Data Analysis Of Ventilation And Gas Based On Mine Internet Of Things

Coal mine ventilation systems and gas prevention have always been the focus of research in the field of coal mine safety.A good ventilation system is one of the keys to ensure the safety of coal mines,and gas accidents are the most influential of all disasters in coal mines,the most serious casualties,and the highest frequency.Therefore,improving the accuracy of ventilation and gas data has more positive practical significance for improving the safety of personnel and equipment in coal mines.This paper focused on the study of ventilation network calculation and gas emission prediction to improve the accuracy of ventilation and gas data.This paper collected basic data on ventilation and gas based on the existing mine IoT data monitoring system,and used the multi-sensor data fusion technology for data analysis to improve the accuracy and availability of the data in the case of complex underground mine environment and data transmission that are easily interfered.Based on the collected data processed by multi-sensor data fusion technology,the loop air volume method using Scott-Hinsley algorithm was improved to establish a ventilation network model to obtain higher accuracy global air volume data.Aiming at the complex and nonlinear dynamic system of gas emission,a gas emission prediction index is established,and a gas emission prediction model based on PCA-MLR-RBF was proposed.Research indicates:(1)Based on the existing coal mine IoT environmental data collection system,the system and topology were analyzed and studied.Multi-sensor data fusion is used to analyze the collected data to improve the accuracy and reliability of the data.Compared with the single node data collection in traditional coal mines,the data accuracy is improved,misreporting and underreporting of data are reduced,and the linkage and reliability of the ventilation system are enhanced.(2)Based on the coal mine ventilation network diagram theory,the coal mine ventilation system was studied.Under the principles of composition processing and simplification,a coal mine local ventilation network diagram was generated;in view of the existing ventilation network solution method,there is a huge loss when looping a complex ventilation network,and the convergence of the network cannot be guaranteed.In this article,the Scott-Hinsley loop air volume method is improved to reduce the number of branches in the generated loop,thus the accumulated error caused by loop iteration is reduced.The generated local ventilation network diagram was used as an example to verify that the solved air volume data had higher accuracy.(3)According to the principle of gas emission prediction index selection,the influencing factors of gas emission are selected as the prediction indicators;utilizing the abilities of principal component analysis to effectively eliminate multivariate collinearity,good processing linear data characteristics of multiple linear regression analysis and excellent nonlinear approximation of RBF neural network,a PCA-MLR-RBF gas emission prediction model is proposed.The experimental results showed that the gas emission prediction model based on PCA-MLR-RBF had greatly improved the accuracy of prediction results compared with the existing prediction methods.