In most real-world problems especially in a wide range of engineering disciplines it is required to optimize a number of objectives which are mostly conflicting with each other as well, this forms MOPs. In these problems since the utopian solution which is considered as best with respect to all objectives scarcely exists, instead of a single point, a set of solutions would exist for the problem where each member of 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 a decision making process.
In the field of power system operation, distribution network reconfiguration is an important way to operate EDS more efficiently which in mathematical terms can be viewed as a constrained nonlinear non-differentiable optimization problem due to the specific features pertaining to this arduous combinatorial MOP which encourages Meta-Heuristics as well as Artificial Intelligence (AI) and heuristic techniques to solve the problem.
The history of EDNRC dates back to 1975 when the first approach was proposed in the field . The technique was basically based on branch and bound optimization technique to find the minimum loss configuration in EDSs; however it suffers from computational deficiency and convergence aspects. Generally speaking, EDNRC approaches can be divided into two discrete categories. The first one consists of heuristic approaches [2–9] and in the second artificial intelligence is used as the optimization tool [10–20].
In Ref.  a heuristic sequential branch opening reconfiguration method is propounded to minimize resistive line losses in EDSs; although the method is fast in terms of computation but it suffers from premature convergence. In Ref.  a heuristic branch exchange reconfiguration method for power loss reduction and load balancing in radial EDSs is propounded; however in terms of computational time their solution technique is highly expensive 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  for loss minimization in EDSs. In Ref.  an interesting heuristic procedure is proposed based on optimal power flow (OPF) to minimize power losses in which power system switches are modeled as continuous variables despite their discrete nature. In Ref.  a prominent heuristic reconfiguration technique is suggested in the framework of a set of heuristic rules based on the experiences of EDS operators and system engineers to minimize power losses in the system. In Ref.  another heuristic plan is developed on the basis of identifying active and reactive power break-points in...