| With the development of the society, safety problem is widely focused on increasingly, although it occurs no more frequently than coal mines in iron mines, and there exists new safety characteristic with the expansion of exploitation range, the deeper extension of working plane and complicate production system with tremendous equipments. So it has become the tendency of the development of mine safety management for the information_based management and safety assessment and prediction with the information attained timely. However, there exist many obstacles in the storage, use, transmission of the immense data. In this process, many problems should be considered and settled, such as how to set up assessment and prediction models by the mathematical theories and how to construct software architecture appropriately. So it is expected that there will be beneficial results in the study on safety assessment and prediction information system based on B/S aimed at the problems mentioned above.Taking safety management in WISCO Cheng Chao Iron works as the example, the main functions of the system based on B/S are implemented. In this process, a study is conducted on accident principles and various math theories are used to set up safety assessment and prediction models considering the local conditions and the safety information arranged. Primary work done is set forth as follows:Firstly, the safety assessment models are built up respectively on different subsystems, including Shaft Hoist System, Mining System, Mining Tunnelling System, Transportation System and Personal Safety System with the math theories of fuzzy mathematics and the FAHP and safety methods of FTA or ETA; Secondly, in order to predict mine roof caving accidents, the improved GM(1,1) has been set up based on acoustic emission method. The improved GM(1,1) takes advantage of the general GM(1,1) in prediction precision. In the meanwhile, it is the first time for data conversion used in GM(1,1); Finally, on the basis of data base created, the related models have been programmed in Python language and the system based on B/S are generated by the technology of middleware. The research above has greatly improved the veracity and timeliness of prediction.It is innovative in regulation models and other safety assessment models based on various math theories, especially in assessment model of personal safety. It is meaningful to some degree in the theory and practice of safety assessment and prediction. |