Research On Theory And Application For Detection And Processing Of Transient Electromagnetic Signal

The mining geology disaster is the major issue which the mine enterprise faces, andalso relates social and the national economy sustainable development, but in our country,there is nearly a blank in high and new technology domain’s research on mine safetyforecast, just pauses in the warning not in the early warning level at present. So, there is anurgent need for developing innovative and high technology on disaster prediction, earlywarning attacks.The theory that combines transient electromagnetic signals with weak signal detectionmakes the mine geological disaster forecasting possible. But under the system of transientelectromagnetic signal, the existing weak signal detection and the processing method wereno longer suitable. In order to achieve the detection of transient electromagnetic signal, thisarticle has made some new theories and new methods of exporation in view of the transientelectromagnetic signal pretreatment technology. Also, it is part of National Natural ScienceFoundation Project named study of the mine water gushing warning technology based onULF signal and Autonomous District billion plan named study of the mine water warningequipment on the base of TEM signal. Main achievements as follows:（1）Presented a weak periodic signal detection algorithm that based on the scalingtransform stochastic resonating. This method makes full use of stochastic resonancedetection of weak signal, and eliminates the stochastic resonating system to the testingsignaling frequency limit by introducing scaling transform, and has realized the function ofwithdrawing weak target signal from the strong noise under the heat insulationapproximate theory. The theoretical analysis and the simulation results indicated that usingthe continual compression algorithm to the unknown frequency target signal that buried inthe strong noise can obtain a suitable input signal to the stochastic resonating system whichmay result in the testing target signal after the inverse transformation operation eachunknown frequency component according to the output signal resonating resonance peak’schange. Meanwhile, it can obtain the ideal test result to detect frequency component oftransient electromagnetism signal by using this system.（2）Proposed the empirical mode decomposition system based on the scalingtransform stochastic resonating noise reduction. The system fully utilize the ability of scaletransform stochastic resonance that can adaptively detect the weak periodic signal in fullfrequency and the ability of empirical mode decomposition that can adaptively decomposemultifrequency signals. The target signal scope value even the testing target signal accuratevalue can be returned to original state in this system by processing each fundamental mode component which was decomposed by the empirical mode. The theoretical analysis and thesimulation results indicated that this method can accurately realize to auto-adapted extracttarget signal under multi-frequency of strong noise background, of course, is different fromthe conventional routes which was only able to detect the target signal frequency from thestrong noise background including chirp the signal, but could not detect the scope valuelimit which each frequency component corresponded. At the same time, when we use thissystem to extract the transient electromagnetic signal in the open country, the geologicalprofile, drawn from the inversion of obtained target signals which was extracted fromnoisy signals is coincided with the practical one.Compared with the traditional methods,probing depth and accuracy have been improved significantly. Therefore, this system hasgood prospects of promotion and application.Finally, expressed affirmative to the achievement that obtains in the actual geologicaldisaster explores by method which proposed to this article, meanwhile, pointed out theinsufficiency exists in this method, and preview problem that need deep research and to besolved.