Abstract: Although dust monitoring and control have been achieved in the coal mine belt transfer point through the installation of sensors and dust reduction equipment,when the concentration of dust in the area is abnormal,the investigation of the abnormal causes still highly relies on manual experience, and there is a problem of response lag. In order to remotely, quickly,and accurately locate the cause of abnormal causes, a dust concentration anomaly tracing algorithm integrating the ordered constraint Apriori algorithm and the traceability tree mechanism is proposed. Based on the layout of dust-proof areas,the Apriori algorithm with ordered constraints is used to mine abnormal association rules and generate a table of abnormal causes with priorities. By combining the dynamic threshold and instantaneous threshold of dust concentration to set the traceability trigger conditions,a multi -level traceability tree integrating backtracking markers and node weights is constructed. When the trigger conditions are met,a depth-first search algorithm based on backtracking markers and node priorities is adopted to quickly locate the cause of the anomaly. Experiments show that in the scenario of mine belt transfer point,this algorithm can effectively identify the high-frequency causes of abnormal dust concentration (such as spray device failure),and the traceability accuracy rate is 96. 96%.