基于LSTM网络的外测级间段数据预测方法

Trajectory Interstage Data Prediction Method Based on LSTM Network

摘要:

针对飞行器飞行试验中外测级间段数据缺失和精度不高的问题，提出了基于长短期记忆(long-short term memory，LSTM）网络的外测级间段数据预测方法。利用遥测视速度数据和外测融合数据建立LSTM网络回归模型，将外测级间段数据作为缺失数据进行预测插值，可将制导工具系统误差以及飞行器初始误差，包括遥外测时间对不准误差，一并利用回归网络表示，从而将遥测视速度数据作为网络输入，得到外测级间段的预测数据。试验数据处理结果证明，基于LSTM网络获得的外测级间段预测数据满足精度要求，所提方法具有实际应用价值。

Abstract:

To solve the problem of loss of data and low precision of trajectory interstage in vehicle flight-test, a trajectory interstage data prediction method based on long-short term memory(LSTM)network was proposed. The LSTM network regression model was established by using the telemetry apparent velocity and the trajectory fusion result data, and the trajectory insterstage data was taken as the loss of data to predict. The guidance instrumentation system error and the initial error, including the remote time error, could be expressed by the regression network. Therefore, the trajectory insterstage data could be predicted by LSTM network which using the telemetry apparent velocity as input. The test data processing results show that the prediction results obtained based on LSTM network meet the accuracy requirements.The proposed method has practical application value.