In-orbit Operational Pattern Monitoring Algorithms Based on LightGBM for Hard X-ray Modulation Telescope Satellite
-
摘要: HXMT卫星的空间硬X射线巡天和定点观测计划切换频繁，需要对卫星有效载荷在轨状态进行实时监测和判别.目前采用的是地面监测人员根据总结的规则进行人工监测的方式，虽然执行方便，可解释性强，但人力消耗较大，且对规则之外的情况无法灵活处理.本文利用HXMT卫星的实时遥测数据，提出一种基于LightGBM机器学习模型的在轨运行模式监测算法，将监测工作规约为多分类问题，并构建判别模型，对卫星在轨运行模式进行判断.在保障判别准确率的前提下，算法模型构建迅速，具有很高的实用性.基于真实遥测数据的试验表明，模型的判别准确率达到99.9%，满足在轨运行模式监测要求，可为HXMT卫星的运行监控任务提供参考依据.Abstract: The frequent switching of space hard X-ray sky survey and fixed-point observation schemes of HXMT (Hard X-ray Modulation Telescope) satellite requires real-time monitoring and identification for the satellite payloads in-orbit status. Manual monitoring according to the rules summarized by experts are used at present. Although the manual monitoring method is easy to execute and explicable, it consumes a lot of manpower and can not deal with the situation outside the rules flexibly. According to the real-time telemetry data of HXMT satellite, an in-orbit operation mode monitoring algorithm based on LightGBM is proposed in this paper. The in-orbit operational pattern monitoring is reduced to a multi-classification problem, and a discriminant model is constructed to efficiently judge the in-orbit operation mode of the satellite. On the premise of ensuring the accuracy of discrimination, the algorithm model is constructed very quickly, which liberates the monitoring personnel from the heavy rule judgment work and has high practicability. The experiments based on the real telemetry data show that the accuracy rate of the model is 99.9%, which can meet the requirement of in-orbit operation mode monitoring, and can provide references for HXMT satellite operation monitoring task.
-
Key words:
- Operational control pattern /
- Real-time monitoring /
- LightGBM /
- HXMT
-
[1] CHEN Wei, HU Guangrui, WANG Yaping. Knowledge acquisition in an expert system for fault diagnosis of certain spacecraft[J]. J. Shanghai Jiaotong Univ., 2000, 34(6):845-847(陈玮, 胡光锐, 汪亚平.飞行器故障诊断专家系统中的知识获取机制[J]. 上海交通大学学报, 2000, 34(6):845-847) [2] KRIZHEVSKY A, SUTSKEVER I, HINTON G. Imagenet classification with deep convolutional neural networks[C]//Neural Information Processing Systems. Lake Tahoe:NIPS, 2012:1106-1114 [3] XIE Lun, LIU Fan, GONG Xiao, et al. Spacecraft fault diagnosis system based on the hybrid intelligence[J]. Inf. Control, 2010, 39(1):106-113(解仑, 刘帆, 巩潇, 等. 基于混合智能的航天器故障诊断系统[J]. 信息与控制, 2010, 39(1):106-113) [4] BALDI P, BLANKE M, CASTALDI P, et al. Combined geometric and neural network approach to generic fault diagnosis in satellite reaction wheels[J]. IFAC, 2015, 48(21):194-199 [5] KE G, MENG Q, FINLEY T, et al. LightGBM:a highly efficient gradient boosting decision tree[C]//Neural Information Processing Systems. Long Beach:NIPS, 2017:3146-3154 [6] FRIEDMAN J H. Stochastic gradient boosting[J]. Comput. Stat. Data An., 2002, 38(4):367-378 [7] CHEN T, GUESTRIN C. XGBoost:a scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco:ACM, 2016:785-794 -
点击查看大图
计量
- 文章访问数: 680
- HTML全文浏览量: 42
- PDF下载量: 67
- 被引次数: 0
下载:
