Load estimation precision for moving forces with position uncertainties through AKF

This paper presents a quantitative analysis of load estimation through augmented Kalman filter (AKF) for dynamic systems subjected to moving forces with position uncertainties. AKF incorporates unknown inputs with estimated system states, enabling the filter to estimate unknown forces. Identification algorithms like AKF often necessitate knowledge of the position of a moving load; however, the uncertainty in force position is frequently overlooked in the analysis. These uncertainties may stem from a variety of factors, including sensor noise, measurement inaccuracies, or the inherent variability in the force’s intended path. For example, vehicle position tracking can be achieved through straightforward methods like entry and exit sensors, assuming a constant velocity model. Alternatively, it can involve more sophisticated remote sensing techniques, such as computer vision or LiDAR-based systems, which provide a more accurate representation of the vehicle’s position. In this context, this paper presents a pioneering effort to quantify the impact of load estimation precision in scenarios involving moving forces with position uncertainties under diverse conditions using AKF. The study focuses on a simply supported bridge case, with the tuning of AKF parameters accomplished through the L-curve method. Tests involving various position noise levels combined with different velocities of the applied forces were conducted.