关键词: 模糊神经网络, 启发式算法, 可重构制造系统, 制造单元

Abstract: To improve the efficiency of manufacturing cell formation and its practicability, a manufacturing cell formation approach combing Fuzzy Adaptive Resonance Theory Neural Network (FARTNN) and heuristic algorithm was proposed. Guided by ideas of matching of similar machining modes, this approach was divided into two stages. Firstly, the improved FARTNN was used to form the basic part family and equipment sets. Then, based on the basic part family and equipment sets, a rule-based heuristic algorithm for resource allocation was presented to resolve the problems of equipment sharing, alternative equipment allocation and unbalanced equipment load in the manufacturing cell, etc. Finally, the validity of the approach was proved by its application in a workshop of a manufacturing enterprise.

Key words: fuzzy neural network, heuristic algorithm, reconfigurable manufacturing system, manufacturing cell