Multiobjective service restoration in distribution networks using an evolutionary approach and fuzzy sets

In this article, the service restoration (SR) problem in electrical distribution networks is dealt with using an evolutionary strategy (ES) with a fuzzy definition of the conflicting objectives. The normal operation status allows the remote control of tie-switches, of capacitor banks and load connection. When a permanent fault occurs, the same remote control actions can be performed with the aim of restoring the service in the concerned areas. The status of these remotely controllable elements is the boolean optimisation variables for the SR problem. Besides this, here the SR problem is dealt with in a multiple objectives (MO) formulation. Indeed, the power losses' term is considered as a further objective to be minimised, together with the primary objective of maximising the number of supplied loads. Generally, the MO formulation of an optimisation problem requires a unique expression for the global objective function. In this particular case, the used ES approach necessarily requires the definition of a 'global performance' index, which is derived on the basis of the fuzzy sets theory, outperforming the weighed sum formulation of the same problem. After a brief discussion on the SR problem and a short review of the state-of-art on the topic, the proposed ES and the fuzzy MO formulation of the SR problem is presented in detail. Results obtained using this procedure applied to a test system are presented and discussed. (C) 1999 Elsevier Science Ltd. All rights reserved.