Multi-agent system for dynamic manufacturing system optimization

Abstract

This paper deals with the application of multi-agent system concept for optimization of dynamic uncertain process. These problems are known to have a computationally demanding objective function, which could turn to be infeasible when large problems are considered. Therefore, fast approximations to the objective function are required. This paper employs bundle of intelligent systems algorithms tied together in a multi-agent system. In order to demonstrate the system, a metal reheat furnace scheduling problem is adopted for highly demanded optimization problem. The proposed multi-agent approach has been evaluated for different settings of the reheat furnace scheduling problem. Particle Swarm Optimization, Genetic Algorithm with different classic and advanced versions: GA with chromosome differentiation, Age GA, and Sexual GA, and finally a Mimetic GA, which is based on combining the GA as a global optimizer and the PSO as a local optimizer. Experimentation has been performed to validate the multi-agent system on the reheat furnace scheduling problem.