N-1 Based Optimal Placement of SVC Using Elitist Genetic Algorithm in Terms of Multiobjective Problem Statement

摘要

Processes of new reactive power compensation devices (RPCD) installation or replacement of old ones are becoming more relevant due to the development of power electronics. An emphasis is made on more efficient and flexible FACTS (Flexible Alternating Current Transmission System) devices - mainly static var compensators (SVC) and static synchronous compensators (STATCOM). In addition, it is necessary to solve a number of problems related to the search for the optimal location of an RPCD and the selection of its parameters. It should be carried out for the purpose of achieving the highest technical and economic effect for a power system as a whole or its region. The multiobjective optimization (MOO) methods, particularly heuristic ones (genetic algorithms, particle swarm methods, simulated annealing etc.), are proven to be efficient in the analysis of power systems planning. The drawbacks of these methods are well known, namely their dependency on forms of objective functions. The importance of taking N-1 criterion into account in questions of optimal placement of an SVC was shown in the article. As a consequence, the security indices were suggested to use as the additional objective functions. Moreover, the range of probabilistic parameters is not limited to expected values of the objective functions components but also it includes moments of higher orders. The test results have shown that the forms of objective functions and plenty of the factors taken into account (e.g. post-contingency states) significantly influence the problem solution. It was demonstrated that the optimization results possess great degree of uncertainty in the case of large number of criteria for the selection of a bus for an RPCD.