Abstract:In response to the intricate installation challenges and the elevated cost of sensors for measuring base shear in large‐scale structures, this paper proposes a noncontact measurement method integrating computer vision and model updating to invert structural base shear. The computer vision part measures physical displacement, while the nonlinear model updating section inverts base shear by refining the structural numerical model, thus achieving cost‐effective, noncontact inverting measurements. In the computer vision component, a highly real‐time and accurate optical flow estimation algorithm was selected and validated in actuator motion tracking tests, yielding a normalized root mean square error of less than 3% between displacement tracking and sensor measurable results. The model‐updating section adopts the Bouc–Wen model, demonstrating through numerical simulations its ability to swiftly calibrate the numerical model within 7000 steps under various noise interference levels, accurately obtaining structural base shear. Moreover, the influence of different response combinations and sampling frequencies on parameter identification for model updating is discussed. Findings indicate that when considering both displacement and acceleration, along with a sampling frequency of 200 Hz, parameter identification meets accuracy requirements due to reduced susceptibility to measurement noise. In addition, a shake table test on a three‐layer shear frame is conducted to further validate the proposed method's feasibility. Test results demonstrate that the amplitude and fluctuation trend of the shake table test's identification results mirror those of the numerical simulation results within the first 25 seconds, with a peak value error of 18.9%. While the error is relatively large, this paper provides a practical research framework for model updating and structural health monitoring. Simultaneously, it reduces the cost of acquiring structural response data during tests, thereby facilitating the application and promotion of computer vision technology in structural response monitoring.

A Model‐driven Method Based on Multisource Data for Reproducing Displacement Time History

Estimation of the displacement time history of high-rise building structures using limited measurement data and structural information

Variational Mode Decomposition-Based Multirate Data-Fusion Framework for Estimating Structural Dynamic Displacement by Integrating Vision- and Acceleration-Based Measurements

Interpretable Machine-Learning Models for Maximum Displacements of RC Beams under Impact Loading Predictions

Error Study of Displacement Model for Seismic Analysis of Long-Span Structures Subjected to Multi-Support Earthquake Excitations

Hybrid surrogate model combining physics and data for seismic drift estimation of shear‐wall structures

An efficient method based on shear models for structural seismic response prediction considering hysteretic characteristics

Model updating of a shear‐wall tall building using various vibration monitoring data: Accuracy and robustness

A Modified Newmark Method for Calculating Permanent Displacement of Seismic Slope Considering Dynamic Critical Acceleration

Dynamic displacement estimation and modal analysis of long-span bridges integrating multi-GNSS and acceleration measurements

An Output‐only Structural Condition Assessment Method for Civil Structures by the Stochastic Gradient Descent Method

A hybrid model-data method for seismic response reconstruction of instrumented buildings

A Smart Multi-Rate Data Fusion Method for Displacement Reconstruction of Beam Structures

Enhanced Measurements of Structural Inter-Story Drift Responses in Shaking Table Tests

Noncontact Measurement Method for Inverting Structural Base Shear

Structural displacement estimation using accelerometer and FMCW millimeter wave radar

Post-earthquake structural damage assessment and damage state evaluation for RC structures with experimental validation

Story drift and damage level estimation of buildings using relative acceleration responses with multi‐target deep learning models under seismic excitation

Shape function-based multi-objective optimizations of seismic design of buildings with elastoplastic and self-centering components

Damage Inference and Residual Capacity Assessment for an E-Defense 2018 Ten-Story RC Test Structure

Seismic Model Parameter Optimization for Building Structures