Dynamic Modeling and Adaptive Frequency Control of Resonance Drilling Systems for Mine Rescue Boreholes

After a mine collapse, rapidly drilling a life support hole is essential for supplying oxygen, enabling communication, and delivering materials to trapped personnel. Resonance drilling offers high rock-breaking efficiency and low disturbance in complex broken strata, but its axial vibration and frequency mismatch under variable working conditions limit drilling efficiency. To address this, a coupled dynamic model of the top drive, drilling tool, and stratum is established. The structural parameters’ influence on resonance frequency is quantified. When the drill string length increases from 100 m to 200 m, the second-order natural frequency decreases by about 39.3%. In low-order modes, titanium alloy drill strings show natural frequencies 30–40% lower than steel, while in higher-order modes they are 10–15% higher. Based on the theoretical model, a closed-loop adaptive resonance frequency tracking control strategy combining model prediction and vibration response feedback is proposed. Simulations show steady-state frequency tracking error below 1.2% under varying depth. Field experiments confirm the predicted frequency attenuation trend, with top–bottom frequency deviation within 4 Hz, indicating surface vibration measurements can reliably reflect downhole resonance. The model prediction error remains within 5%, and the control strategy maintains resonance conditions. These results provide a theoretical basis for the design and self-optimization of intelligent drilling equipment for mine emergency rescue.