在应用 HP(Hodrick-Prescott)滤波的卫星遥测数据预测方法的基础上进一步研究,提出了基于二分法和琴生不等式判别的自适应 HP 滤波的遥测数据预测方法,有效解决了平滑参数的选择困难.同时,趋势项和波动项预测在应用 HP滤波的方法的基础上有创新和优化,分别采用了线性预测模型和自回归单整移动平均(ARIMA)模型,有利于有效趋势特征提取和模型参数调整范围的优化.应用文章方法对卫星在轨行波管阳极电压数据进行分析,证实了方法的正确性和有效性.此方法可有效提高预测精度,实现遥测数据中长期预测,在卫星故障诊断和预警方面具有工程应用价值.
Prediction Method for Satellite Telemetry Data Using Adaptive HP Filter
Based on the application of HP(Hodrick-Prescott)filter for satellite telemetry data pre-diction,an adaptive HP filter method for telemetry data prediction is proposed based on binary and Jensen inequality discrimination,which effectively solves the difficulty of selecting smoothing parameters.At the same time,the prediction of trend and fluctuation terms is innovative and op-timized based on the application of HP filter method,using linear prediction model and ARIMA(autoregressive integrated moving average)model respectively,which is conducive to effective trend feature extraction and optimization of model parameter adjustment range.The application to analyze the anode voltage data of in-orbit satellite TWT(travelling wave tube)confirms the cor-rectness and effectiveness of the method proposed in the paper.This method can effectively im-prove prediction accuracy,achieve medium and long term prediction of telemetry data,and has engineering application value in satellite fault diagnosis and early warning.
satellite telemetry dataadaptive HP filterARIMA modeldata prediction
国家科技期刊平台
NSTL
万方数据
