Research on alarm method of coal mine extraordinary accidents based on sound recognitio

Coal mine gas and coal dust explosion will produce explosive sound. Coal and gas outburst will produce the sound of coal cannon and the supports will produce squeaking and cracking sound. Rock burst will produce huge rock breaking sound and vibrations. Coal mine water inrush will produce "hissing" water sound, and a large amount of water inrush will produce water flow sound. Coal mine roof fall will produce roof cracking sound, coal rock hitting the ground sound and support damage sound. According to the characteristics of the sound of extraordinary accidents in coal mines, the alarm methods of accidents of mine gas and coal dust explosion, coal and gas outburst, rock burst, water inrush and roof fall are proposed. The characteristics of the time domain and frequency domain of each accident sound are different from the characteristics of other sounds, and the sound can be monitored in real time by mine explosion-proof sound pickup equipment and system. Therefore, accidents can be sensed and alarmed through the intelligent analysis of the sound and the analysis of the characteristic parameters of the sound frequency, amplitude and short-term energy. By monitoring and analyzing the sound intensity and other characteristics of different monitoring locations, the sequence of occurrence and the damage sequence of explosion-proof sound pickup equipment, the accident location is able to be determined. Based on the characteristics of each accident, the accident identification method of multi-information fusion analysis is proposed to reduce the sound interference of coal fall from the working face, blasting operations, coal mining equipment, excavation equipment, transportation and lifting equipment, power supply equipment, emulsion pumps, water pumps and local ventilators. The paper discusses the advantages and disadvantages of different sound pickup equipment and proposes that the microphone arrays should be used for mine sound pickup equipment. Furthermore, the paper proposes a sound recognition classifier applicable to coal mine extraordinary accidents.