实时领域交互行为模型转换方法及非功能语义规约研究

In the process of real-time industrial control software design with model-driven approach, how to convert the design model with both control functions and non-functional Statutes into an executable model accurately is the key issue. To solve this question, the specific real-time behavioral collaboration model conversion and the non-functional semantic restrict must be carried out successfully. Current researches focus on the design and conversion of the model framework, but not on the studies for specific behavior modeling and conversion, It will not support multi-point collaborative behavior modeling and conversion in the field of real-time control and those research can not meet the non-functional semantics consistency constraints in real-time field. In this study those behavior ontologies of real-time domain are identified and the interpretation are presented to built a real-time ontology knowledge base.As a result an ontology framework for multi-task interactive based on real-time semantic model is established .real-time semantic model is constructed as an intermediate model and so the model transformation process from the source model to the target model is devided into two steps :source model -semantic model transformation and semantic model - target model transformation.The semantic mapping of the behavior, structure, type, rules and timing in both functional behaviors and non functional behaviors are completed respectively. This research emphasizes on the modeling and model transformation method for real-time multi-point interactive behavior and the constraints of non-functional semantics in the conversion process. thus the aim of which is to provide theoretical and experimental foundations for the collaboration behavior modeling, conversion and semantic restraint.

在使用模型驱动方法进行实时工业控制软件设计的过程中，将带有控制功能与非功能性规约的设计模型完整准确地转换成可执行的实现模型是关键问题，其中具体的实时行为交互模型的转换及非功能性语义约束是难点。现有研究集中于模型的框架设计及转换，而较少研究具体行为建模及转换，因而无法支持实时控制领域的多点协作交互行为建模和转换，更难以满足实时领域内非功能性语义的一致性约束要求。研究拟通过对实时领域行为本体的识别和演绎建立实时本体知识体系，在此基础上给出实时多任务交互行为本体应用框架，构建实时语义模型作为中间模型，把模型转换的过程分解为源模型-语义模型转换及语义模型-目标模型转换，分别完成行为、结构、类型、规则和时序等功能性和非功能性语义映射。研究重点在于实时多点交互行为的建模及模型转换方法和时序等重要非功能性语义在转换过程中的约束问题，进而为交互行为模型的描述、转换及语义约束等普遍问题提供理论与实验基础。

实时领域中多任务并行和交互是系统建模和实现的关键技术，由于时序、资源等非功能性规约众多，使得在实时系统模型的设计和转换过程中，如何完整准确地表现这些非功能功能规约并进行模型转换尤为重要。我们以实时嵌入式系统为对象，围绕这一问题展开了深入研究。首先我们研究了实时系统模型构建过程中的关键问题，给出了实时嵌入式系统中图形元模型的表述方法，为实时非功能性语义提供了合适的载体；其次对实时行为建模过程中的时序和并发控制的约束展开了分析和研究，获得了实现非功能语义规约的条件集合，并提出了确定实时语义的形式化方法；对模型转换过程中的语义规约方法进行了研究，构建了基于本体的实时语义库，给出了融合实时语义库的交互行为模型转换方法，并以上述理论研究为基础实现了实时系统建模平台和转换平台，提供了以实时交互行为作为基本元模型的建模环境和非功能语义的可编辑环境。同时通过转炉炼钢实时终点判断系统、隧道实时安全监测系统和人眼特征点实时跟踪方法等典型的实时系统设计进行了相关研究内容的验证。