An empirical time‐domain trend line‐based bridge signal decomposing algorithm using Savitzky–Golay filter

Abstract

Summary This paper develops a trend line‐based algorithm for signal decomposition in which the adjusted Savitzky–Golay filter is utilized to initiate the decomposition process. In this line, the proposed algorithm determines some special trend lines, mainly composed of the natural frequency of a bridge. An easy‐to‐implement algorithm is then provided to formulate this process and to decompose the given signal into its components in a systematic way. Additionally, a residual signal is generated by the proposed algorithm to store the detected noise and to reconstruct the original signal. To verify the proposed algorithm in the field of bridge health monitoring, a set of numerical and experimental examples are offered in which the proposed algorithm is employed to decompose the signal and provide the constituent components. Moreover, the application of the proposed algorithm in damage localization of the bridge is addressed in the appendix using a simply supported bridge under a moving vehicle. Finally, the bridge example is solved by empirical mode decomposition, as a promising benchmark method, to further illustrate the accuracy of the results and compare them in detail.

Gaussian bare‐bones gradient‐based optimization: Towards mitigating the performance concerns

“…In this part, we apply the GOMGBO algorithm to solve standard engineering problems, and three standard engineering problems are used to verify the effectiveness of GOMGBO for solving practical cases. Practical problems, such as privacy-preserving optimization, 112 image and signal processing, [113][114][115][116][117] learning models, 118 smart grids, 119 and many other complex feature spaces, 120,121 often have many constraints and limits. The three engineering problems involved in this part are including welded beam design problem (WBD), tension-compression spring problem (TCSD), and gear train design problem (GTD).…”

Forecasting Daily Electricity Price by Hybrid Model of Fractional Wavelet Transform, Feature Selection, Support Vector Machine and Optimization Algorithm

“…Indeed, energy is identified as one of the most important and strategic issues in the global economy [37][38][39][40][41][42][43][44][45][46][47][48]. There are four forecast patterns based on the expected time frame [49][50][51][52][53][54][55][56][57][58][59][60]: (1) real-time, (2) dayahead, (3) midterm, and (4) long-term prediction patterns [61][62][63][64][65][66][67][68][69][70][71][72]. In the real-time method, forecasting is generally performed for an hour later or a fraction of it.…”

Constrained mode decomposition method for modal parameter identification

“…They are typically determined as the average of spline-approximated upper and lower envelopes of the signal, but definitions based on signal trendlines are also possible. 22 Originally defined for single-channel data, the method was recently extended to multivariate EMD (MEMD) to process multivariate signals. 23 Ensemble EMD (EEMD) is a noise-assisted version that processes an ensemble of intentionally noise-contaminated signals in a procedure that is expected to average out nonphysical IMF components and to automatically distribute the signals to the appropriate reference scales, which helps to avoid mode aliasing of EMD.…”

“…They are respectively based on the empirical mode decomposition (EMD) and the variational mode decomposition (VMD).The EMD method 20 has been first proposed by Huang et al 21 The response components, called the intrinsic mode functions (IMFs), are extracted recursively from high to low frequency. They are typically determined as the average of spline‐approximated upper and lower envelopes of the signal, but definitions based on signal trendlines are also possible 22 . Originally defined for single‐channel data, the method was recently extended to multivariate EMD (MEMD) to process multivariate signals 23 .…”