Abstract: Industrial noise-induced occupational hearing loss has come up to the second place among occupational diseases in China. At present, the occupational health of the coal industry is mainly focused on dust control while few research work has so far been made on noise control monitoring and early warning. In order to explore the approaches for control and monitoring of noise generated in coal preparation processes, several schemes for noise reduction at sound sources and control of noise through transmission pathways are proposed. The schemes are proposed based on the case study on control of pollution of noise generated in the screening-crushing shop of Xiwan Mine Coal Preparation Plant, as well as analysis of the factors causing the generation of noise from the vibrating screens, scraper conveyor, chutes, etc. The noise control schemes include use of all polyurethane screenplate and rubber air spring on vibrating screen; laying buffer rubber plate between wear-resistant lining plate and steel plate on inner surface of chute; changing the scraper conveyor′s early-warning mode to electromechanical mode; installation of sound insulation covers on high-noise equipment; and laying a constrained damping layer in chute. For realizing real-time monitoring and early warning of noise, an intelligent noise monitoring and early warning system is developed based on deep learning sound recognition technology. Study results show that with implementation of the comprehensive noise control measures, the noise level in the screening-crushing shop is up the state-specified standard without exceeding the permissible occupational noise exposure limit of 97 dB (A); in addition to intelligent monitoring of noise, the system can give forth early warning according to the zoning and grading principle; the use of the system can ensure not only occupational health of operators in screening-crushing shop, but also give forth early warning of equipment health for extending their service life.