The thesis addresses the problem of control in complex cyber-physical production systems, where traditional methods are no longer effective due to the complexity and interconnectedness of elements. A multiagent approach to control is employed, enabling flexible and adaptable management of challenging scenarios. A multiagent system model is developed, in which autonomous agents collaborate based on local information and goals to achieve effective global system behavior. The hypothesis that this approach is more scalable, robust, and resilient than centralized approaches is tested through two case studies: pressure control in industrial compressed air systems and task allocation to autonomous mobile robots in intralogistic systems. The results demonstrate that the system's scalability, robustness, and resilience are improved by the multiagent approach, confirming the hypothesis.
|
