Mathematical Capture of Human Data for Computer Model Building and Validation (Briefing Charts).

摘要

The model development process created at the TBRL has successfully created models from LVC simulation data, simulated models over time to generate simulated crowd data per condition, performed analysis of observed and simulated data, and compared simulated and observed data against crowd behavioral measures. The process allows for quantitative means for validation of both the mathematical (statistical) and computational models against empirical data. These findings support the claim that more accurate human behavioral models may be derived using laboratory observed data. Furthermore the findings support the necessity of a continuation of the work to explore ways to improve the model development and validation processes.