Node-Depth Encoding with recombination for multi-objective Evolutionary Algorithm to solve loss reduction problem in large-scale distribution systems

Power loss reduction problem can be achieved by an adequate Distribution System Reconfiguration (DSR). Usually DSR for power loss reduction is formulated as a nonlinear, multi-objective and multi-constrained optimization problem, containing thousands of constraints equations for large-scale networks. As a consequence, to find an adequate solution for such problem is computationally complex. Recently a practical and efficient approach to solve this problem which overcomes such a hurdle was developed. This approach, called MEAN, combines a multi-objective Evolutionary Algorithm (EA), based on subpopulations tables, with a new tree encoding, named Node-Depth Encoding (NDE). This paper presents a new genetic recombination operator, named Evolutionary History Recombination NDE operator, which improves the MEAN performance to solve the loss reduction problem in large-scale networks. Tests with large networks show that with the new operator the MEAN can find network configurations with loss reduction of 29.57% for networks with 5,166 switches. This result is surprising since the practitioners used to expect at most a 10% reduction by means of DSR.