Nonlinear Model Predictive Control for Industrial Manufacturing Processes with Reconfigurable Machine Tools

Abstract

Manufacturing companies are faced with challenges to respond to volatile market demands quickly and flexibly while maintaining a cost-effective level of production. Instead of flexible working times, we adopt Reconfigurable Machine Tools (RMTs) to compensate for unpredictable events in case of bottleneck. To include these tools effectively on the operational layer, we propose a complementing feedback approach using model predictive control (MPC) together with genetic algorithm and branch and bound to achieve a better compliance with logistics objectives and a sustainable demand oriented capacity allocation. Further, the system stability is guaranteed by a trajectory-based unconstrained MPC scheme associated with the principle of flexible Lyapunov functions. The effectiveness and plug-and-play availability of the proposed method is demonstrated via a four-workstation job-shop system, which shows that the work in process can be practically asymptotically stabilized by usage of RMTs.