promela/spin vs uppaal 嵌入式建模的优缺点
promela/spin vs uppaal pros and cons for embedded modelling
uppaal 和 spin/promela 对嵌入式系统建模的优缺点是什么？
我有点困惑 - 谢谢
一篇过时的论文，但在比较中有一些有用和相关的要点
http://www.brics.dk/RS/96/24/BRICS-RS-96-24.pdf
设计
Considering the design languages, the obvious distinction is the possibility of modelling real–time systems in UPPAAL. In the case study it is shown that interesting bounded liveness properties can be expressed and verified in UPPAAL. Another beneficial feature of UPPAAL is the possibility of committed locations which makes possible a quite natural modelling of the broadcast behaviour needed in the case study. In contrast PROMELA can not apply theatomicity construct on sequences of send- and receive statements as these mightbe blocking
验证
Considering the verification phase, the kind of properties expressible in the property language of UPPAAL are restricted to invariance and possibility properties. Other properties as e.g. the bounded liveness properties of our case study needs to be expressed as separate test automata probing the design. In section 3.4 we present ideas on how to extend the property language and automatically generate the test automata. This is already possible in SPIN for transforming LTL properties to never automata
uppaal 和 spin/promela 对嵌入式系统建模的优缺点是什么？
我有点困惑 - 谢谢
一篇过时的论文，但在比较中有一些有用和相关的要点
http://www.brics.dk/RS/96/24/BRICS-RS-96-24.pdf
设计
Considering the design languages, the obvious distinction is the possibility of modelling real–time systems in UPPAAL. In the case study it is shown that interesting bounded liveness properties can be expressed and verified in UPPAAL. Another beneficial feature of UPPAAL is the possibility of committed locations which makes possible a quite natural modelling of the broadcast behaviour needed in the case study. In contrast PROMELA can not apply theatomicity construct on sequences of send- and receive statements as these mightbe blocking
验证
Considering the verification phase, the kind of properties expressible in the property language of UPPAAL are restricted to invariance and possibility properties. Other properties as e.g. the bounded liveness properties of our case study needs to be expressed as separate test automata probing the design. In section 3.4 we present ideas on how to extend the property language and automatically generate the test automata. This is already possible in SPIN for transforming LTL properties to never automata