Noise Cancellation Method Based on TVF-EMD with Bayesian Parameter Optimization

To separate the noise and important signal features of indoor carbon dioxide (CO2) concentration signal, we proposed a cancellation method, based on time-varying, filtering-based empirical mode decomposition (TVF-EMD) with Bayesian optimization (BO). The adaptive parameters TVF-EMD, that is, bandwidth threshold ξ B-spline order n, were determined by BO algorithm, correlation coefficient for kurtosis index (CCKur) constituted objective function. Initially, function CCKur was introduced to systematically identify anomalous signals while preserving feature extraction between modes input signal. Subsequently, model TVF-EMD algorithm employed, along Hurst exponent, extract sensitive mode. An examination indices decomposed intrinsic functions (IMFs), namely CC, Kur, MI, EE, EEMI, CCKur, revealed synthetic measurement fitness reasonable effective. method exhibited better performance, compared default values, EMD, moving average exponential smoothing method.