The effective utilisation of monitoring data of the coal mine is the core of realising intelligent mine. The complex and challenging underground environment, coupled with unstable sensors, can result in “dirty” data in monitoring information. A reliable data cleaning method is necessary to figure out how to extract high-quality information from large monitoring data sets while minimising data redundancy. Based on this, a cleaning method for sensor monitoring data based on stacked denoising autoencoders (SDAE) is proposed. The sample data of the ventilation system under normal conditions are trained by the SDAE algorithm and the upper limit of reconstruction errors is obtained by Kernel density estimation (KDE). The Apriori algorithm is used to study the correlation between monitoring data time series. By comparing reconstruction errors and error duration of test data with the upper limit of reconstruction error and tolerance time, cooperating with the correlation rule, the “dirty” data is resolved. The method is tested in the Dongshan coal mine. The experimental results show that the proposed method can not only identify the dirty data but retain the faulty information. The research provides effective basic data for fault diagnosis and disaster warning.

Data sets gathered continuously in air monitoring systems are never entirely complete. The problem of missing data in monitoring measure series often has to be solved by modeling. A new method of air monitoring data modelling was tested in the paper. Regional diurnal concentration courses (RDCCs) were used as the main source of knowledge of predicted time series during specified days. The paper presents a comparison of predicted and measured diurnal concentration patterns of two frequently used parameters in air monitoring (PM10 and NO2). The analysis was based on hourly time series of these air pollutants collected in a 3-year period at nine monitoring stations in the Lodz Region. It was shown that well determined regional diurnal concentration patterns could be useful to missing data modelling at the specified monitoring site. Improvement of modelling accuracy is possible after modification of modelling results by adding local difference vectors (LDVs), describing the specificity of the monitoring station.

There are several large karst caves at haunch part of the Lidong Tunnel during construction, together with inrush water due to high pressure within these caves. In light of it, this paper takes YK342+113 section as an example and adopts finite difference software FLAC 3D, so as to analyze tunnel deformation when crossing karst caves under six different working conditions, including with or without karst cave, before and after karst treatment, along with support locations. According to analysis results: First, the wall rock mainly had deformation at tunnel vault when evacuating at the third bench, which is a critical monitoring focus for tunnel construction; Second, karst cave treatment contributed to better conduct forces on both sides of wall rock, thus reducing vault settlement, while not affecting horizontal convergence and upturn of vaults; Third, treatment measures were proved to be effective in minimizing wall rock deformation by comparing deformation curves under different conditions; Fourth, after treatment measures, the angular points within the cave’s chamber had stress concentration, which might cause secondary collapse. Field monitoring data revealed that the final settlement of the tunnel vault was relatively consistent with the numerical analysis results, with a distinct change in daily settlement after initial support construction. By integrating numerical analysis and field monitoring, the rationality of the karst treatment plan was fully verified, providing a valuable reference for similar projects.