| 基于证据合成与贝叶斯网络推理的航天器发射风险评估模型 |
| |
| 引用本文： | 董学军,杜建洲.基于证据合成与贝叶斯网络推理的航天器发射风险评估模型[J].系统工程理论与实践,2019,39(8):2170-2178. |
| |
| 作者姓名： | 董学军 杜建洲 |
| |
| 作者单位： | 1. 酒泉卫星发射中心 技术部, 酒泉 732750;2. 酒泉卫星发射中心 发测站, 酒泉 732750 |
| |
| 基金项目： | 国家自然科学基金（70971131，71071156） |
| |
| 摘 要： | 为解决航天器发射风险评估中统计数据较少问题,基于证据合成和Bayesian网络推理,构建了航天器发射风险评估模型.模型将灾难征兆信息转化成专家意见,然后使用证据合成技术对专家意见进行融合,并形成风险评估结论.模型针对专家中少数不同意见,使用网络推理技术将相关的灾难征兆转化成灾难发生可能性的判断.最后,通过STS-51L实例分析验证了模型的可行性.
|
| 关 键 词： | 航天器发射 证据合成 Bayesian网络推理 风险评估模型 灾难征兆 |
| 收稿时间： | 2018-04-11 |
A risk assessment model based on evidence combination and Bayesian network reasoning |
| |
| DONG Xuejun,DU Jianzhou.A risk assessment model based on evidence combination and Bayesian network reasoning[J].Systems Engineering —Theory & Practice,2019,39(8):2170-2178. |
| |
| Authors: | DONG Xuejun DU Jianzhou |
| |
| Institution: | 1. Technology Department, Jiuquan Satellite Launch Center, Jiuquan 732750, China;2. Launch Test Station, Jiuquan Satellite Launch Center, Jiuquan 732750, China |
| |
| Abstract: | A risk assessment model based on evidence combination and Bayesian network reasoning was built in order to solve the problem of less statistical data in spacecraft launch. The model converts the signs of disaster into the opinions of multiple experts. Then, the model integrates those expert opinions and forms risk assessment conclusions by using evidence combination techniques. In addition, based on a few expert opinions, the model uses Bayesian network reasoning to convert disaster signs into judgments of the possibility of disasters. Finally, the feasibility of the model was verified by the STS-51L case study. |
| |
| Keywords: | spacecraft launch evidence combination Bayesian network reasoning risk assessment model disaster sign |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《系统工程理论与实践》浏览原始摘要信息 |
| 点击此处可从《系统工程理论与实践》下载免费的PDF全文 |
