Multi-source Test Data Fusion and Evaluation Based on Improved p-Bayesian Method

An improved multi-source data fusion evaluation method based on p-Bayes method was proposed in the paper. In view of the deficiency of constructing the inheritance factor subjectively, this paper proposed to calculate the inheritance factor based on the mean square error(MSE).Combined with experience and inheritance factors, test data fusion from different sources such as multi-stage, different regions, modified products and similar products was realized, which effectively improved the evaluation accuracy of success or failure type indicators under small sample flying test, and to reduce the test risk.