Integration of a Holonic Organizational Control Architecture and Multiobjective Evolutionary Algorithm for Flexible Distributed Scheduling

Based on the concept of autonomous cooperating holons, this paper presents a holonic command and control ( C ²) organizational control architecture (OCA) that models a C ² organization as an integration of holonic multilevel decentralized decision-making networks. The OCA consists of two levels: operational- and tactical-level controls. Authority and control are highly distributed among agents belonging to different levels of the holarchy to empower the edges, whereas the integration of decisions is ensured to achieve overall mission objectives. In order to complete a mission in real time in dynamic environments, the decision makers (DMs) at different control levels need to coordinate their actions extensively and be prepared to adapt their schedules. Based on the proposed OCA, we present a holonic multiobjective evolutionary algorithm that produces flexible distributed schedules that account for the unexpected changes in the mission environment, such as asset breakdown, appearance of new events, DM failure, etc. This approach generates multilevel Pareto optimal solutions and, as a consequence, produces a set of ranked neighboring schedules. The actual schedule is a combination of different phases from alternative neighboring schedules that adapt to environmental disturbances. Moreover, the cost of adaptability is reduced while maintaining the stability of the organization. Numerical experiment shows the advantages of the proposed OCA, viz., simplicity, efficiency, and flexibility, which enable an organization to achieve high performance under dynamic and uncertain environments.