Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
Multi-Objective Optimization Using Evolutionary Algorithms
Multi-Objective Optimization Using Evolutionary Algorithms
UAI'96 Proceedings of the Twelfth international conference on Uncertainty in artificial intelligence
| Hi-index | 0.00 |
This paper proposes a new multi-objective approach for the problem of power restoration in (n-1) contingency situations. It builds on a previous, mono-objective approach introduced in Mendes et al. (2010) [14]. Power restoration normally relies on network reconfiguration, and typically involves re-switching and adjustment of tap-changers and capacitor banks. In this work, we focus on re-switching strategies. The quality of the re-switching strategy is measured in terms of voltage deviations, number of consumers still affected after the reconfiguration, number of overloaded branches and number of switches changes. Due to the number of criteria and conflicting objectives, power restoration is a prime candidate for multi-objective optimisation. The method studied is based on a genetic algorithm and was tested using two real-world networks, with up to of 1,645 branches and 158 switches. We present a contingency example for each network and discuss the results obtained. Finally, we discuss the approach's convergence by analysing the evolution of the solutions that compose the Pareto frontier.