Font Size: a A A

Prediction Of The Mine Air Temperature Based On Support Vector Machine

Posted on:2014-02-09Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y DuanFull Text:PDF
GTID:2251330422950246Subject:Heating, Gas Supply, Ventilation and Air Conditioning Engineering
Abstract/Summary:PDF Full Text Request
With the increase of mining depth and the improvement of mechanized extraction, thesafety in mine production and underground workers’ physical and mental health arethreatened by high temperature and heat harm, which increase the mine accident rate,reduce workers’ productivity and impact on normal coal production. Therefore,minecooling becomes more and more important. An accurate prediction of mine air temperatureis the primary task and important foundation for the rational design of ventilation and airconditioning system.Every heat source’s influencing factors of Mine air temperature is studied according tothe heat exchange principle of the air flowing. Based on this, the main factors that affect theair temperature of portal, roadway and working face are determined, providing an importantbasis for the air temperature prediction model input factors determination.Because of the good ability of nonlinear systems, the support vector machine caneffectively capture the non-linear variation and characteristics of the airflow temperatureand is suitable for predicting small sample data to solve the problem of limited mineprediction samples. Therefore, we established and optimized the mine air temperatureprediction model based on support vector machine. We applied the support vector machineinto the prediction of the air temperature in the mine water spray portal, determining thesupport vector machine type and kernel function and selecting relevant parameters By themethod of theoretical analysis and cross-over trial. The predict results show that the supportvector machine model has the advantage of high prediction accuracy, the fitting effect andfast computation, etc.Take advantages of C language programming, the researchers has developed the MineAir Temperature Prediction software which is applied to predicate temperature of the portal, roadway and working face of excavation in mines. The software has two functions ofmodeling and prediction. It has advantages of high visual degree, high speed calculation,friendly man-machine interface and easy to operate according to the demarcation of thefunction of the Mine air Temperature Prediction software.
Keywords/Search Tags:Mine, Thermal harm, Wind temperature prediction, Support vector machine(SVM)
PDF Full Text Request
Related items