Scenario analysis to account for photovoltaic generation uncertainty in distribution grid reconfiguration

Abstract

This paper considers hourly reconfiguration of a low voltage distribution network with the objectives of minimizing power loss and voltage deviation. The uncertainty in photovoltaic (PV) generation which in turn will affect the optimum configuration is tackled with the help of scenario analysis. In the proposed strategy of the scenario analysis called the Internal method, final non-dominated solutions (configurations) will have a relative acceptable performance in all scenarios. The use of scenario analysis helps in reducing the number of switches that needs to be installed in order to cater for the uncertainty in PV generation. This is possible because the PV generation values for a given hour during different days are taken into account by means of defining various scenarios when finding out the optimum configuration. Non-dominated sorting genetic algorithm (NSGA-II) used in this paper generates non-dominated solutions, demonstrating their trade-offs. Finally, the best compromise solution can be selected depending on the decision maker's requirement.