Recent advances in the modeling, scheduling and control of flexible automation

Abstract

This paper initially discusses the state-of-the-art and the current limitations in the modeling, scheduling and control of flexible automation. To model flexible automation, it is argued that the simulation tools must provide enhanced capabilities to consider both controller interactions and the flow of resources that support production.It is also demonstrated that scheduling and control must be considered concurrently in real-time to effectively manage flexible manufacturing systems (FMSs). The complexity of the modem FMS further requires that the integrated scheduling and control function must be distributed among several coordinators with the system. The second pti of the paper presents several research developments pertaining to the modeling, scheduling and control of flexible automation including: a Recursive, Object-Oriented Control Hierarchy for the integrated distribution of scheduling and control; a Hierarchical ObjectOriented Programmable Logic Simulator for the detailed modeling of FMSs; and a Hierarchical System Coordinator for implementing real-time scheduling and control. A physical emulator for an FMS which is being constructed is discussed as the example application in this presentation. Finally, the futureresearchplan and theeducationalprogram based upon the developments is outlined.

Multiprocessor Scheduling with Machine Allotment and Parallelism Constraints

“…Denote by w OPT (I ) the length of an optimal schedule of an instance I . First, we show that an optimal schedule can be obtained by relaxing the parallelism constraint 5 of each job by one. Specifically, if for all j, we run J j in parallel on ρ j + 1 (instead of ρ j ) machines, we can obtain a schedule of length w OPT (I ).…”

“…Production processes [5] typically involve the usage of consumable resources (i.e., special materials) which cannot migrate from one machine to another, and mobile resources (e.g., human supervision), which allow flexibility in the choice of machines. The maximal amount of consumable resources determines the allotment parameter of a production process; the available amount of mobile resources determines its parallelism parameter.…”

An object-oriented simulation model for communication network traffic at a general mail facility

“…Earlier simulation studies conducted in our laboratory demonstrated the importance of modeling controller interactions which coordinate the flow of all entities including jobs and supporting resources, such as tools (see Dullum andDavis 1992 andDavis et al 1993). As a result of this study, we now develop the complete object-oriented definition of the modeled system before attempting to employ a simulation language.…”

Initializing on-line simulations from the state of a distributed system

“…To this end, we have developed an even more comprehensive modeling paradigm called the Hierarchical Object-Oriented Programmable Logic Simulator (HOOPLS), which simulates the distributed control architecture by modeling the interactions among the controllers, see Davis, Macro, and Setterdahl (1997) and Davis, Setterdahl, Macro, Izokaitis, and Bauman (1993). The HOOPLS paradigm is based upon the belief that interactions among the controllers must be considered by the simulation model in order to accurately model a system with a distributed control architecture.…”