Abstract:
This dissertation focuses on the development of an intelligent prone methodology for efficient and effective handling of scheduling problems. In industrial concerns the issues/problems related to resource scheduling arise from abrupt and sudden demand and want patterns due to clustered and unbalanced supply of resources and assets. A novel technique in this respect has been developed and implemented on cases from industry. This technique takes into stride the efficiency demonstrated by various PSO (Particle Swarm Optimization) inspired techniques, combined with the intelligent prone ANNs (Artificial Neural Networks). Novelty and uniqueness is demonstrated through amalgamation of these approaches to introduce a term: NaACO (Neural Augmented ACO). This formulation is done under the umbrella of introduction of yet another unique approach i.e i-ACO (Intelligent ACO) theme through Neu(Tau) or Neuτ.This thesis also focuses on how ACO takes into account and absorbs the neural aspect of supervised and unsupervised learning. The intention of this research is to come up with a unique, customized and yet efficient way to handle the problems of the industry under given limitations and constraints. A complete model for this approach is built and for the application of the model a high technology aviation maintenance industry (case study I) is selected along with a medium technology manufacturing setup (case study II). The usage of ACO meta-heuristic is taken as an ideal reference point with which every problem set can be converged towards a best fit solution. Subsequently ANN is used to come up with a combitorial dialog box to prompt for the inputs and evaluate the outputs of the given problem. The thesis contributes to current research by introducing NaACO, i-ACO through intelligent scheduling (hence introducing neuτ); which proposes many solutions of the existing problems. The discussion and conclusions part at the end summarizes the research and the future areas of research are also elaborated to assist and appreciate future researchers who are interested to endeavor in this field and related applications.
