Validation Methods Of Simulation Model Based On Static And Dynamic Consistency Analysis

With the development of technologies and human progress, simulation modelshave become increasingly complicated and people gradually pay more attention tothe credibility of simulation models. From the aspect of purpose, the start ofVV&A work made a solid foundation for credibility of simulation models.Especially, the validation process of simulation models is just the procedure toevaluate it. Based on the circumstances above, it made a deep research onvalidation methods.Firstly, an overview of the background and status of simulation modelvalidation was summarized in detail. Also it described the theoretical basis formodel validation.Secondly, for the validation of the static performance, the data pre-processingmethods were introduced. Especially, focused on parameter estimation, hypothesistesting and estimation based on Bayesian theory, it summarized the scope andapplicable conditions of the above methods.Thirdly, for the validation of the dynamic performance, the data pre-processing methods were introduced. On one hand, focused on TIC coefficientmethod, grey relation analysis, maximum entropy spectral analysis and regressionanalysis method, it summarized the scope and applicable conditions of the abovemethods. On the other hand, the modified algorithm of grey relation analysis wasput forward to make it suitable for simulation model validation. The comparisonwas made between Yule-Walker method and Burg method.At last, based on the research of validation methods, the tool for validationassistance is completed via the platform of VC and Matlab. The tool providedguidance for customers and integrated numerous validation data-preprocessingmethods and data-analysis methods, greatly reducing costs and improving theefficiency of model validation.In this paper, it not only made a deep research on validation methods, but alsoput them into practice to evaluate the credibility of simulation models and hadachieved perfect results. The work of this paper provided a guidance for futureresearch on validation methods.