Qi Qi

ChronoSteer: Bridging Large Language Model and Time Series Foundation Model via Synthetic Cross-Modal Alignment Dataset

Conventional forecasting methods are trained end-to-end on unimodal time series, which limits their ability to exploit textual information and undermines their generalization in data-scarce scenarios. Recently, large language models (LLMs) and time series foundation models (TSFMs...

Damage identification and model updating using bridge displacement under moving vehicles with uncertain weight

Bridge damage incurred during long-term service leads to discrepancies between the physical structure and its digital twin finite element (FE) model, necessitating regular model updates to ensure reliable health monitoring. Traditional methods typically adopt a two-step framework...