小卫星测试数据在线时间序列判读方法

小卫星自动化测试是提高测试效率、缩短研制周期、降低研制成本和保证可靠性的重要手段,数据判读是测试系统的重要组成部分.传统测试数据判读多依赖于测试人员的手工比对,效率较低,难以适应于实时性较高的小卫星系统测试.针对这一问题,设计了小卫星测试系统方案和数据在线判读模型,建立了测试数据的时间序列模型,在此基础上完成数据预测算法设计和判读方法设计.最后利用飞轮和陀螺的测试数据对该方法进行了工程验证.结果表明,设计的数据预测算法和判读方法兼顾测试数据自身特点和实时性要求,能在较小的先验数据集上全面地检测出数据异常,从而显著提高测试的全面性和准确性.

An on-line data interpretation method in a micro-satellite ground test system based on a time series analysis

Automatic testing of micro-satellites is an important means of improving test efficiency,shortening the development cycle,reducing development costs,and ensuring the reliability.Data interpretation is becoming one of the most critical components of test systems.Traditional data interpretation methods depended on testing operators and had a low efficiency.Therefore,it was difficult to adapt to test systems with higher real-time requirements.To solve this problem,a micro-satellite test system and online data interpretation models were designed,the time series model of test data was established,and the data prediction algorithm and interpretation method were obtained based on those models.Finally,taking advantage of the flywheel and gyroscope test data,the methods proposed were verified.The results show that the data prediction algorithm and interpretation methods take into account the characteristics of test data and real-time requirements and can detect data abnormalities comprehensively on a smaller priority data set,thereby significantly improving the comprehensiveness and accuracy of the tests.