Research On Rock-burst Prediction Model Of Multi-parameter Information Mining Based On Vibration Signal

The stable, sustained and healthy development of coal resources is directlyrelated to the security of national energy. The disasters caused by mine exploitationare an unavoidable reality. Disastrous accidents caused by deep exploitationrepresented by rockbursts, mine earthquakes pose more destructive, sudden andserious threatens to the safety production of mine, which has caused the enormouseconomic losses and casualties in China.Many foreign countries with serious rockburst have established a national minemicroseismic monitoring network to monitor, forecast and reduce the risks ofrockburst. In recent years, China has introduced and developed a number ofmicroseismic monitoring systems for different research purposes and monitoringscales in order to monitor and predict rockbursts, mine earthquakes and othernonlinear dynamic disasters. Studies on the domestic application and theory ofmicroseismic monitoring technique are still on the primary stage, mainly focusing onthe study of the source location, motion events, statistical analysis etc. Research israre on source rupture mechanism, impact damage effect and information of vibrationsignal. There is less research on how to analyze the information revealed by thesignals from a large number of data acquired from advanced instruments.Through the field experiment to obtain the microseismic monitoring data, thedissertation focuses on the following key issues:1. Through the establishment of the seismic source equivalence models ofdifferent coal rock vibrations, the dissertation systematically analyzes different impactrupture models and their energy radiation characteristics; studies the correspondingrelationship between rupture models of coal rocks and vibration signal models andproposes an identification method based on Fourier‘s analysis of the vibration signalsfrom the underground coal mines. The paper, taking the coal pillar blasting vibrationsignal and fault slip vibration signal as an example, has established the correspondingrelationship between the rupture model of seismic source and the vibration signals andtherefore has successfully puts forward a method to figure out the rupture model ofthe seismic source through the identification of the model of the signal.2. There exists a close relationship between the research of the attenuation law ofenergy and frequency and the accurate prediction of the rock burst when the vibrationsignal transmits in coal rocks. This paper through a research of the attenuation law of the energy decay and frequency changes in the vibration signals, has established thephysical model of the vibration wave when it transmits in coal rocks and has realizedits mathematical model through the digital signal processing technology, which lays afoundation for collected vibration signal which can invert the focus to monitor theamount of energy under monitoring.3. Location accuracy of the seismic source can directly affect the function of theearly-warning system of the rockburst. This paper, starting from the layout of sensorsand optimizing the number of channels, has proposed a study on how to improve theprecision of micro-seismic monitoring. The paper has also brought forward theobjective function based on optimal layout of sensor networks so as to establish thesolution mode based on genetic algorithm which is intended to solve the issue ofcombinatorial planning of large-scale sensor networks. In addition, through simulationexperiment and field application the superiority of the algorithm has also beenverified. Meanwhile, the paper, with the microseismic monitoring system to locate,touches upon the relationship between the number of the channels and the locationaccuracy of the seismic source, so as to provide guidance for increasing the accuracyof locating operation.4. Based on previous experience and various research findings on predictionsystems and in view of different mine geological conditions and actual exploitationconditions, the paper aims at establishing the optimal exploitation paradigm. In themean time, through acoustic emission monitoring and dynamic mining stressmonitoring and through a combination of fuzzy logic and support vector machineaccording to the fuzzy nature of variables, the paper aims at establishinganearly-warning model for the risks on the basis of double-layer coal and rock impact ofvague support vector machine.