<p>IEC&#xa0;61499 is often presented as a more modular and distributed alternative to IEC&#xa0;61131-3, but empirical comparisons usually conflate the effect of the <i>standard</i> with that of the underlying <i>design</i>, pitting a modular IEC&#xa0;61499 application against a monolithic IEC&#xa0;61131-3 baseline. We introduce a controlled, artifact-level method that separates the two. For two reference problems—a multi-floor elevator and an ISA-88 batch dosing unit—we build three functionally equivalent implementations: a monolithic IEC&#xa0;61131-3 program (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(I_L\)</EquationSource> </InlineEquation>), a modular IEC&#xa0;61131-3 design (<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(I_M\)</EquationSource> </InlineEquation>) in the same engineering tool, and an IEC&#xa0;61499 design (<InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(I_E\)</EquationSource> </InlineEquation>). The step <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(I_L\rightarrow I_M\)</EquationSource> </InlineEquation> isolates modular design; <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(I_M\rightarrow I_E\)</EquationSource> </InlineEquation> evaluates the additional effect of adopting IEC&#xa0;61499. Each version undergoes two classes of change—a parametric extension and a cross-cutting modification—measured by a reproducible canonical-diff protocol, without physical hardware. Shared-state coupling between functional units drops from 6 and 2 variables in <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(I_L\)</EquationSource> </InlineEquation> to 0 in both modular designs, yet change impact differs by at most one unit between <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(I_M\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\(I_E\)</EquationSource> </InlineEquation>, and the monolith is least affected in two of the four changes. For the studied cases, source-level maintainability is driven mainly by modular design rather than by the standard alone, and depends on the shape of the change. The contribution is a reproducible method for separating design effects from standard effects, not a claim of IEC&#xa0;61499 superiority.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Design discipline or standard effect? a controlled artifact-level comparison of IEC 61131-3 and IEC 61499 for control software maintainability

  • Zakaria Bencherki,
  • Khaled Benfriha,
  • Vincent Meyrueis,
  • Alexandre Sava

摘要

IEC 61499 is often presented as a more modular and distributed alternative to IEC 61131-3, but empirical comparisons usually conflate the effect of the standard with that of the underlying design, pitting a modular IEC 61499 application against a monolithic IEC 61131-3 baseline. We introduce a controlled, artifact-level method that separates the two. For two reference problems—a multi-floor elevator and an ISA-88 batch dosing unit—we build three functionally equivalent implementations: a monolithic IEC 61131-3 program ( \(I_L\) ), a modular IEC 61131-3 design ( \(I_M\) ) in the same engineering tool, and an IEC 61499 design ( \(I_E\) ). The step \(I_L\rightarrow I_M\) isolates modular design; \(I_M\rightarrow I_E\) evaluates the additional effect of adopting IEC 61499. Each version undergoes two classes of change—a parametric extension and a cross-cutting modification—measured by a reproducible canonical-diff protocol, without physical hardware. Shared-state coupling between functional units drops from 6 and 2 variables in \(I_L\) to 0 in both modular designs, yet change impact differs by at most one unit between \(I_M\) and \(I_E\) , and the monolith is least affected in two of the four changes. For the studied cases, source-level maintainability is driven mainly by modular design rather than by the standard alone, and depends on the shape of the change. The contribution is a reproducible method for separating design effects from standard effects, not a claim of IEC 61499 superiority.