A Simulation Model Validation Method Based on Functional Data Analysis.

Contrapose the disaccord between simulation data and real data in the process of simulation model validation, a simulation model validation method based on functional data analysis is proposed. This method is a kind of dynamic data analysis method, the simulation data and real data can be regarded as the sample data of two random processes, so the data also can be regarded as infinite dimension vector. Fit the data for function, so the data can be tackled in function form, the correlation analysis of the two random process of the simulation model and real system can be processed through the function data sample, the simulation model validation can be realized. This method proposes a new metrics to evaluate the credibility of the simulation model without the requirement of the time series consistency of simulation data and real data. An example is provided to prove the feasibility of the method.