Safety Evaluation And Dynamic Prediction For The Rescue Operation After Mine Thermo Dynamic Disasters

Thermo dynamic disaster is a major type of disaster in coal mine, accompanying with a sudden and rapid development, a large number of casualties, serious secondary disasters, and the difficult rescue. The ventilation disorder of the disaster area resulted from the effects of thermal activities further complexes and blurs the disaster, even induces the gas explosion during the rescue. In addition, rescue decisions tend to encounter a dilemma situation, the danger of the disaster will significantly increases as long as a wrong rescue decision was made. Consequently, it is meaningful to study the safety evaluation and dynamic method of the mine thermo dynamic disaster rescue, which will provide the theoretical guidance for the rescue decision making.Firstly, the causes, influence factors, types of disaster, and spread characteristics of the thermo dynamic disaster in coal mine is analyzed and summarized. Types and forms of the thermo dynamic secondary disaster in coal mine are analyzed. The characteristics and rules of coal disasters are studied. The danger of the secondary disaster is comprehensively evaluated from the standpoint of the rescuers. It is found that there are strong fuzziness and difficult to predict characteristics with thermo dynamic secondary disaster under the influence of thermal action. Especially the secondary gas explosion is more difficult to accurately determine, and it is a serious threat for the rescuers.Secondly, combined conditions and influencing factors of gas explosion with the characteristics and laws of gas explosion in coal mine, the determination method of gas explosion was proposed. This determination method can be applied to predict time, possibility, and location of the gas explosion. According to increasing characteristics of gas and oxygen after the thermo dynamic disaster, gas explosion is divided into high gas concentration explosion and low gas concentration explosion. Accordingly, time ranges of their explosions were classified. Based on the characteristics of the gas accumulation, the underground area is divided into two regions, which are closed region and tunnel region. Applying the proposed determination method into these two regions, characteristics of the gas explosion occurred at these two regions were analyzed. Then, the scientific and reasonability of the proposed method was verified.Thirdly, based on modern accident-inducing theory, factors impacting the safety of mine rescue were classified into 14 categories which were further divided into 50 secondary classification by employing analyzing method of "man machine environment management". The evaluating indices indicating the safety of the mine rescue was established for the first time. The selection of evaluation indices and the optimization method were analyzed and summarized. Combined these with the characteristics of evaluation target and building principles of the determination indices, the evaluation indices of a mine thermal dynamic disaster rescue was optimized according to Delphi method and AHP method. The optimized evaluation indices of a mine thermal dynamic disaster rescue was obtained. On the basis of the characteristics of safety evaluation indices of mine thermo dynamic disaster rescue, methods of processing, quantification, and weight assignment of evaluation indices is put forward.Fourthly, according to the proposed modeling ideas of evaluation model and optimized evaluating indices, AHP-FCE evaluation model is established based on AHP analysis and the fuzzy theory; the model was verified via three specific thermo dynamic disaster rescue cases in coal mine; Based on the similarities and differences between coal and non-coal mines, the model was firstly applied to the rescue of non-coal mines to test the reasonability of the "manmachine-management" and model. In addition to this, the model was applied to evaluate the safety grade of collected 60 groups of disaster in 20 rescue cases to evaluate. This work provides scientific and accurate data for the subsequent dynamic prediction model based on the artificial intelligence algorithm.Finally, the theory and algorithm of extreme learning machine(ELM) were introduced based on evaluating characteristics of the thermo dynamic disaster. Steps and characteristics of several common algorithms of ELM were analyzed and summarized. Then, dynamic prediction model, which is based on the improved incremental extreme learning machine(I-ELM) algorithm, was established to determine rescue safety of thermo dynamic disaster in coal mine. Finally, the model is used in predicting the safety of field actual thermo dynamic disaster rescue case quantitatively, and the accuracy and reliability of the prediction method are testified.Results of the study will provide guidance for quantificational and efficient evaluating the safety of the mine rescue based on the real time information collected during the rescue process. This is of importance for mine thermo dynamic disaster rescue.