Integrated OVMD-BiGRU-SMAC Framework for Forecasting Construction Accidents in the Kingdom of Saudi Arabia

Construction Accidents (CA) remain a major concern for occupational safety, especially within high-risk environments such as those found across the expanding construction sector in the Kingdom of Saudi Arabia (KSA). This research introduces an innovative prediction framework that combines signal dec...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE access Ročník 13; s. 124543 - 124555
Hlavní autori: Alsulami, Badr T., Khattak, Afaq
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Piscataway IEEE 2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:
Accidents
Configurations
Construction accidents
Construction accidents & safety
Construction industry
Construction site accidents
Decomposition
Deep learning
Forecasting
Image reconstruction
Injuries
Noise
Occupational safety
Optimization
Predictive models
Safety management
signal processing
time series
Time series analysis
Time-frequency analysis
Saudi Arabia
ISSN:2169-3536, 2169-3536
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Construction Accidents (CA) remain a major concern for occupational safety, especially within high-risk environments such as those found across the expanding construction sector in the Kingdom of Saudi Arabia (KSA). This research introduces an innovative prediction framework that combines signal decomposition with advanced deep learning methods to estimate CA trends. Initially, the framework applies Optimized Variational Mode Decomposition (OVMD) to break down historical CA time series into distinct temporal components known as Intrinsic Mode Functions (IMFs). These IMFs are individually forecasted using Bidirectional Gated Recurrent Unit (BiGRU) models, which are capable of learning sequential patterns in both temporal directions. To enhance the predictive accuracy, the hyperparameters of each BiGRU model are optimized using the Sequential Model-based Algorithm Configuration (SMAC) technique. The proposed framework is trained on monthly CA data in the KSA from June 2010 to March 2023. Among the tested configurations, the proposed OVMD-BiGRU-SMAC model produced the most reliable and better results and achieves RMSE value of 17.26, MAE of 14.02, and R2 of 0.874. In comparison, the OVMD-TCN-SMAC model showed the weakest performance, with an RMSE of 23.93, MAE of 19.11, and R2 of 0.742. These results demonstrate the effectiveness of combining signal decomposition with deep learning techniques in order to caputer the irregular and nonstationary patterns of CA data and provide more reliable forecasts to support safety management and proactive planning efforts.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2025.3589024