On adaptive direction in competitive co-evolutionary process

抄録

In this paper an agent-based scheduling technique is proposed which capitalizes on the adaptive nature of competitive co-evolutionary process in scheduling sporadic user-submitted jobs for a manufacturing setup. The jobs are represented by agents who interact with other agents in forming groups based on mutual interest. Grouping the jobs provides adaptive direction for the co evolutionary process. The constant competition between jobs in a group and between groups themselves allows the scheduler to adapt to the dynamic nature of job submissions and come up with the best possible job sequence within the given constraints. Simulation results indicate that our proposed technique outperforms other conventional algorithms.