Power loss reduction can be considered as one of the main purposes for a distribution system’s designers and operators, especially for recent non-governmental networks. Moreover, the nature of power loss challenges different methods to solve this problem, while various studies indicate effectiveness of reconfiguration and its high portion for this case. Thus, “reconfiguration” can be introduced as an optimization procedure to obtain economical high quality operation by changing the status of sectionalizing switches in these networks. Some major points such as using different switch types, considering number of switching and time varying loads, which are almost neglected or not applied simultaneously in most pervious essays, are the main motivation to propose this chapter. A heuristic practical scheme is proposed to perform optimal reconfiguration, and all previous neglected topics are fully discussed. Proposed method will apply to sample distribution networks, and the effectiveness of this method will be discussed through several case studies and comparisons.
TopIntroduction
Distribution network reconfiguration has a history about 35 years1 and literatures indicate initial efforts by Merlin and Back (1975). Their heuristic work focused on power loss reduction where it started with closing all network switches to perform a mesh configuration. Switches are then opened successively to restore radial configuration with less power loss. Thereafter, this method got some improvements by Shirmohammadi and Hong (1989) where introduced as “Sequence Switch Operation Method”2. Here also, the reconfiguration procedure starts by closing all network switches which are then opened one after another so as to establish the optimum power flow in the network. This is while, many approximations of pervious method have been overcome in this algorithm and the computation takes less time.
Civanlar et al. (1988) made use of method which is known as a “Branch Exchange” operation. This method suggested the opening of any switch was required to correspond to the closure of another switch, ensuring that the radial configuration of the distribution network would be preserved. A year after, Baran and Wu (1989) succeeded to improve the method of Branch Exchange by offering two approximated power flow Equations in 1989. Their power flow Equations introduced by recursive approximation of P, Q and V for each branch.
Liu (1989) in the same year, Taylor, Chiang and Jeon (1990) in next years presented various heuristic method for distribution feeder reconfiguration. Wagner (1993), compares different major reconfiguration methods. This comparison depicts considerable economic advantages by using presented heuristic methods especially for real time operations. He indicated that an important loss reduction was obtained through simulations in Canadian networks during a one year period.
Genetic Algorithm applied in distribution system reconfiguration for reduction of real power loss by Nara et al. (1992) for the first time. They compared obtained results with the results of Simulated Annealing (SA) method and concluded that although the genetic algorithm has less assurance but it can be faster that SA. After introducing GA method and due to its effectiveness, many works done by refine or suggest heuristic techniques to this algorithm to compensate long solution time.
Sarfi et al. (1996) presented an algorithm based on “Network Partitioning Method” after their efforts to use refined genetic algorithm for reconfiguration of distribution networks. Before that, Chen and Cho (1993) have performed an analysis of hourly reconfiguration scheme. They have studied the hourly load patterns over an interval of a year in order to define the hourly load conditions for each season. They have used Branch and Bound technique for obtaining minimum loss configuration considering hourly load patterns over annual intervals. It was the year that Kashem et al. (1999) investigated the load rearrangement and proposed their famous method for load balancing in distribution systems.
Huang and Chin (2002) proposed an algorithm based on fuzzy operation in order to deal with feeder reconfiguration problem. They suggested this method to reduce power loss as well as to acquire the load balance. More Artificial intelligence based methods, such as (Ahuja, 2007; Mendoza, 2009; Milani 2010) are other useful procedures proposed for reconfiguration of distribution networks, recently.