Adaptation of multi-agent manufacturing control by means of genetic algorithms and discrete event simulation

In this paper we apply Genetic Algorithms to adapt the decision strategies of autonomous controllers in heterarchical manufacturing systems. The basic idea of our approach is to let the control agents use pre-assigned decision rules for a limited amount of time, and to define a rule replacement policy propagating the most successful rules to the subsequent populations of concurrently operating agents. The twofold objective of this schema is to automatically optimize the performance of the control system during the steady-state unperturbed conditions of the manufacturing floor, and to improve the reactions of the agents to unforeseen disturbances (e.g. failures, shortages of materials) by adapting their decision strategies. Results on a simulated benchmark confirm the effectiveness of the approach.