POWER DISTRIBUTION SYSTEM RECONFIGURATION USING FAST NONDOMINATED SORTING GENETIC ALGORITHM AND GRAPH THEORY

摘要

D i s t ribution system reconfiguration (D SR) is a multiobjective,rnnon-linear problem. This paper introduces arnnew, fast, non-dominated sorting genetic algorithmrn(FNSGA) for solving the DSR problem in normalrnoperation by satisfying all objectives simultaneously withrna relatively short computational time. The goals of thernproject are to minimize real power losses, and improvernthe voltage profile and load balancing index withrnminimum switching operations. Instead of dealing with arnsingle objective function while the others are formulatedrnas constraints, as in the traditional methods, the FNSGArnis applied to optimize all the objectives according to thernoperator’s wishes. Moreover, an adapted mutationrnoperation is applied instead of a random one to speed uprnconvergence. Radial topology is satisfied using graphrntheory by formulating the branch-bus incidence matrixrn(BBIM) and checking the rank of each topology. Thernalgorithm is applied to two different systems, the IEEE16rnbus and the IEEE 32 bus test systems. The results showrnthe efficiency of this algorithm as compared to otherrnmethods in terms of both achieving all the goals andrnminimizing the computational time with reasonablernpopulation and generation sizes.