The Key Technologies of Road Elevation Detection Based on Sensor Fusion

Area of Science:

• Civil Engineering
• Geomatics Engineering
• Transportation Engineering

Background:

• Pavement elevation undulation detection is crucial for slope, flatness, 3D modeling, and quality assessment.
• Traditional methods using levels and manual coordination are slow and inaccurate.
• High-speed, non-contact vehicle-mounted detection offers a more efficient alternative.

Purpose of the Study:

• To develop and compare advanced methods for detecting long-distance pavement elevation undulation.
• To evaluate the accuracy and reliability of different sensor fusion and filtering techniques.

Main Methods:

• Proposed three detection methods: rotary encoders, accelerometers, attitude sensor data fusion; GPS RTK; and Kalman filtering.
• Employed modeling and experimental comparisons to validate the methods.
• Focused on high-speed, non-contact, vehicle-mounted road undulation elevation detection.

Main Results:

• Kalman filtering demonstrated robustness against environmental interference.
• Kalman filtering achieved higher detection accuracy than current international standards.
• Compared the performance of sensor data fusion, GPS RTK, and Kalman filtering.

Conclusions:

• Kalman filtering is a highly accurate and reliable method for detecting pavement elevation undulation.
• The proposed non-contact, vehicle-mounted approach significantly improves detection efficiency and accuracy.
• This method provides a superior data basis for pavement management and quality control.