基于机器学习建模的航天器健康管理平台研究
2022,30(12):112-118
摘要：近年来，随着深度学习等技术的快速发展和航天器系统数据量的不断增加，新型的机器学习平台凭借其友好的流程化分析框架、丰富的即插即用机器学习工具、分布式的服务等诸多优点，为航天器等领域复杂问题分析处理提出了新思路。在分析了航天器故障预测与健康管理方面存在的难点以及机器学习优势基础上，提出了面向机器学习建模的航天器健康管理平台设计方案与方法，分析了多语言融合的健康管理算法模型构建、基于分布式的健康管理计算服务引擎等关键技术，并以某卫星电源系统太阳电池阵功率预测等案例详细说明平台实际应用情况，验证结果表明研究成果能够为基于机器学习建模的航天器健康管理技术研究与应用提供技术参考，最终提高卫星、空间站等航天器的安全性。
关键词：航天器;健康管理平台 机器学习建模;故障诊断 故障预测
Research of Spacecraft Health Management Platform based-on Machine-Learning Modelling
Abstract：In recent years, with the rapid development of deep learning and the continuous increase in the amount of data in spacecraft systems, the new machine learning platforms, which have friendly process analysis framework, rich plugs and play machine learning tools and distributed services, can provide new ideas of complex problem handling in the fields of spacecraft. On the basis of the analysis of the difficulties in spacecraft prognostic and health management, it proposed the design scheme and method of spacecraft health management platform based on machine learning modeling, and the key technologies such as multi-language fusion health management algorithm model construction and distributed health management computing service engine are analyzed, too. Then, the actual application effect is verified with the satellite solar array power forecasting and other cases. The experiment results show that the research can provide technical reference for the research and application of spacecraft health management technology based on machine learning modelling, and ultimately improve the safety of the spacecraft.
Key words：spacecraft, PHM system, machine learning modelling, diagnostic, prognostic
收稿日期：2022-09-30
基金项目：
