Optimization of Supply Chain based on JIT Pull Control Policies: An Integrated Fuzzy AHP and ANFIS Approach

摘要

This Paper introduces an integrated Fuzzy Analytical Hierarchy (FAHP) and Adaptive neuro-Fuzzy Inference System (ANFIS) to evaluate Just In Time (JIT) control strategies. Three Pull Control Policies (PCPs), Kanaban, ConWIP, and Kanabn-ConWIP Hybrid systems, are identified for implementation. The proposed approach of this study creates a framework for identifying the alternatives and criteria, evaluating the PCPs, and comparing the performance of each policy. The approach is examined by studying a real multi-echelon, multi-stage, and multi-product supply chain network from automotive parts industry. The approach exemplifies the PCPs mechanisms, measurement criteria formulations, and integration of fuzzy theory with multi criteria decision making (MCDM) methods and ANFIS. The evaluation of the PCPs are based on JIT criteria such as inventory level, lead time, and lost demands. Discrete event computer simulation results are the basis of expert interpretation of each policy's performance. The FAHP method is applied to systematically measure the performance of each system. Then, Three ANFIS models are developed for each PCP based on the FAHP input-output results. Finally the ANFIS and FAHP methods are compared as well as the three PCPs.