1846 words - 8 pages

In most real-world problems, especially in a wide range of engineering disciplines it is required to optimize a number of objectives that are often in contradict with each other as well, these forms MOPs. In these problems since the utopian solution scarcely exists, instead of a single point, a set of solutions would exist for the problem. Each member in the solution set represents a trade-off between the multiple objectives of the MOP. These solutions, which are known in the literature as either the non-inferior, Pareto-efficient or the Pareto optimal solutions are of great importance. In fact, these solutions equip the decision-maker with available, efficient scenarios when he/she comes to ...view middle of the document...

In Ref. [3] a heuristic branch exchange reconfiguration method for power loss reduction and load balancing in radial EDSs is propounded; however the method is slow in terms of computation and dependent on the initial network configuration. In Refs. [4,8] heuristic network reconfiguration schemes for voltage stability enhancement in EDSs are proposed. A geometrical reconfiguration approach is propounded in [5] for loss minimization in EDSs. In Ref. [6] an interesting heuristic procedure is designed based on optimal power flow (OPF) to minimize power losses in which power switches are modeled as continuous variables, despite their discrete nature. In Ref. [7] a prominent heuristic reconfiguration technique is suggested in the framework of a set of heuristic rules inspired from the experiences of EDS operators and system engineers to minimize power losses in the system. In Ref. [9] another heuristic plan is developed on the basis of identifying active and reactive power breakpoints in EDSs to minimize power losses. The approaches surveyed above are representative of some of the state-of-the-art heuristic approaches proposed so far in the field.

In the second group, we have EC-based approaches which are basically based on different types of meta-heuristics and evolutionary algorithms such as genetic algorithm (GA) [10–12]; Tabu search (TS) [13,14]; Harmony Search Algorithm (HSA) [15]; Ant Colony Optimization (ACO) [16]; Hybrid Optimization Algorithms like (PSO-SFLA) [17] and (PSO-NM) [18]; Imperialist Competitive Algorithm (ICA) [19] and Artificial Immune System (AIS) [20] to tackle the mentioned features pertaining to the network reconfiguration problem.

Nara et al. [10] were the pioneers in opening up and flourishing the application of artificial intelligence in solving power system problems by introducing a GA optimization scheme to solve EDNRC in 1992. In Ref. [11] a binary GA approach for network reconfiguration with an adaptive mutation process has been promoted to minimize losses in EDSs. Enacheanu et al. [12] used matroid theory and developed a GA approach to minimize power losses in distribution networks. Zhang et al. [13] introduced a Tabu search algorithm to find minimum loss configuration in large-scale EDSs. In Ref. [16] Ant colony optimization algorithm is deployed to solve the problem for the minimum loss configuration. There are also some other research works designed on hybrid EA reconfiguration optimizers such as PSO and shuffled frog leaping algorithm (PSO-SFLA) [17]; PSO in fuzzy frame incorporated with Nelder–Mead numerical method [18] suggested for loss reduction in the power system.

In terms of optimization technique, the proposed approaches in the field can be subsumed under three broad categories: Classic, Fuzzy and Modern approaches depending on the type of optimization technique employed to solve the MOP. In the first group we have Classic approaches [21–24] which typically use different types of meta-heuristics...

Get inspired and start your paper now!