Stable fault adaptation in distributed control of heterarchical manufacturing job shops

In this paper, a control theoretic model is developed for analyzing the dynamics of distributed cooperative control systems for manufacturing job shops with multiple processing steps with parallel dissimilar machines in which parts control their release times autonomously. The model allows an arbitrary number of part types to be produced using an arbitrary number of machines with an arbitrary number of alternate routings. Conditions for global stability of the resulting distributed control system with nonlinearities are shown using results from Lyapunov stability theory. System stability is found to be robust to a variety of faults and disturbances that may be encountered in a manufacturing environment as long they are bounded in the mean. Feedback enables implicit adaptation to faults in real time, which allows the flexibility in the systems to be fully utilized to compensate for faults and disturbances. Numerical simulation experiments are used to illustrate the global stability and the distributed fault adaptation capability of the system without requiring explicit notification or compensation to conditions such as machine deterioration, multiple machine failures, and network communication delays. Simulation results for job shops with 2000 parts are also presented to illustrate the scalability of the approach.