Sensor Selection and State Estimation for Unobservable and Non-Linear System Models

Article Engineering, Mechanical

Advanced optimal sensor placement for Kalman-based multiple-input estimation

R. Cumbo et al.

Summary: The paper presents a Kalman-based methodology for inverse load identification, introducing two alternative metrics to improve sensor selection and enhance the accuracy of estimated quantities.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2021)

添加到收藏夹

Article Automation & Control Systems

Uncertainty propagation on a nonlinear measurement model based on Taylor expansion

Min-Hee Gu et al.

Summary: This paper presents the propagation of uncertainty in a nonlinear measurement model using a higher-order Taylor series, taking into account partial derivatives and correlations among input variables. By generating multivariate random variables with copulas, the study obtained covariances among the products of various input variables, leading to results matching those of a Monte-Carlo simulation. This approach offers a potential alternative to time-consuming simulations and can be used to assess the linearity of uncertainty.

MEASUREMENT & CONTROL (2021)

添加到收藏夹

Article Engineering, Mechanical

A robust algorithm to test the observability of large linear systems with unknown parameters

X. Shi et al.

Summary: This work presents a robust algorithm for examining the observability of linear systems in order to estimate the dynamic states and parameters of the system. The algorithm's development is motivated by addressing the significant memory requirements of the Observability Rank Condition and aims to improve efficiency for better performance in engineering systems. Through random realizations and modular operations, the algorithm demonstrates superior performance in large linear dynamical models.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2021)

添加到收藏夹

Article Engineering, Mechanical

Fast greedy optimization of sensor selection in measurement with correlated noise

Keigo Yamada et al.

Summary: In this paper, a greedy algorithm is proposed for sparse-sensor selection in reduced-order sensing, which maximizes the determinant of the Fisher information matrix using a Bayesian estimation operator. By improving the computational efficiency of the algorithm, more accurate reconstructions can be obtained with reasonable increase in computational time.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2021)

添加到收藏夹

Article Engineering, Mechanical

State, parameter and input observers based on multibody models and Kalman filters for vehicle dynamics

Antonio J. Rodriguez et al.

Summary: The aim of this work is to introduce a novel estimator for vehicle dynamics by using a dual Kalman filter, incorporating force estimation and uncertain modeling parameters. The performance of the estimator is tested in simulation environment and shows high accuracy in estimating variables of interest for vehicle dynamics.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2021)

添加到收藏夹

Article Chemistry, Analytical

A Discrete-Time Extended Kalman Filter Approach Tailored for Multibody Models: State-Input Estimation

Rocco Adduci et al.

Summary: Model-based force estimation utilizing high-fidelity multibody models and a Kalman filter-based estimator is presented for accurate input load identification. The methodology is experimentally validated on a slider-crank mechanism, showing stability and accuracy in estimating input torque and system states. The framework efficiently handles redundant state descriptions and nonlinear dynamics, demonstrating the potential for practical applications in mechatronics.

SENSORS (2021)

添加到收藏夹

Article Chemistry, Analytical

A Novel Kalman Filter Design and Analysis Method Considering Observability and Dominance Properties of Measurands Applied to Vehicle State Estimation

Julian Ruggaber et al.

Summary: In Kalman filter design, the focus is usually on the filter algorithm and prediction model. This paper introduces a novel holistic method to design and evaluate Kalman filters, using nonlinear optimization algorithm to compute optimal filter parameters and quantify the contribution of measurands to estimation. Two new quantitative nonlinear observability measures are presented to analyze optimal filter design and influence of measurands on estimation.

SENSORS (2021)

添加到收藏夹

Article Chemistry, Analytical