Behavior Analysis and Testing Software (BATS)

摘要

Analysis of model behavior is mainly conducted in a pattern-based manner in system dynamics (SD) methodology. In pattern-based evaluation of model outputs, similarity of the overall behavior pattern (e.g. S-shaped-growth, oscillations) and of specific pattern characteristics (e.g. inflection points, periods, amplitudes) are more important than point-by-pointsimilaritymeasuressuch as sum-of-squared errors. Although some output analysis tools/software that address this special pattern focus are available, they lack usability and are fragmented. In this study, new standalone analysis software, namely Behavior Analysis and Testing Software (BATS), is developed. It integrates a pattern classification algorithm and a set of statistical methods for analysis of steady-state behaviors. Apart from enabling comparison of behaviors with these algorithms/methods, BATS includes structured processes that enable user to conduct automated hypothesis testing, behavior space exploration, and sensitivity analysis. In its current state, BATS can seamlessly communicate with SD modeling software (Vensim) and other common data sources. This study provides illustrative examples of how BATS can assist the modeler and/or analyst in various phases of modeling; indirect structure testing, output evaluation, sensitivity analysis, policy analysis. Considering its pattern-orientation, user-friendly interface, and communication with modeling software BATS can be an important contribution to the analysis toolset of SD methodology.