自适应指数平滑法预测遥测数据实时趋势研究

针对目前航天领域在对运载火箭遥测数据进行监控过程中，人工监测反应速度慢、工作量比较大这一问题，提出了对遥测数据进行趋势预测的思想，并采用自适应指数平滑法进行趋势预测研究。仿真实验证明：该方法预测精度较高，跟踪速度较快，实时性较好，能够实现对遥测数据的实时趋势预测。应用该预测方法可提前预测潜在的故障趋势，为指挥员正确决策提供理论支持。

Research on Exponent-smoothing Model with Variable Parameters Applied in Real-time Trend Prediction of Telemetric Data

Aiming at slow reactivity and great workload of human based monitoring test during the monitoring process of telemetric data from launch vehicles in the spaceflight experimentation, a tentative has been put forward as real-time trend prediction of telemetric data, and an exponent-smoothing model with variable parameters has been introduced to implement such research. Simulation proves that such model can promise a higher precision,a faster track and has been more real-time in prediction, which can achieve real-time trend prediction of telemetric data. Moreover, it can forecast the fault tendency in advance, which can provide a scientific basis for commander in right decision-making.