In this study, a hybrid approach between two new techniques, Genetic Algorithms and Artificial Neural Network, is describes for generating Job Shop Schedules (JSS) in a discrete manufacturing environment based on nonlinear multi-objective function. Genetic Algorithms (GA) is used as an effective search technique for finding an optimal schedule via a population of gene strings which represent alternative feasibility schedules. GA propagates a new population of genes through a number of cycles called generations by implementing natural genetic mechanism. The other technique is an Artificial Neural Network that performs multi-objective schedule evaluator. The intention is to establish an effective model that maps a complex set of values provided by experienced expert schedulers. This neural network is later combined with Genetic Algorithm. Specifically, GA uses the Neural Network evaluator to access the fitness performance of gene strings stimulated.
The proposed approach is prototyped and tested on four different JSS problems based on the sizes of the problem, namely small, medium, large, and real problems provided by a company. The comparative results indicate that the proposed approach is consistently better than those of heuristic algorithms used extensively in industry.
Return to Dissertations.
Last update: January 18, 1996