A Kernel-Based Method for Modeling Non-harmonic Periodic Phenomena in Bayesian Dynamic Linear Models

  title={A Kernel-Based Method for Modeling Non-harmonic Periodic Phenomena in Bayesian Dynamic Linear Models},
  author={Luong Ha Nguyen and Ianis Gaudot and Shervin Khazaeli and James A. Goulet},
  journal={Frontiers in Built Environment},
Modeling periodic phenomena with accuracy is a key aspect to detect abnormal behaviour in time series for the context of Structural Health Monitoring. Modeling complex non-harmonic periodic pattern currently requires sophisticated techniques and significant computational resources. To overcome these limitations, this paper proposes a novel approach that combines the existing Bayesian Dynamic Linear Models with a kernel-based method for handling periodic patterns in time series. The approach is… 

Real‐time anomaly detection with Bayesian dynamic linear models

The Rao‐Blackwellized particle filter with the Bayesian dynamic linear models (BDLMs) is presented, which succeeds in detecting the anomaly caused by the refection work on the dam as well as the artificial anomalies that are introduced on the original dataset.

Bayesian dynamic linear model framework for structural health monitoring data forecasting and missing data imputation during typhoon events

A Bayesian dynamic linear model (BDLM) framework for data modeling and forecasting is proposed to evaluate the performance of an operational cable-stayed bridge, that is, Ting Kau Bridge in Hong

The Gaussian multiplicative approximation for state‐space models

The Gaussian multiplicative approximation (GMA) is proposed that is applicable to multiplicative state‐space models that are often encountered in practical SHM applications and exceeds the performance of the existing nonlinear Kalman filter methods in terms of accuracy along with the computational cost.

Anomaly detection using state‐space models and reinforcement learning

An anomaly detection framework is proposed that combines the interpretability of existing Bayesian dynamic linear models, a particular form of state‐space models, with the long‐term planning ability of reinforcement learning and is demonstrated on monitoring data collected on a bridge in Canada.

Automatic quality detection system for structural objects using dynamic output method: Case study Vilnius bridges

This paper presents an automated structural health monitoring approach based on the algorithm-driven definition of the first resonant frequency value from a noisy signal, acquired from trafficcreated bridge vibrations, and compares three methods of data processing.



Anomaly detection with the Switching Kalman Filter for structural health monitoring

This paper proposes an anomaly detection method that combines the existing Bayesian Dynamic Linear Models framework with the Switching Kalman Filter theory and shows that the approach succeeded in identifying the anomaly caused by refection work, without triggering any false alarm.

Bayesian dynamic linear models for structural health monitoring

In several countries, infrastructure is in poor condition, and this situation is bound to remain prevalent for the years to come. A promising solution for mitigating the risks posed by ageing

Time Series: Modeling, Computation, and Inference

A graduate-level account of Bayesian time series modeling and analysis, a broad range of references to state-of-the-art approaches to univariate and multivariate time series analysis, and emerging topics at research frontiers are included.

Time Series Prediction with Periodic Kernels

  • M. Michalak
  • Computer Science
    Computer Recognition Systems 4
  • 2011
This short article presents the new algorithm of time series prediction: PerKE. It implements the kernel regression for the time series directly without any data transformation. This method is based

Bayesian Filtering and Smoothing

  • S. Särkkä
  • Computer Science
    Institute of Mathematical Statistics textbooks
  • 2013
This compact, informal introduction for graduate students and advanced undergraduates presents the current state-of-the-art filtering and smoothing methods in a unified Bayesian framework and learns what non-linear Kalman filters and particle filters are, how they are related, and their relative advantages and disadvantages.

A Course in Time Series Analysis

Introduction (D. Pe?a & G. Tiao). BASIC CONCEPTS IN UNIVARIATE TIME SERIES. Univariate Time Series: Autocorrelation, Linear Prediction, Spectrum, State Space Model (G. Wilson). Univariate

Effect of vehicular loading on suspension bridge dynamic properties

Since the 1970s, many researchers have attempted to use changes in natural frequencies as means for condition assessment of large civil engineering structures such as bridges, but have faced the

An Introduction to Kalman Filter

A practical introduction to the discrete Kalman filter, a set of mathematical equations that provides an efficient computational means to estimate the state of a process, in a way that minimizes the mean of the squared error.