PHYSICS-AWARE TUNING OF UNSCENTED KALMAN FILTER TO ESTIMATE PARAMETERS, QUANTIFY ABRUPT PARAMETER CHANGE, AND QUANTIFY UNCERTAINTY IN IDENTIFICATION OF A DYNAMIC SYSTEM WITH MANY DOFS
Bibliographic record
Abstract
The Unscented Kalman Filter (UKF) represents a robust method for estimating latent states and parameters within specified nonlinear equations of dynamic systems under noisy sensor data. Nonetheless, adjusting the filter's hyperparameters (HP)s is essential for effective performance and poses difficulties, especially in systems with many parameters and states to identify, or when it is necessary for the filter to both detect and measure anomaly levels in parameters. Building on the authors previous research on introducing a physics-aware objective function to tune the UKF, this study advances the capabilities of the objective function to tune different adaptive variants of the UKF which facilitates virtual sensing, joint state-parameter estimation, damage quantification and uncertainty quantification under partial observation for systems with many degrees of freedom (DoF) as an open-ended question in structural health monitoring field. To validate the framework, a three DoF damped mass-spring system experiencing a sudden change in physical characteristics is used. Subsequently, the filter's precision in estimating parameters and states is evaluated using a ten DoF system with 40 states and unknown parameters, featuring sparsely placed sensors. Furthermore, the Lorenz attractor under partial observation is used as another case study to highlight why and how the physics-aware objective outperforms other commonly used data-driven objective functions. These results demonstrate the potential of the proposed framework for addressing challenging identification problems in dynamical systems such as tracking sudden changes and evaluating the uncertainties linked to both modeling and measurement, particularly those with limited and noisy sensor data.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".