Abstract: Starting from the specific requirements of manual patrol inspection and the actual requirements of on-site safety patrol inspection in the Code for Coal Mine Safety, in view of the low degree of automation and informatization of the current coal mine safety inspection, the reliability of the inspection results is untrusted. This paper proposes a supplementary method of artificial patrol inspection, by means of information technology and intelligent research and decision to improve the efficiency of security patrol inspection and the reliability of patrol data in typical scenarios. Based on the analysis of the characteristics of artificial wind patrol scene, underground gas patrol scene and methane sensor calibration patrol scene, the architecture of safety patrol information system is proposed. Corresponding to three typical patrol scenarios, the reliability research model of patrol data is established by using the data of precise personnel positioning system, security monitoring system and security patrol information system. Taking the methane sensor calibration patrol data of the fully mechanized mining face as an example, the reliability of the calibration data of the four methane sensors in the air inlet, upper corner, working face and air return roadway is evaluated. The results of analysis are highly consistent with the actual situation in the field, which can improve the validity and reliability of security patrol to a certain extent.