考虑随机不确定性的常态仿真模型确认试验设计

模型确认试验是一种新的试验,其目的在于度量仿真模型的可信度.为了得到低成本、高可信度的模型确认试验方案,提出一种随机不确定性模型确认试验设计方法.首先,基于面积确认度量指标提出一种新的无量纲的模型确认度量指标(面积确认度量指标因子),并且在其基础上发展了基于专家系统的仿真模型准确性定性评判准则;然后,建立随机不确定性模型确认试验优化设计模型,提出该优化模型的求解方法;最后,通过两个数值算例对提出的模型确认试验设计方法进行验证.结果表明,小样本情况下,试验方案的随机性会影响模型评判结果的可信度;面积度量指标因子随试验样本数量的增加而收敛;随机不确定性模型确认试验设计方法能够避免试验方案对模型确认结果的影响.

Model validation experiment design of time-invariant model with random uncertainty

Model validation experiment is a new type of experiments, of which the goal is to determine the credibility of the simulation model. In order to obtain low cost and high credibility of a model validation experiment, a methodology of model validation experiment design considering random uncertainty is presented. Firstly, by extending the concept of area metrics, a dimensionless model validation metric (area metric factor) is proposed, and a criterion for qualitative assessment of accuracy of simulation models based on expert systems is developed. Then, an optimization model for model validation experiment design considering random uncertainty is constructed. Meanwhile, an optimization method is developed to solve the optimization model. Finally, two simulation examples are used to illustrate the proposed methods. The simulation results show that the randomness of experiment design affects the reliability of the model evaluation results in the case of small samples. The area metric factor converges with the increase of the number of experimental observations. The proposed method for model validation experiment design can avoid the effect of testing program on the result of model validation.