Improvement Of Moment Tensor Inversion Method And Layout Optimization Of Microseismic Monitoring System

The determination of rock fracture mechanism is the research foundation in the field of geotechnical engineering,and the acoustic emission/microseismic(AE/MS)moment tensor inversion method is an effective means to complete this kind of research.However,there are many factors that can affect the accuracy of moment tensor inversion.Large moment tensor errors can even mislead the understanding of rock/rock mass fracture,which seriously limits the wide application of moment tensor inversion methods in the field of geotechnical engineering.This thesis aims to obtain more accurate moment tensors in acoustic emission/microseismic monitoring.Starting from different scales(such as,Brazilian splitting test,uniaxial compression test,geothermal field and hydraulic fracturing),focusing on many aspects(such as,sensor calibration method,moment tensor inversion method,sensor layout),some in-depth research has been carried out,which has greatly improved the accuracy of moment tensors.The innovative results achieved are as follows:(1)A new sensor calibration method for amplitude is proposed:the Search Calibration.This method can obtain more accurate Double Couple(DC)components and stress states in a mixed-noise and high-noise environment,and can effectively improve the calibration accuracy of the amplitudes collected by sensors.Traditional sensor calibration methods in microseismic monitoring(e.g.,Network Calibration Method and Hybrid Moment Tensor method)are easily affected by systematic errors,which are detrimental to the overall monitoring results.Search calibration can overcome the weaknesses of the original calibration methods and has been successfully applied to the optimization of the moment tensor results in the microseismic monitoring of the Geysers geothermal field in California,USA.(2)A new method of moment tensor inversion is proposed:the Weighted Inversion Method.According to the effect and evolution characteristics of different quality amplitudes in acoustic emission event,this method adjusts the weight factor of each amplitude in the moment tensor inversion process to obtain more accurate moment tensor inversion results.Through theoretical calculation and multi-condition simulation analysis,it is verified that the accuracy of the results of the weighted inversion has been greatly improved,compared with the results of standard inversion and the nonlinear STC inversion.Three methods are simultaneously applied to the research of moment tensor inversion of Brazilian splitting test.The results of weighted inversion method can describe the process and form of disc splitting more accurately than the other two methods.(3)The concept of "optimal offset-to-depth ratio(R)interval" is proposed,and the optimization of regular grid sensor layout in surface microseismic monitoring is accomplished based on this concept.The results prove that the sensor layout parameters are closely related to the accuracy of the moment tensor inversion results.On the basis of the regular grid layout,the concept of circular layout and corresponding optimization methods are proposed.Through theoretical calculation and simulation analysis,it is proved that the circular layout can obtain lower moment tensor errors than regular grid layout,star layout and uniform spherical layout.In addition,the optimization process and corresponding conclusions are not limited by factors such as the number of sensors,source types and magnitude.