A learning automaton-based approach for power loss minimization and voltage profile enhancement in large-scale distribution systems

The reconfiguration problem consists of finding a new network topology with minimal power losses, while all the system constraints such as radial structure, lines power flow below capacity limits, node voltage magnitude within limits and all nodes connected are satisfied. This is a combinational optimization problem where the aim is to specify the final status of all switches (open/closed,) in a large-scale distribution system. Although there are a plenty of methods in the literature, but comprehensive analysis of bus and line failure has not been accomplished and just a limited version of failure has been studied. This paper presents a learning automaton-based algorithm for reconfiguration of large-scale distribution systems with assumption of probabilistic failure in both of the buses and lines. The main objective of the proposed algorithm is to minimize the power loss and voltage deviation, and also to maintain the distribution system in radial structure. To demonstrate the applicability of the proposed algorithm, it is tested on two standard IEEE sample systems and the obtained results are compared with other methods. Also, the numerical result indicates that the proposed method supplying power to the non-faulted areas with minimal power loss and maintains the radial structure of distribution system under abnormal conditions.