What question did this study set out to answer?

The aim is to develop a model that accurately forecasts fatal construction accidents to enhance safety management.

April 16, 2026Buildings2 citationsOpen Access

Forecasting Fatal Construction Accidents Using an STL–BiGRU Hybrid Framework: A Multi-Scale Time Series Approach

YCYuntao CaoShenyang Jianzhu University RZRui ZhangQingdao University of Technology ZQZiyi QuQingdao University of Science and Technology

Key Points

The aim is to develop a model that accurately forecasts fatal construction accidents to enhance safety management.
Utilized a hybrid STL–BiGRU framework for forecasting
Decomposed accident time series into trend, seasonal, and residual components
Trained independent BiGRU models on each decomposed component
Aggregated forecasts from individual components
Compared performance against traditional forecasting models including ARIMA and LSTM.
Achieved a short-term RMSE of 6.8522 and medium-term RMSE of 7.0568
Demonstrated consistent improvements in Mean Absolute Error and Mean Absolute Percentage Error
Outperformed conventional models like GRUs and SVR in forecasting accuracy.

Abstract

Accurate forecasting of fatal construction accidents is critical for proactive safety management; however, accident time series exhibit strong non-stationarity, nonlinear dynamics, and multi-scale temporal patterns that challenge conventional models. This study proposes a hybrid STL–BiGRU framework that integrates Seasonal–Trend decomposition using Loess (STL) with a Bidirectional Gated Recurrent Unit (BiGRU) network to deliver robust and interpretable forecasts tailored to construction safety needs. STL first decomposes the original monthly accident series (January 2012–December 2024, OSHA) into trend, seasonal, and residual components, reducing structural complexity and mitigating non-stationarity. Independent BiGRU models are then trained on each component to capture bidirectional temporal dependencies, and final forecasts are reconstructed through component aggregation. Comparative experiments against Gated Recurrent Units (GRUs), Long Short-Term Memory (LSTM), Recurrent Neural Networks (RNNs), Support Vector Regression (SVR), Autoregressive Integrated Moving Average (ARIMA), and their STL-enhanced variants demonstrate that the proposed STL–BiGRU model achieves superior performance across both short-term and medium-term horizons. The model achieves the lowest error levels, with a short-term Root Mean Squared Error (RMSE) of 6.8522 and a medium-term RMSE of 7.0568, and shows consistent improvements in Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). Results indicate that multi-scale decomposition combined with bidirectional deep learning provides a practical, forward-looking tool. It helps regulators and contractors anticipate high-risk periods, optimize resource allocation, and reduce fatal accidents through targeted preventive measures.

Perguntar à IA

Bookmark

View Full Paper