Satisfaction assessment of multi-objective schedules using neural fuzzy methodology

In the rapidly changing manufacturing environment and marketplace of today, the selection and evaluation of an appropriate manufacturing policy is a vital issue. However, this is a difficult task for decision-makers because manufacturing objectives are multiple, complex, conflict and often vague. This paper presents a methodology for a satisfaction assessment of multi-objective schedules with results from a scheduling simulator. In addition, a general objective structure appropriate for multi-objective manufacturing scheduling is also proposed for use in satisfaction assessment methodologies. The general objective structure is a hierarchical structure of the scheduling objective that consists of a goal, subgoal and criteria. The assessment methodology was based on a neural network and Technique for Order Performance by Similarity to Ideal Solution (TOPSIS), which is a common technique for multi-attribute decision-making problems. This paper extends TOPSIS by using fuzzy logic to deal with inaccurate and linguistic attributes in the general objective structure and a neural network for the weight calculation of inter-attribute importance.