Summary Interest in the use of multiple processors in real-time, on-line machinery control systems has grown rapidly in recent years. The trend set is toward higher levels of performance and sophistication with low cost mass produced processors. While multiple processor systems can be structured to logically mimic a single processor, a prospect for the future has been the use of multiple interacting processors. Decentralization and distribution of control in a network of processors leads to increased reliability, fault tolerance and flexibility when compared to a single processor system of comparable performance. These benefits are obtained through the application of distributed processing concepts to control system design. This paper describes the essential characteristics of distributed processing systems and relates the attributes of alternative systems to the requirements of industrial control. Choice to be made in the design of such a system are reviewed as are alternative network architectures and methods for synchronizing events in more than one processor. An application example is also discussed.
[1]
D. L. Parnas,et al.
On the criteria to be used in decomposing systems into modules
,
1972,
Software Pioneers.
[2]
E. Douglas Jensen,et al.
The Honeywell Experimental Distributed Processor - an Overview
,
1978,
Computer.
[3]
E. Douglas Jensen,et al.
Computer Interconnection Structures: Taxonomy, Characteristics, and Examples
,
1975,
CSUR.
[4]
Leslie Lamport,et al.
Time, clocks, and the ordering of events in a distributed system
,
1978,
CACM.
[5]
K. R. Dimond,et al.
Distributed Industrial Control Using A Network Of Microcomputers
,
1977
.
[6]
George A. Anderson.
Interconnecting a distributed processor system for avionics
,
1973,
ISCA 1973.
[7]
John G. Bollinger,et al.
Multiple CPU controls for manufacturing systems
,
1977
.