基于自适应卡尔曼滤波的光学监控信号处理

针对光学监控系统(OM S)实际工作环境的复杂性,导致监控信号中存在不稳定量测噪声,采用基于遗忘因子的自适应卡尔曼滤波(AKF)对光学膜厚监控信号进行去噪处理。建立了针对非线性OM S的AKF模型,仿真分析中针对存在突变信噪比(SNR)的输入信号设置相应AKF初始参数使滤波性能达到最优。结果表明自AKF输出监控信号精度明显优于常规卡尔曼滤波;同时利用AKF输出监控信号判读反射率极值点对应膜厚,最终理论膜厚监控误差<2nm。

Adaptive Kalman filter for monitor signal processing of optic thin film

Aiming at the complexity of the optical monitor system(OMS) in actual working environment,which will lead to the existence of the unstable noise in the monitor signal,this paper adopts adaptive Kalman filter(AKF) based on oblivious coefficient to filter the noise in optical monitor signal.The AKF model of nonlinear OMS has been established.The initial AKF parameters have also been set up to optimize the performance of the filter according to the abruptly variational signal-to-noise(SNR) of the OMS Signal.The results indicate that the precision of AKF output monitor signal is distinctly more excellent than general Kalman filter.By using AKF output monitor signal to obtain the reflectivity peak position,the final error of the theoretic film thickness is less than 2nm.