1、翻译部分英文原文：Prediction of Coal Mine Safety Level Based on LSSVMDesheng Liu1, Zhiru Xu1, Wei Wang2, Lei Wang11 College of Information and Electrics Technology Jiamusi University Jiamusi, ChinaE-mail:desheng9122 College of Engineering Technology Northeast Forestry University Harbin, ChinaE-mail: cedar3Ab
2、stractCoal mine disaster has a serious threat to production and safety, mine safety prediction is an extremely challenging problem from many perspectives. This paper describes a generic fusion model for coal mine safety combining information from several physically different sensors aiming to the de
3、tection, monitoring and crisis management of such natural hazards. A conduct model base on least squares support vector machine (LSSVM) is proposed. Experimental results from the coal mine sensors are presented.Keywords-coal mine safety; multisensor fusion; support vector machine; I. INTRODUCTIONCoa
4、l mine safety evaluation is an extremely challenging problem from many perspectives .The most crucial problem is how to eliminate hidden dangers, the eradication of illegal, protect the personal safety of workers, improve the level of safety management and reduce the incidents. Therefore, coal mine
5、safety evaluation method and its application research is an urgent problem which is needed to address currently in the coal industry safety. However, the long-term mine monitoring system is in a state of uninterrupted work, and poor working conditions underground coal mines, all kinds of interferenc
6、e would have serious implications for monitoring system performance.In recent years, many scholars committed to addressing mine safety monitoring issues, such as ：remote sensing1, 2, BP Network3, X-ray diffractometry4, infrared ray gas detection 5, underground coal gasification (UCG) technique 6 .Th
7、e above methods are the study of one aspect of the mine disaster, such as fires, gas explosions, collapses and other accidents. In the practice, condition of the mine is Considerable bad, a variety of disasters may occur simultaneously and affect each other, which has brought a great difficulty abou
8、t the prevention of coal mine safety; On the other hand, the traditional method is to use a single sensor to detect underground environmental parameters, when the sensor fails, the information is not reliable, this method will lose its monitoring functions. In the case, single sensor can not be comp
9、letely due to provide the required information which will lead to uncertainty of the situation, how to protect the stability and effectiveness of the system is the fundamental problem. To detect the fault mode and to study diagnostic fault methods of the sensor is important for improving the reliabi
10、lity of systems of safety supervision. The appropriate increase in the number of similar sensors, eliminating the limitations posed by a single sensor and the application of multi-sensor data fusion theory to deal with multi-source data, it is an effective way to solve this problem. There are many w
11、ays about multi-sensor data fusion in the literature, according to coal mine safety class structure and features, in this paper, we choose the support vector machine for multi-source sensor analysis, first the classify model is be carried out using a small sample of empirical data, then predict secu
12、rity level by the unknown sample the data, compare prediction accuracy and prediction of the standard.The paper is structured as follows: section 2 describes coal mine safety class, focused on analyzing the impact of fire and gas disaster mechanism. Support vector machine(SVM) fusion model, sensing
13、process and model structure are given in Section 3. This section also describes least squares support vector machine algorithm and its terminology definition. Experiment data analysis and discussions are given in section 4. II. E VALUATION OF COAL MINE SAFETY LEVELSafety evaluation is a principle or
14、 method in the application security system engineering, for the proposed or existing project, the system may exist in dangers and possible consequences which are comprehensively assessed and predicted, and according to the risk of accidents, make the corresponding security countermeasures in order t
15、o achieve engineering systems security. In China, monitoring, and control system is mainly to monitor underground gas, carbon monoxide, wind speed, temperature, air pressure, smoke and throttle switch and so on. However, the key monitoring sensor parameters are gas and the fire signals, and fires, a
16、nd gas disasters are the most serious and common. Mine the event of fire, not only to the affected units causing huge economic losses, but also to work in the coal mine workers, employees in the relevant place and the personal safety of urban dwellers is seriously threatened, when there is gas and coal dust explosion hazard mine may also cause gas and coal dust explosion, the harm is more serious. Thus, it is very importan
