In this paper, we introduce a genetic algorithm approach for set covering problems. Since the set covering problems are constrained optimization problems we utilize a new penalty function to handle the constraints. In addition, we propose a mutation operator which can approach the optima from both sides of feasible/infeasible borders. We experiment with our genetic algorithm to solve several instances of computationally difficult set covering problems that arise from computing the 1-width of the incidence matrix of Steiner triple systems. We have found better solutions than the currently best-known solutions for two large test problems.
|出版狀態||已出版 - 1994|
|事件||Proceedings of the 1st IEEE Conference on Evolutionary Computation. Part 1 (of 2) - Orlando, FL, USA|
持續時間: 27 6月 1994 → 29 6月 1994
|???event.eventtypes.event.conference???||Proceedings of the 1st IEEE Conference on Evolutionary Computation. Part 1 (of 2)|
|城市||Orlando, FL, USA|
|期間||27/06/94 → 29/06/94|