| The mining industry is the guarantee foundation of the national resource security strategy and also a high-risk industry.It is difficult to prevent and control risks.In particular,high and steep slopes,tailings ponds(dams),etc.,as the "hazard sources" associated with mining,are the most important factors in the safety production of mines,because once they get out of control,they are likely to cause serious accidents that may lead to mass casualties.The traditional contact point monitoring technology is easily affected by the environment,which has the problems of difficult location of hidden dangers and low early warning efficiency.In recent years,the newly developed ground-based real aperture radar is based on the phase interference principle of microwave remote sensing.Compared with the synthetic aperture system with fixed antenna beam center,it adopts the point by point scanning mode,and can obtain high-precision deformation and displacement information all day/all weather.It has become a research hotspot of slope deformation monitoring technology.The key technologies for accurate positioning and effective early warning of landslide hazards based on the monitoring results of slope deformation and displacement are accurate identification and positioning of radar imaging targets and monitoring data matching with slope space This thesis focuses on three problems in real ground-based real aperture radar,namely,multiple interference of radar imaging targets,large error of target slant range parameters and low registration accuracy of non-uniform sparse point sets.The achievements have been applied to the self-developed Ku band 1.2m aperture offset paraboloid based real aperture slope radar system(S-RAR).Through the monitoring application of the actual mine slope,the effectiveness and practicability of the key technologies for slope deformation monitoring proposed in this thesis has been verified.The main innovations of this thesis are as follows:1.A 3D imaging target recognition method of ground-based real aperture radar based on slope smoothness constraint is proposed,which can effectively eliminate interference targets generated under the peak constraint criterion of more than 90%.S-RAR scans and images the slope at the set signal sampling position of the turntable.Due to the large coverage of the spot beam and the complex interference of engineering machinery,vehicles and other vehicles on the slope monitoring site,there are several peak point targets in the near and far range of radar one-dimensional range profile.Affected by interference,it is difficult for conventional peak constraint criteria to identify the imaging target with stable strong scattering characteristics that best conforms to the spatial distribution of the slope from a large number of imaging candidates,resulting in that the radar deformation monitoring data cannot accurately reflect the slope surface deformation trend.This innovation point establishes a multi threshold joint optimization algorithm and a smoothness analysis model of radar monitoring unit based on surface roughness and slope gradient characteristic parameters.Through the normalized amplitude deviation,coherence and other parameters,the multi threshold joint optimization identification of redundant peak point targets in the slope monitoring area is carried out,and the candidate point set of 3D imaging targets is established;The 3D imaging target is finally determined according to the spatial continuity of the true profile of the slope.In the 3D imaging application of mine slope,the method proposed in this thesis can identify the imaging target of the main scatterer of slope stability under the objective condition of satisfying the spatial continuous distribution of slope,and effectively eliminate the interference target generated under the peak constraint criterion of more than 90%.2.The slant range parameter is the key to the accurate positioning of targets in space.This thesis proposes a method to measure the slant range of ground-based real aperture radar targets.It is found in the thesis that the slant range parameter is mainly caused by the positioning error of spherical coordinate system origin composed of target slant range,turntable azimuth angle and turntable pitch angle and the equivalent range measurement error of system delay.For the problem of origin location,because the positioning accuracy of the azimuth angle and elevation angle of the turntable can be controlled by the motor and fed back by the driver,but the rotation center of the antenna is a virtual point in three-dimensional space,it is difficult to obtain its position through other measurement methods.This is a unique problem of the ground-based real aperture radar system,and there is no ready-made origin coordinate estimation method.Therefore,according to the structural characteristics of the S-RAR system,A fitting estimation method of S-RAR station center coordinate system parameters is constructed,and the coordinate origin and coordinate axis vector are determined;There is time delay in the radar RF host,RF link and other systems.The time difference between the transmission timing of the electromagnetic wave and the receipt of the slope reflection echo includes the system time delay.If the system time delay is not corrected,the slant range measurement error will occur.The system time delay problem has not been considered in previous calculations.Therefore,the equivalent range measurement problem of the time delay generated by the above systems is addressed,In this thesis,a radar slope calibration function model based on the plane slope is proposed.Using the origin coordinates,the one-dimensional range information Rm obtained by multiple spatial scans and the target slope range R are iteratively optimized to extract the equivalent range of the delay,and the accuracy can reach0.0001 m.The equivalent range of S-RAR system delay given by this model is 5.6224 m.Due to the influence of weather environment,the system delay will change.Before each system application,this model can be used to extract the equivalent range of delay.By correcting the equivalent range of delay,the measurement accuracy of slope target slant range parameters can be improved.3.How the ground-based real aperture radar imaging target reflects the actual target of the slope is the key to the slope deformation monitoring.Therefore,this thesis proposes a slope surface heterogenous point set registration method based on the iterative nearest neighbor algorithm.In order to intuitively study and judge the degree of slope deformation and improve the reliability of landslide hazard analysis,it is necessary to project the radar deformation monitoring data onto the three-dimensional digital elevation model of the slope.At present,three-dimensional laser is often used to quickly obtain the digital elevation model.Therefore,how to match the three-dimensional image of the ground-based real aperture radar with the three-dimensional point cloud data is the key to achieve projection,which is essentially a problem of heterogeneous data registration.The difficulty lies in different equipment monitoring geometry Data resolution is different.In this innovation,the 3D image of ground-based real aperture radar is regarded as a non-uniform sparse point set.First,the coarse registration of heterogeneous point sets on the slope surface is realized through the spatial transformation of the radar station center coordinate system;Then,by combining the spatial interpolation resampling processing with the iterative nearest neighbor algorithm,the precise registration of heterogeneous point sets on the slope surface is realized.Through the application of slope deformation monitoring in Heidaigou Open pit Coal Mine in Inner Mongolia,it is proved that the root mean square error of the target registration results of more than 94% of the ground real aperture radar 3D images processed by the proposed method is less than one resolution unit. |
PDF Full Text Request