arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1503.04917cs.SEcs.LO
收藏

基于模糊逻辑的交互系统形式化描述

A Formal Approach based on Fuzzy Logic for the Specification of Component-Based Interactive Systems

Vasileios Koutsoumpas

摘要:

形式化方法被广泛认为是分布式交互系统的规范、仿真、开发和验证的一种强有力的工程方法。然而,大多数形式化方法依赖于二值逻辑,因此受限于该逻辑的公理:规范有效或无效,组件行为可实现或不可实现,安全属性保持或被违反,系统可用或不可用。特别是当问题域具有不确定性,不精确性和不确定性时,这些方法的应用成为一项具有挑战性的任务。为了克服形式化方法的严格工作方式所带来的局限性,本文的主要目的是通过使用模糊逻辑来放松形式化规范的布尔概念。本方法是基于焦点理论,一个基于模型的和严格的形式化方法,基于组件的交互式系统。这项工作的贡献是双重的:i)我们引入了规范技术的基础上,模糊逻辑,可用于顶部的焦点开发正式规范的定性方式; ii)我们部分扩展焦点理论的模糊,它允许规范的模糊组件和模糊的相互作用。虽然前者提供了一种方法来近似I/O行为下的不精确性,后者能够捕捉到一个更定量的规格属性,如可实现性。

英文摘要:

Formal methods are widely recognized as a powerful engineering method for the specification, simulation, development, and verification of distributed interactive systems. However, most formal methods rely on a two-valued logic, and are therefore limited to the axioms of that logic: a specification is valid or invalid, component behavior is realizable or not, safety properties hold or are violated, systems are available or unavailable. Especially when the problem domain entails uncertainty, impreciseness, and vagueness, the appliance of such methods becomes a challenging task. In order to overcome the limitations resulting from the strict modus operandi of formal methods, the main objective of this work is to relax the boolean notion of formal specifications by using fuzzy logic. The present approach is based on Focus theory, a model-based and strictly formal method for componentbased interactive systems. The contribution of this work is twofold: i) we introduce a specification technique based on fuzzy logic which can be used on top of Focus to develop formal specifications in a qualitative fashion; ii) we partially extend Focus theory to a fuzzy one which allows the specification of fuzzy components and fuzzy interactions. While the former provides a methodology for approximating I/O behaviors under imprecision, the latter enables to capture a more quantitative view of specification properties such as realizability.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑