system. So it's significant for electric enterprise to make a good fist at substation location and feeder planning.To solve substation location problem we often abstract it and found a mathematic model at first. Then we can use the algorithms in existence or present a new algorithm to slove the mathematic model. Many mathematic models of substation location have been found, and many algorithm have been presented to solve them. Classical single and multiple continuum location optimal models are mostly used. They abstract loads as points, and simplify the issue to find the point which makes running cost of electric power system minimum as optimal substation location. The single location optimal model is solved by the method of iteration which is often used to solve extremum problem. The multiple location optimal model is solved by the method which is called location, distribution by turns. Methods such as branch and bound algorithm, mixed integer programming, transportation model and separable linear programming method have been used in recently years. Others are heuristic methods like genetic algorithm. These methods simplify the issue, so their purposes are very clear.However, aforementioned methods have shortcoming. At first, substation location models are getting more and more complex, and their dimensions become more. To solve the problem they need predigest it, so the result will be affected. Secondly, their aims are made the investment of substations, feeders or running cost of electric powersystem minimum. But they don't take into account of land utilization, street accessibility of right of way, etc. , but demands of electric power system. Generally, a straight line distance, namely, Euclid distance, was used for calculating the distance between a load point and a substation, which is a straight forward but non-practical way for most planning techniques. So they can hardly give the optimum solution in real life. Finally, some methods automatically select the optimal location of distribution substations without candidate substation location. And they may be optimal in a shortterm. But the substation location can't be optimal in a long range unless it is directed by the long term distribution planning. We select substation locations from potential position to make the distribution system optimal in a long term.With a Geographic Information System (GIS), W. M. Lin, et al, in a paper present an effective method, with the use of actual street routes in feeder planning, to determine the optimal substation location and feeder configuration based on the shortest path.In this paper, we present a model based on GIS, and solve it by a disperse method. The method's aim is made the investment of substations and feeder minimum. It can model physical constraints with land utilization and street accessibility and used actual street routes in the substation location and feeder planning, which is important especially in a suburban or city area. We can take into account the opinion of electric experts which greatly affected the result. With GIS we can obtain element data such as coordinate of load, distance from one load to another load. A shortest-path algorithm based upon dynamic programming is used in this paper to determine the distance between a load point and a substation along the city streets which we called generalized Euclid distance. According to the physical distance, optimal substations location can be selected, and in sequent a load point is assigned to a specific service area. Thus the optimal feeder planning can be determined.Substation location requires not only making the investment of substations and feeders minimum but also meeting the demands such as load, lan...
|
PDF Full Text Request