Abstract
This chapter deals with the control system design problem for automated machine systems from the viewpoint of a discrete event system approach. Behaviors of automated systems have externally the features of discrete event asynchronous, concurrent processes, which implies the necessity of distributed architecture for intelligent cooperative control. Based on the generalization of multi-axis machine activities, the detailed control functions of a machine task are hierarchically represented by an interpreted form of the Petri net. The necessary control conditions and rules to ensure that the control system is well-defined, including synchronization and conflict resolution, are provided as a conceptual model of machine task. Due to hierarchical decomposition of Petri net models, the structure of the whole control system as well as the contents of each machine task is easily understood so that the task planning, monitoring and modification of the control system can be done effectively.
TopBackground
Petri nets and automata are the most used to describe discrete event system control. Modeling and analysis of discrete event systems with controllable and uncontrollable events, turned on and off by the supervisor, were proposed based on automata theory (Ramadge & Wonham, 1987). However, finite automata present some drawbacks such as the difficulty to model parallelism, synchronization and resource sharing. Although control commands generation was related to state transitions, since the connection of several finite automata is no longer a finite automaton, the communication specifications between modules can not be achieved using the finite automata framework (Lima & Saridis, 1996; Stadter, 1999; Silva et al., 2014). In spite of a great number of researches concerning advanced methods and tools to analyze the distributed models, these models are mostly constructed by empirical methods based on the knowledge of experts and customized for a particular application. Most approaches are limited by the combinatorial explosion that occurs when attempting to model complex systems.
On the other hand, Petri nets incorporate the notion of a distributed state of a system and a rule of state change (Murata, 1989; David & Alla, 1992), providing a mathematical formalism and a graphic tool for the formal representation of a system whose dynamics is characterized by concurrency, synchronization, nondeterministic decision, mutual exclusion and conflict. Computerized automation systems have the same, typical and most important features as industrial distributed systems.
Key Terms in this Chapter
Formal Methods: Mathematical techniques for the development of functional specifications and designs of real-time software.
Synchronization: In multiprocessing systems, joining multiple independent processes in order to reach an agreement or commit to a certain sequence of actions.
Discrete Event Systems: A dynamic system where the state evolution depends entirely on the occurrence of asynchronous discrete events over time, consisting of discrete state spaces and event-driven state transition mechanisms.
Manufacturing Machines: Complex, high-speed machines which process and combine materials as they move through the machine to produce a finished product.
Software Control: Control of electro-mechanical actuator system, where the actuator function is defined by software and can be readily modified or reprogrammed.
Multi-Axis Mechanisms: Mechanisms which have complex motion requirements by multiple independent drives, in place of the traditional centralized cam-actuated mechanisms.