Multimodal Fusion-Based Self-Calibration Method for Elevator Weighing Towards Intelligent Premature Warning

Area of Science:

• Engineering
• Mechatronics
• Sensor Technology

Background:

• Elevator load-weighing systems are crucial for safety but suffer accuracy loss.
• Nonlinear deformation of rubber buffers, influenced by environmental factors, causes precision degradation.
• This leads to safety risks like missed overload alarms and false empty status detections.

Purpose of the Study:

• To develop an online self-calibration method for improving elevator load-weighing accuracy.
• To address the precision deterioration caused by rubber buffer deformation.
• To create a robust and automated solution for load monitoring.

Main Methods:

• Constructed a reference detection model linking load to rubber buffer compression.
• Integrated draw-wire sensor displacement data with target detection model outputs.
• Derived a displacement-based compensation term and implemented dynamic error compensation with an online computation framework.

Main Results:

• The proposed method effectively compensates for rubber buffer deformation.
• Online self-calibration eliminates the need for manual intervention.
• Field experiments confirmed the method's effectiveness and system stability.

Conclusions:

• The multimodal information fusion and adaptive calibration approach resolves load-weighing precision degradation.
• This offers a novel technical paradigm for enhancing elevator safety monitoring.
• The system provides a robust solution for real-time load monitoring under complex conditions.