An intelligent multi-objective stochastic framework to solve the distribution feeder reconfiguration considering uncertainty

This paper deals with the optimal operation management of the distribution feeder reconfiguration (DFR) considering the uncertainty effects. In contrast to the conventional objective functions, this paper considers the System Average Interruption Frequency Index (SAIFI) as a reliability index. Meanwhile, the total active power losses and the voltage deviation objective functions are considered as the other targets too. In order to make the analysis more practical, the uncertainty associated with the active and reactive load forecast errors are modeled in a stochastic framework based on 2m Point Estimate Method (PEM). In the proposed stochastic optimization framework, an external memory called repository is defined to store the non-dominated solutions which are found during the optimization process. Also, a fuzzy based clustering technique is defined to keep the size of the repository within the predefined limits. Since the proposed problem is a nonlinear, discrete complex optimization problem, this paper proposes an intelligent self adaptive modified optimization algorithm based on theta-firefly algorithm to solve the optimal multi-objective stochastic DFR problem suitably. The proposed self-adaptive modification method consists of three sub-modification techniques which let each firefly choose the sub-modification method that best suits its situation adaptively. The feasibility and superiority of the proposed method is tested on a standard IEEE test system.