摘要 |
PROBLEM TO BE SOLVED: To simulate the autonomous actions of workers by producing a rule for every worker to define a place to move next and a job to do next every time a job is over and performing the simulation based on the produced rule. SOLUTION: A product is moved onto a work bench P2 for every worker and every time the job of a process 1 is over (step 1). The process 1 proceeds to a process 10 (step 3) if an in-process product exists at an in-process buffer Q4 (Yes of step 2). If no in-process product exists at the buffer Q4 (NO of step 2), the process 1 proceeds to a process 2 (step 4). Thus, a rule is produced for every job to be done and the simulation is performed based on this rule. Then a variable included in the rule is optimized by learning. |