Solving Extremely Difficult MINLP Problems Using Adaptive Resolution Micro-GA with Tabu Search

Non convex mixed integer non-linear programming problems (MINLPs) are the most general form of global optimization problems. Such problems involve both discrete and continuous variables with several active non-linear equality and inequality constraints. In this paper, a new approach for solving MINLPs is presented using adaptive resolution based micro genetic algorithms with local search. Niching is incorporated in the algorithm by using a technique inspired from the tabu search algorithm. The proposed algorithm adaptively controls the intensity of the genetic search in a given sub-solution space, i.e. promising regions are searched more intensely as compared to other regions. The algorithm reduces the chances of convergence to a local minimum by maintaining a list of already visited minima and penalizing their neighborhoods. This technique is inspired from the tabu list strategy used in the tabu search algorithm. The proposed technique was able to find the best-known solutions to extremely difficult MINLP/NLP problems in a competitive amount of time. The results section discusses the performance of the algorithm and the effect of different operators by using a variety of MINLP/NLPs from different problem domains.