Fuzzy Neural Network PID-Based Constant Deceleration Control for Automated Mine Electric Vehicles Using EMB System

Area of Science:

• Mining Engineering
• Control Systems Engineering
• Automotive Engineering

Background:

• Underground electric trackless rubber-tired vehicles (UETRVs) currently use multi-level constant braking torque control.
• This system lacks closed-loop control for target deceleration, leading to safety and comfort issues.
• Inaccurate stopping distances are a significant drawback of existing systems.

Purpose of the Study:

• To develop and validate a constant deceleration braking system for UETRVs.
• To improve the safety, comfort, and stopping precision of automated mining vehicles.
• To introduce a closed-loop control algorithm for enhanced braking performance.

Main Methods:

• Described the architecture and working principle of a constant deceleration braking system with an electro-mechanical actuator.
• Proposed and simulated a deceleration closed-loop control algorithm using fuzzy neural network PID in Matlab/Simulink.
• Built and tested an actual brake control unit (BCU) in an industrial field setting.

Main Results:

• The fuzzy neural network PID algorithm achieved constant deceleration control within a ±0.1 m/s² deviation.
• This performance surpasses fuzzy PID (±0.11 m/s²) and classic PID (±0.13 m/s²).
• The system demonstrated rapid achievement of target deceleration (-2.5 m/s²).

Conclusions:

• The developed constant deceleration control algorithm is feasible for UETRVs.
• The new BCU provides active and smooth deceleration, enhancing vehicle safety and automation.
• This advancement improves the electrification and automation levels of mine transport machinery.