Evaluation of the performance of space reduction technique using AC and DC models in Transmission Expansion problems

Abstract

Transmission Expansion Planning (TEP) is anoptimization problem that has a non-convex and combinatorialsearch space so that several solution algorithms may converge tolocal optima. Therefore, many works have been proposed to solvethe TEP problem considering its relaxation or reducing its searchspace. In any case, relaxation and reduction approaches shouldnot compromise the quality of the final solution. This paper aimsat analyzing the performance of a search space technique using aConstructive Heuristic Algorithm (CHA) admitting that the TEPproblem is then solved using a Discreet Evolutionary ParticleSwarm Optimization (DEPSO). On one hand the reductionquality is performed by analyzing whether the optimal expansionroutes are included in the CHA constrained set and, on the otherhand, the relaxation quality of the DC model is analyzed bychecking if the optimal solution obtained with it violates anyconstraint using the AC model. The simulations were performedusing three different test systems. The results suggest that theproposed CHA provides very good results in reducing the TEPsearch space and that the adoption of the DC model originatesseveral violations if the full AC model is used to model theoperation of the power system.