基于Bayesian的测试性虚实一体化试验方案的设计

Design of Testability Virtual-Physical Integrated Test Scheme Based on Bayesian

实物试验与虚拟试验相结合是测试性试验的发展趋势.经典Bayesian样本量确定方法仅能确定单一试验类型样本量,不满足一体化试验方案设计要求,本文针对此问题,引入了设计效应指标,利用测试性虚拟试验可信度和Bayesian最大后验区间(HPD)平均长度构造设计效应,建立虚实样本等效模型,将虚实一体化试验样本转换为实物试验样本,进而给出了不计虚拟试验代价的测试性虚实结合试验方案设计方法.最后针对某装备控制系统舵伺服子系统提出的测试性指标评估要求,利用本文提出的方法设计试验方案,分析了试验方案与虚拟样机可信度的变化关系,验证了本文方法能有效减小实物试验样本量.

It was the tendency of testablity test to combine the physical test with the virtual test. While the classical Bayesian method for sample size determination could apply to the pure physical or virtual test,but not the virtual-physical integrated test. In order to solve this problem,the index of design effect was introduced in this paper,then the index with the credibility of virtual test and the average length of the Bayesian highest posterior density（ HPD） internal was established. After that,the equivalence model of virtual and physical sample was built,based on which the virtual test samples could be converted into physical test samples,and then the design method of virtual-physical integrated testability test scheme was introduced ignoring the cost of virtual test.At last,to evaluate testability indices of a control system＇s subsystem of rudder servo,the test scheme was designed based on the proposed method,then the relationship between the Integrated test scheme and the virtual test credibility was analyzed. The results showed that the proposed method can effectively reduce the sample size of physical test.