Robust Supervisory Control for Automated Manufacturing Systems with Minimal Decentralized Switch-Buffer Controllers

Abstract

Automated manufacturing systems require effective supervisory control mechanisms to ensure their efficient operation and prevent fatal errors. Majority of the deadlock supervisory control policies for automated manufacturing systems are formulated based on the presupposition that the systems' resources are all reliable. However, in actuality systems' resources experience failure. Deadlocks and blockages could still occur in a system with failure-prone resources to which a supervisory controller is deployed if a failure-prone resource in the system fails. Extra buffers (redundant buffers) are usually used on production floors to store work-in-process parts in order to prevent deadlocks or blockages caused by the failure of failure-prone resources. However, additional buffers signify extra costs to the enterprise. To this end, this paper presents a robust supervisory control framework for automated manufacturing systems that utilizes minimum amount of decentralized buffer units and switch controllers, viz., switch-buffer controllers. The policy combines control places computed to prevent deadlocks in the system and switch-buffer controllers to prevent deadlocks/blockages in the event of failure-prone resource failure. Both the plant and the proposed supervisory controller are modeled using Petri nets. Examples are used to validate and demonstrate the effectiveness of the proposed approach. Furthermore, the results, compared with some existing techniques, indicate that the proposed technique has more advantages over them.