CAJ | 학술논문

为了评估小样本情况下光纤陀螺的可靠性,开展基于多源信息融合技术的 Bayes(贝叶斯)评估方法研究,对样本的先验信息进行分析,采用随机加权法通过关键器件的信息来确定先验分布,结合小样本试验数据所确立的似然函数,在不降低可信度的情况下对产品的可靠性进行评估.由仿真结果可知,在相同置信度下,Bayes评估方法将可信区间长度缩短为传统方法的1/16.该方法提高了估计结果精度对解决小样本下的可靠性评估问题具有重要的理论意义和参考价值.
위료평고소양본정황하광섬타라적가고성,개전기우다원신식융합기술적 Bayes(패협사)평고방법연구,대양본적선험신식진행분석,채용수궤가권법통과관건기건적신식래학정선험분포,결합소양본시험수거소학립적사연함수,재불강저가신도적정황하대산품적가고성진행평고.유방진결과가지,재상동치신도하,Bayes평고방법장가신구간장도축단위전통방법적1/16.해방법제고료고계결과정도대해결소양본하적가고성평고문제구유중요적이론의의화삼고개치.
To estimate the reliability of fiber optic gyroscope on the small sample testing data, the Bayes (bayesian) estimation methods were studied based on the information fusion technology of sources. The priori information of stylebooks was analyzed, and the prior distribution was determined by the key apparatus’s information through the stochastic weighted method, and the likelihood function was combined based on the small sample testing data, to estimate the reliability of product without lowering the credibility. The simulation results show that the length of the confidence interval of Bayes estimation methods is shortened to 1/16 compared with that of traditional methods in the same confidence levels. The accuracy of the estimated results is improved by the Bayes estimation methods and it has important theoretical significance and reference value for solving the reliability assessment problem with small sample.