基于贝叶斯动态模型的某器件性能预测

在导弹武器系统当中，及时准确的故障预报对提高导弹的安全性具有极其重要的意义。本文根据导弹惯性器件故障预报系统的设计要求，利用贝叶斯动态线性模型对导弹某惯性器件的性能进行了预测研究。贝叶斯预测是利用历史信息和从样本获得的信息来获取后验分布的一种预测方法，该方法不需要平稳性的假设，而且充分利用了已有的信息，可以用来对电子设备的性能进行预测。实例研究表明，该方法具有比较好的预测效果。

Bayesian dynamic model based performance prediction for a certain device

Bayesian prediction uses historical data and those data acquired from the samples for obtaining the posterior probability. This method need not the hypothesis of the stationary performance, and can make full use of the information we have got. This method can be used for predicting the performance variation of electrical devices and for evaluating the reliability of it in time. The result of study shows that this method of the Bayesian prediction has very good effect.