改进小波阈值的遥测数据平滑算法

Construction Method of Wavelet Function for Improving Threshold of Telemetry Data

为解决常用遥测数据平滑方法收敛速度慢、拟合精度差和适用性弱等缺点,本文提出了改进阈值函数的小波方法。利用Matlab对遥测到的飞行器空速、发动机转速、机舱温度和湿度数据进行算法去噪仿真实验。试验结果表明:本文方法改善了传统小波软、硬阈值函数存在的数据跳变和过度平滑现象,能够有效进行多类遥测参数噪声处理,降低了数据测量误差,且曲线拟合相似度均在80%以上,使测量值更接近真实,能够更加真实的反映出原始测量数据的变化趋势,具有较强的适用性和平滑性。

The common functions for telemetry data have the disadvantages of slow convergence speed,poor fitting accuracy and weak applicability.In order to solve these problems,a wavelet method of improved threshold functions is proposed.The Matlab simulation test was conducted to denoise the remotely sensed aircraft airspeed,engine speed,engine room temperature and humidity data.The experimental results show that the proposed method can solve the problems of the data jump and excessive smoothing existing in the traditional wavelet soft and hard threshold function,that it can effectively deal with the noise of various telemetry parameters and reduce measurement errors and that the similarity of curve fitting is more than 80%,making the measured values closer to reality.The measured curve can provide a true reflection of the change trend of the original values.