A graph-theoretic approach to integrated process and control system synthesis

Hangos, Katalin M.; Friedler, Ferenc; Varga, J.B.; Fan, L. T.

doi:10.1016/s1474-6670(17)47965-1

Cited by 2 publications

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These control algorithms can also be designed by various graphbased methods. For example, [12] is based on a CP-graph model. Usually the control algorithm itself is already available as a dataflow-like graph.…”

Section: Special Requirements Of Dcsmentioning

confidence: 99%

Application of a System-Level Synthesis Tool in Industrial Process Control Design

Arató¹,

Nagy²,

Rácz³

2019

ACTA POLYTECH HUNG

View full text Add to dashboard Cite

Complex digital systems usually demand some kind of a multiprocessing architecture. The requirements to be fulfilled (energy and communication efficiency, speed, pipelining, parallelism, the number of component processors, cost, etc.) and their definable priority order may cause conflicts. Therefore, the best choice of the component processors (beside general-purpose CPUs also DSPs, GPUs, FPGAs and other custom hardware) is very important. Such resulting architectures are called heterogeneous multiprocessing architectures (HMA). The system-level synthesis (SLS) methodology can be applied beneficially in designing the HMAs. In this way, the design procedure can get rid of the most intuitive trial and error steps including also the partly reusing of existing structures. Therefore, the SLS methods help to optimize HMAs by reducing the intuitive steps. A high degree of similarity can be observed between HMAs and modern distributed industrial process control systems (DCS). This paper illustrates the procedure of adapting and applying an SLS tool in redesigning of an existing DCS as a benchmark for analyzing, evaluating and comparing the results. Through this adaptation, all such SLS functions become executable on the traditional and standardized documentation form of a DCS.

show abstract

Section: Special Requirements Of Dcsmentioning

confidence: 99%