Analysis of Methods for Long Vehicles Speed Estimation Using Anisotropic Magneto-Resistive (AMR) Sensors and...

Area of Science:

• Engineering
• Computer Science
• Physics

Background:

• Intelligent transportation systems (ITSs) are crucial for managing increasing urban traffic.
• Accurate estimation of long vehicle (L > 10 m) speed and length is essential for ITS applications.
• Traditional methods like cross-correlation are computationally intensive and unsuitable for long vehicles.

Purpose of the Study:

• To develop and evaluate a novel system for estimating long vehicle speed and length.
• To introduce an adaptive signature cropping algorithm for improved magnetic signature analysis.
• To validate the system's performance against ground truth data.

Main Methods:

• A self-developed system utilizing two anisotropic magneto-resistive (AMR) sensors.
• An adaptive signature cropping algorithm applied to magnetic signatures.
• Validation using piezoelectric polyvinylidene fluoride (PVDF) sensors and video cameras for ground truth.
• Performance evaluation using Mean Absolute Percentage Error (MAPE).

Main Results:

• The proposed method accurately estimates long vehicle speed and length.
• The adaptive signature cropping algorithm overcomes limitations of traditional methods.
• Experimental results with 600 unique vehicles show an average speed MAPE error below 3% for speeds between 40-100 km/h.
• The system demonstrated robustness under various traffic and weather conditions.

Conclusions:

• The developed AMR sensor system with the adaptive signature cropping algorithm offers a viable solution for accurate long vehicle speed and length estimation.
• This method is suitable for integration into intelligent transportation systems (ITSs).
• The system provides reliable performance in real-world urban traffic environments.