AN OBJECT-ORIENTED FUZZY CONTROLLER FOR DYNAMIC SEQUENCING OF FLEXIBLE MACHINES

Most real world manufacturing systems operate under conditions that entail significant information delays. Hence, there is a need to model, analyze and evolve control strategies that can perform well under such delays. This paper introduces a fuzzy associative memory-based control strategy, the generalised fuzzy sequencing strategy (GFSS), to cope with status information delays. In demonstrating the use of GFSS for the single machine, three queue dynamic sequencing problem wherein information delays manifest in the form of machine setup times, the paper identifies suitable input and output control variables and suggests their appropriate fuzzification. We define the relative work-in-process, relative setup time, relative service rate, and relative traffic intensity as four input fuzzy control variables. The output fuzzy variable is the queue confidence level. Using the input and output fuzzy variables, the design of GFSS is described together with its implementation in an object-oriented environment. Simulation results for the average part flowtime and average work-in-process level performance measures demonstrate the efficacy and potential of using fuzzy control in situations where machine setup times are significant. Importantly, although designed for the three-queue single-machine dynamic sequencing problem, GFSS is generalizable for the n-queue case in a straightforward manner.