The Appliction Of Data Mining In Colliery Safety Monitoring

The coal domain is one of the most important basic industries in our country, which is belonging to high risk vocation determined by its connatural characters. The factors which could influence the colliery producing safely was very complex, the natural disaster is the main factors such as gas concentration, the coping pressure, carbon monoxide concentration, firedamp concentration, temperatures and so on. How to efficiently identify the relations in the factors thus improving the management level of the production safety of colliery is the major items with which was faced in our country's coal industry.In this thesis, the data mining based method Association Rules was used to find out the relations between the disaster factors in the producing process of colliery on the samples which were obtained from the information system of colliery safety monitoring. The purpose of this study is to provide some favorable means for enhancing the efficiency and degree of colliery safety monitoring and realizing the colliery safety monitoring automatically and expert systems of colliery safety monitoring. This thesis includes several works as following:1.Describing the details of data mining techniques, the Association Rules was employed to analysis the samples obtained form the information system of colliery safety monitoring combining with the features of the monitoring dataset generated in the colliery safety production processes.2.The data selection, purge and ensemble, data discretization and concept hierarchy were used in the original dataset pretreatment.3.The Apriori algorithm was improved and used in finding out the dataset obtained from the information system of colliery safety monitoring, the results indicated that the improved algorithm showed higher effectiveness for a given period of time. The results obtained by the Association Rules presented preferable effects for advancing the safety monitoring and alerts of colliery.