Research On Simulation And Feature Recognition Method Of Earth-rock Dam Disease Based On Stochastic Medium Algorithm

Earth-rock dams have the advantages of low cost and simple structure.They account for a relatively high proportion of water conservancy projects in my country.However,due to their high probability of failure,research on the safety and reliability of earth-rock dams is of great significance.With the development of on-site non-destructive testing of dangerous reservoirs in my country and the construction of a large number of earth-rock dam projects,the application of geological radar method to timely,quickly and accurately find the location of the disease in the earth-rock dam project and determine its type,for the safety inspection of the earth-rock dam It is of great significance.At present,there are two main problems in the application of geological radar in the detection of earth-rock dams.First,the geological radar image recognition is more empirical;at this stage,the geological radar image recognition of earth-rock dam diseases requires long-term experience accumulation,and the forward simulation geoelectric models for auxiliary recognition are mostly established with reference to homogeneous media or layered media.Consider the non-uniformity and dispersion of the discrete material of earth-rock dam.Second,the geological radar image recognition of earth-rock dam diseases has poor timeliness;the existing manual recognition methods have a long recognition cycle and cannot meet the requirements of timely and rapid detection.At the same time,the manual recognition results are missed and misdetected.The probability is higher.Aiming at the above problems,this article aims to realize accurate simulation of earth-rock materials and rapid identification of disease types of earth-rock dams.The main research contents are as follows:(1)Based on the random medium theory,a three-phase discrete random medium model is constructed according to the particle size characteristics and electrical characteristics of different components of the soil-rock mixture.The dielectric properties of the model at a low rate are compared with the homogeneous medium model,and the propagation characteristics of electromagnetic waves in the two models are analyzed.(2)According to different types of earth-rock dam diseases,the causes and characteristics are analyzed,the corresponding three-phase discrete random medium geoelectric model is constructed,the geological radar forward simulation is carried out,and the corresponding forward simulation is studied by using complex signal analysis technology.The simulation results analyze the characteristics of each disease type from the perspective of amplitude,phase,and frequency.(3)Based on the Faster R-CNN algorithm,an earth-rock dam disease target detection model was constructed.The model was trained and tested in the self-made earth-rock dam disease geological radar image data set,and the detection effect of the model was analyzed and tested in engineering examples.The application effect of the model.