Efficient Utilization Method of Motorway Lanes Based on YOLO-LSTM Model

Area of Science:

• Computer Science
• Civil Engineering
• Transportation Science

Background:

• Urban development exacerbates traffic congestion, diminishing motorway transport efficiency.
• Real-time traffic management is crucial for mitigating congestion and optimizing road networks.

Purpose of the Study:

• To develop and validate an intelligent system for real-time traffic congestion classification and management.
• To provide a data-driven approach for optimizing emergency lane usage on motorways.

Main Methods:

• An improved ML-YOLO (YOLOv8n) video data extraction model integrated with DeepSORT for object tracking.
• Real-time extraction of traffic flow parameters to classify congestion levels using K-means clustering.
• Implementation of Long Short-Term Memory Dropout (LSTM-Dropout) and an emergency lane opening model for congestion warnings.

Main Results:

• The model accurately classifies traffic congestion levels and provides real-time traffic flow parameter extraction.
• The system successfully implemented road congestion warnings and an emergency lane management strategy.
• Verification through LSTM time series model showed low relative error, confirming model stability and practicality.

Conclusions:

• The developed model offers a reliable method for motorway traffic managers to decide on emergency lane operations.
• The system effectively alleviates motorway congestion, enhances traffic efficiency, and holds significant practical value.