Research on Temperature Effect Separation Method of Bridge Deflection Monitoring Data Based on Complementary Ensemble Empirical Mode Decomposition and Fast Independent Component Analysis

Abstract Bridge deflection monitoring signals encompass the responses of vehicle loads, temperature loads, and shrinkage creep of concrete, and the temperature effect of bridge deflection can mask the signal characteristics reflecting the bridge operational status, thereby influencing the bridge status assessment. This paper proposes a temperature effect separation method for bridge deflection monitoring data, based on complementary ensemble empirical mode decomposition (CEEMD) and fast independent component analysis (FastICA). First, the single‐channel bridge deflection signal is decomposed into a series of intrinsic mode functions (IMFs) by using CEEMD; next, the energy entropy increment‐based discriminant method is used for identifying and eliminating the false IMF components, and those IMF components with large energy entropy increments compose the input signal; finally, the input signal is separated via FastICA. Based on the separation results and periodic pattern of temperature change, the temperature difference effect of bridge deflection monitoring data is recognized and separated. The correlation coefficients between the simulation analysis results of the separated signals and the source signal are above 0.95. Furthermore, in‐service bridge deflection monitoring data analysis results indicate that the correlation coefficient between the source signal and the separation signal, as well as the symmetrical measuring point signals, is greater than 0.95, which demonstrates the applicability and practicality of the method in engineering applications.