基于时态逻辑的管制员响应执行差错建模与分析方法研究

发布时间：2018-08-28 11:57

【摘要】：管制员响应执行是空中交通管制的重要环节,也是民航空管运行中易引发不安全事件的关键过程。为了对涉及管制员响应执行过程的不安全事件进行深入分析,提高事件分析的规范性和自动化程度,避免过分利用自然语言描述的缺点,提出了基于时态逻辑的形式化推理方法。首先,总结分析了时态逻辑建模语言(Temporal Trace Language,TTL)的相关特征,概括性地介绍了Agent的理论基础和Multi-Agent的系统构建特征,并阐述了利用TTL语言对Multi-Agent系统进行形式化建模的方法。借鉴Wickens提出的人类信息加工模型,加入管制运行的相关要素,并根据场景角色的交互特征建立响应执行Multi-Agent系统模型,描述交互过程中Agent之间的关系。其次,针对建立的响应执行交互模型,从时态逻辑的角度出发构建响应执行场景形式化模型与仿真模型。首先,对场景中涉及的相关静态概念进行收集,然后抽象出属性要素之间的动态关系,接着根据TTL的规范将概念、关系以及属性形式化,最后,介绍了仿真建模语言(Language and Environment for Analysis of Dynamics by SimulaTi On,LEADSTO)语义、语法规则以及动态属性关系构建方法,并阐述了建模与仿真的具体步骤,包括模型概念化、形式化、仿真以及评估分析。以一种典型的跑道侵入事件中管制员响应执行场景为例,建立其时态逻辑模型,并构建场景需要满足的规则属性,描述规则属性之间的层级关系。选取此场景的不安全事件案例,对事件进行形式化建模、仿真建模以及逻辑推理分析,自动推理出事件轨迹中不满足规则属性的动态属性,从而找出诱发事件发生的危险要素;再变更仿真案例中不同的属性因子组合,进行仿真对比分析,抽象出诱发差错和事件发生的多种因素,探索出阻止差错和事件发生的有效措施,进而提出有效地缓解措施。结果表明,该方法能实现逻辑化、动态性的仿真,并通过形式化推理识别出事件中的危险要素,克服了单纯利用自然语言进行危险分析的缺点,为后续此类事件的分析提供了科学的方法指导。
[Abstract]:Controller response is an important part of air traffic control, and it is also a key process of causing unsafe events in the operation of civil aviation air traffic control. In order to conduct an in-depth analysis of unsafe events involving controllers' response to the execution process, to improve the normative and automated level of event analysis and to avoid overusing the shortcomings of natural language descriptions, A formal reasoning method based on temporal logic is proposed. Firstly, this paper summarizes and analyzes the relevant features of temporal logic modeling language (Temporal Trace Language,TTL), introduces the theoretical basis of Agent and the characteristics of Multi-Agent system construction, and expounds the formal modeling method of Multi-Agent system using TTL language. Based on the human information processing model proposed by Wickens, this paper adds the relevant elements of control operation, and establishes the response execution Multi-Agent system model according to the interaction characteristics of scene roles, and describes the relationship between Agent in the process of interaction. Secondly, based on the response execution interaction model, the formal model and simulation model of response execution scenario are constructed from the point of view of temporal logic. First, the static concepts involved in the scene are collected, then the dynamic relations between attribute elements are abstracted, then the concepts, relationships and attributes are formalized according to the TTL specification. Finally, This paper introduces the semantics, syntax rules and dynamic attribute relation construction method of simulation modeling language (Language and Environment for Analysis of Dynamics by SimulaTi On,LEADSTO), and expounds the concrete steps of modeling and simulation, including model conceptualization, formalization, simulation and evaluation analysis. Taking a typical runway intrusion event as an example, the temporal logic model of controller response execution scene is established, and the rule attributes that need to be satisfied in the scene are constructed to describe the hierarchical relationship between the rules attributes. This paper selects the unsafe event case of this scene, carries on the formalization modeling, the simulation modeling and the logical reasoning analysis to the event, automatically inferences the dynamic attribute which does not satisfy the rule attribute in the event track, thus finds out the dangerous factor which induces the event to occur; Then changing the different attribute factor combination in the simulation case, comparing and analyzing the simulation, abstracting various factors that induce errors and events, exploring effective measures to prevent errors and events, and then putting forward effective mitigation measures. The results show that the method can realize logical and dynamic simulation, and identify the dangerous elements in the event by formal reasoning, which overcomes the disadvantage of using natural language to carry out hazard analysis. It provides scientific guidance for the analysis of this kind of events.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V355.1

【相似文献】