Multi-sensor information fusion detection system for fire robot through back propagation neural network.

Area of Science:

• Robotics and Automation
• Artificial Intelligence
• Sensor Technology

Background:

• Firefighter safety is paramount, necessitating advanced detection systems.
• Existing fire detection methods may lack the comprehensive data integration required for complex environments.
• Multi-sensor information fusion offers a promising approach to improve detection accuracy and reliability.

Purpose of the Study:

• To investigate a fire sensor multi-sensor information fusion detection system.
• To enhance firefighter safety by reducing risks during fire incidents.
• To develop a robust system for early and accurate fire detection.

Main Methods:

• Utilized a back propagation neural network (BPNN) for feature layer fusion.
• Employed fuzzy control for decision layer fusion.
• Designed a multi-sensor information fusion system integrated into a robot for data collection and processing.

Main Results:

• Optimized BPNN performance with 7 hidden layer nodes.
• Achieved lowest error after 127 iterations, demonstrating excellent accuracy.
• The trained BPNN exhibited a mean square error of 0.0013, confirming high precision.

Conclusions:

• The BPNN-based multi-sensor fusion system provides theoretical support for forest fire detection.
• The developed robot system is applicable to forest fire inspection and processing globally.
• This research contributes to advancing robotic fire detection systems for enhanced safety.