Konferensartikel

The simulation model validation and calibration (SMVC) is a complicated work, including uncertainty description, many simulation experiments execution and complex data analysis etc. Moreover, there are many uncertainty factors such as model form assumptions and solution approximations, random variability of model inputs, etc. need to be considered when requiring a precise model. For assisting the SMVC effectively, this paper develops a software platform to validate and calibrate the simulation models when some quantities may be affected by uncertainty. First, an unprecedented process model, which includes uncertainty description, simulation experiment design, model validation and model calibration, is presented to explicate the procedures of SMVC under uncertainty. In the process model, many new model validation and calibration algorithms under uncertainty are applied based on our previous work. Second, the design of platform is divided into two parts, which consist of structure design and function design, and the software technique based on strategy pattern is introduced to integrate and maintain the SMVC algorithms. Then this platform is implemented according to its expected uses and key design requirements. Finally, application example of model validation and calibration of a flight vehicle kinematic control system is illustrated how to use the platform.

model validation, model calibration, uncertainty description; software platform