基于知识的仿真模型的验证方法

Knowledge-based Method for Simulation Model Validation

仿真模型的验证研究的是模型的可信度问题。由于模型验证需要关于真实系统,建模理论以及数据分析等多方面的知识,基于知识的智能化方法适合于仿真模型的自动验证。这类方法是建立在三个知识库的基础上的:存储验证相关知识的验证知识库(VKB),用于智能推理和决策的决策知识库(DKB),以及由各类验证算法所构成的验证技术库(VTB)。通过分析验证知识库的验证域,系统选择合理的验证技术和相关的验证过程。在此基础上,系统自动设计仿真实验,获得运行结果,并自动分析数据,执行验证过程,生成验证报告。此外,我们也讨论了验证相关知识的获取过程和细节。

Validation of simulation models is concerned with fidelity of models. As many kinds of knowledge, such as system specific, modeling and simulation, and also data analysis are required within it, knowledge-based intelligent method can be suitable for model validation. This method becomes possible because of three knowledge bases: validation knowledge base (VKB) which is stored with validation relevant knowledge, decision knowledge base (DKB) for intelligent inference and decision, and validation techniques base (VTB) making up of general validation algorithms. By analyzing the context of VKB, validation system decides the rational validation techniques and corresponding validation process. And then it designs experiment, analyses simulation results, executes validation process and generates validation report automatically. In addition, the acquisition of validation relevant knowledge was discussed.