Optimization Of Fuzzy Logic Controllers With Genetic Algorithm For Two-Part-Type And Re-Entrant Production Systems
Homayouni, Seyed Mahdi (2008) Optimization Of Fuzzy Logic Controllers With Genetic Algorithm For Two-Part-Type And Re-Entrant Production Systems. Masters thesis, Universiti Putra Malaysia.
Improvement in the performance of production control systems is so important that many of past studies were dedicated to this problem. The applicability of fuzzy logic controllers (FLCs) in production control systems has been shown in the past literature. Furthermore, genetic algorithm (GA) has been used to optimize the FLCs performance. This is addressed as genetic fuzzy logic controller (GFLC). The GFLC methodology is used to develop two production control architectures named “genetic distributed fuzzy” (GDF), and “genetic supervisory fuzzy” (GSF) controllers. These control architectures have been applied to single-part-type production systems. In their new application, the GDF and GSF controllers are developed to control multipart- type and re-entrant production systems. In multi-part-type and re-entrant production systems the priority of production as well as the production rate for each part type is determined by production control systems. A genetic algorithm is developed to tune the membership functions (MFs) of input variables of GDF and GSF controllers. The objective function of the GSF controller is to minimize the overall production cost based on work-in-process (WIP) and backlog cost, while surplus minimization is considered in GDF controller. The GA module is programmed in MATLAB® software. The performance of each GDF or GSF controllers in controlling the production system model is evaluated using Simulink® software. The performance indices are used as chromosomes ranking criteria. The optimized GDF and GSF can be used in real implementations. GDF and GSF controllers are evaluated for two test cases namely “two-part-type production line” and “re-entrant production system”. The results have been compared with two heuristic controllers namely “heuristic distributed fuzzy” (HDF) and “heuristic supervisory fuzzy” (HSF) controllers. The results showed that GDF and GSF controllers can improve the performance of production system. In GSF control architecture, WIP level is 30% decreased rather than HSF controllers. Moreover the overall production cost is reduced in most of the test cases by 30%. GDF controllers show their abilities in reducing the backlog level but generally production cost for GDF controller is greater than GSF controller.
Repository Staff Only: Edit item detail