在小样本量下,经典测试性评估方法风险性大,为解决这个问题,充分利用先验信息,分别采用最大熵法和Chebyshev多项式展开方法确定出先验分布,然后运用Bayes理论对发射装置的测试性进行评估,并与经典评估方法作比较,结果表明经典评估方法过于保守,Chebyshev多项式展开方法存在评估结果不稳定的问题,而最大熵法确定出的先验分布偏差较小,评估结果最好。

The risk of classical testability assessment method is high under the condition of small sample. To avoid this problem, the prior information was fully utilized and the maximum entropy method and Chebyshev polynomial expansion method was used to determine the prior distribution. Then the Bayes theory was applied to evaluate the testability of the launcher and compared with the classical method. By contrast, the results evaluated by the classical method are too conservative, and the ones evaluated by the Chebyshev polynomial expansion method are not stable, but the maximum entropy method is the best and accurate.