Comparison of fuzzy inference system (FIS), FIS with artificial neural networks (FIS + ANN) and FIS with adaptive neuro-fuzzy inference system (FIS + ANFIS) for inventory control

Conventional inventory models mostly cope with a known demand and adequate supply, but are not realistic for many industries. In this research, the fuzzy inference system (FIS) model, FIS with artificial neural network (ANN) model and FIS with adaptive neuro-fuzzy inference system (ANFIS) model in which both supply and demand are uncertain were applied for the inventory system. For FIS model, the generated fuzzy rules were applied to draw out the fuzzy order quantity continuously. The order quantity was adjusted according to the FIS model with the evaluation algorithm for the inventory model. The output of FIS model was also used as data for FIS + ANN and FIS + ANFIS models. The FIS + ANFIS model was studied with three membership functions; trapezoidal and triangular (Trap), Gaussian and bell shape. Inventory costs of the proposed models were compared with the stochastic economic order quantity (EOQ) models based on previous data of a case study factory. The results showed that the FIS + ANFIS_Gauss model gave the best performance of total inventory cost saving by more than 75 % compared to stochastic EOQ model.     

Lexicographic fuzzy multi-objective model for minimisation of exceptional and void elements in manufacturing cell formation

This research proposes a lexicographic fuzzy multi-objective model based on perfect grouping for concurrent solving the part-family and machine-cell formation problems in a cellular manufacturing system. New simplified mathematical expressions of exceptional and void elements are proposed, opposing conventional quadratic and absolute functions. The main objectives of the proposed solution model, that is, the minimisation of both the number of exceptional elements and the number of void elements is defined by fuzzy goals as pre-emptive ordering. A lexicographic fuzzy goal model is developed to enhance cell performance and machine utilisation simultaneously. A satisfactory efficient solution can easily be obtained, and alternative solutions can also be generated by capturing flexibility of the proposed fuzzy multi-objective programming model. The formulated model can be solved by existing integer programming solvers. Finally, the evaluation of cell formation problems is briefly discussed to show the performance of the proposed model.     

Genetic algorithm for short-term scheduling of make-and-pack batch production process

This paper considers a scheduling problem in industrial make-and-pack batch production process. This process equips with sequence-dependent changeover time, multipurpose storage units with limited capacity, storage time, batch splitting, partial equipment connectivity and transfer time. The objective is to make a production plan to satisfy al constraints while meeting demand requirement of packed products from various product fam-ilies. This problem is NP-hard and the problem size is exponentially large for a realistic-sized problem. Therefore, we propose a genetic algorithm to handle this problem. Solutions to the problems are represented by chromo-somes of product family sequences. These sequences are decoded to assign the resource for producing packed products according to forward assignment strategy and resource selection rules. These techniques greatly reduce unnecessary search space and improve search speed. In addition, design of experiment is carefully utilized to de-termine appropriate parameter settings. Ant colony optimization and Tabu search are also implemented for com-parison. At the end of each heuristics, local search is applied for the packed product sequence to improve makespan. In an experimental analysis, al heuristics show the capability to solve large instances within reason-able computational time. In al problem instances, genetic algorithm averagely outperforms ant colony optimiza-tion and Tabu search with slightly longer computational time.