A method and system for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau

Abstract

The present invention discloses a method and system for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau. The method comprises the following steps: obtaining resistance information of underground ice in permafrost areas; obtaining spatial distribution information of underground ice in permafrost areas based on the resistance information; and generating a detection image based on the spatial distribution information of underground ice in permafrost areas; acquiring the underground distribution image of the permafrost areas using radar; generating a spatial distribution report of underground ice in the permafrost areas; establishing a dataset based on the spatial distribution report and constructing a permafrost area underground distribution detection model; training and testing the permafrost area underground distribution detection model; and inputting the spatial distribution report of underground ice in the permafrost areas into the trained and tested permafrost area underground distribution detection model to obtain detection results of the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau. The method utilizes an intelligent ground resistance meter and radar device to improve the accuracy of underground ice distribution detection in permafrost areas of the Qinghai-Tibet Plateau. Real-time updates of the distribution report improve the detection accuracy of the model and the accuracy of the detection results.

Description

Technical Field

[0001] The present invention relates to the field of geological detection technology, and in particular to a method and system for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau. Background Art

[0002] China's Qinghai-Tibet Plateau, endowed with unique natural geography and hydrothermal effects, boasts extensive permafrost, making it the world's highest and largest permafrost region at mid- and low-latitudes. The extensive distribution of permafrost on the Qinghai-Tibet Plateau is sensitive to the formation, development, and changes of climate locally, in eastern China, and throughout East Asia, earning it the nickname "warning zone" or "initiator zone" for climate change. Accurately determining the distribution and lower limit of subsurface ice beneath the permafrost on the Qinghai-Tibet Plateau has long been a scientific challenge for cryogenics researchers in my country. Clarifying the distribution of subsurface ice beneath the permafrost on the Qinghai-Tibet Plateau will further our understanding of climate-permafrost interactions and is essential for engineering projects on the plateau. It can provide essential baseline information for climate modeling, cold-region research, planning, and engineering. Summary of the Invention

[0003] In order to solve the problem of low accuracy in detecting the distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau in the above-mentioned prior art, the present invention provides a method and system for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau. The method generates an image of the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau, processes the image through smoothing filtering, and retains important information such as the image contour and edge. At the same time, an intelligent ground resistance meter is used to improve the accuracy of detecting the distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau.

[0004] To achieve the above technical objectives, the present invention provides a method for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau, comprising:

[0005] S1, using a detector to obtain the electrical resistance information of underground ice in permafrost areas;

[0006] S2, obtaining spatial distribution information of underground ice in the permafrost region based on the resistance information, and generating a detection image based on the spatial distribution information of underground ice in the permafrost region;

[0007] S3, using radar to obtain an underground distribution image of the permafrost area;

[0008] S4, generating a spatial distribution report of underground ice in the permafrost area based on the detection image and the underground distribution image;

[0009] S5, establishing a data set based on the spatial distribution report, dividing the data set into a training set and a test set, and constructing a permafrost area underground distribution detection model using a deep learning method;

[0010] S6, training and testing the permafrost area underground distribution detection model based on the training set and the test set;

[0011] S7, inputting the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model of the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0012] Optionally, the calculation formula of the resistance information is:

[0013] R 地 = “Magnification scale” reading × “Measurement scale” reading, where R 地 is the resistance information.

[0014] Optionally, before obtaining the resistance information of underground ice in the permafrost area in S1, parameter calibration is performed on the detector.

[0015] Optionally, the spatial distribution report includes location data of underground ice in the permafrost area and road environment information in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0016] The present invention also provides a system for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau, comprising: a detection module, an image generation module, a first radar detection module, a processing module, a model building module, and a detection module;

[0017] The detection module is used to obtain the resistance information of underground ice in the permafrost area;

[0018] The processing module is used to obtain spatial distribution information of underground ice in the permafrost area based on the resistance information;

[0019] The image generation module is used to generate a detection image based on the spatial distribution information of underground ice in the permafrost area;

[0020] The first radar detection module is used to obtain an underground distribution image of the permafrost area;

[0021] The processing module is further configured to generate a spatial distribution report of underground ice in the permafrost region based on the detection image and the underground distribution image; and to establish a data set based on the spatial distribution report, and to divide the data set into a training set and a test set;

[0022] The model building module is used to build a permafrost area underground distribution detection model, and train and test the permafrost area underground distribution detection model based on the training set and the test set;

[0023] The detection module is used to input the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model in the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0024] Optionally, the calculation formula for the detection module to obtain the resistance information of underground ice in the permafrost area is: R 地 = “Magnification scale” reading × “Measurement scale” reading, where R 地 is the resistance information.

[0025] Optionally, the detection module includes but is not limited to an intelligent ground resistance meter.

[0026] Optionally, the spatial distribution report includes location data of underground ice in the permafrost area and road environment information in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0027] The present invention has the following technical effects:

[0028] Using an intelligent ground resistance meter and radar, the resistance information of subsoil and subsurface ice in permafrost areas is matched to radar-generated detection images, improving the accuracy of ground ice distribution detection in permafrost areas on the Qinghai-Tibet Plateau. The spatial distribution of subsurface ice in permafrost areas of the Qinghai-Tibet Plateau is then mapped and processed using a smoothing filter, preserving important information such as outlines and edges. Real-time updates of distribution reports enhance the accuracy of model detection and improve the accuracy of detection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0030] Figure 1 This is a flowchart of a method for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau according to embodiment 1 of the present invention. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0032] Example 1

[0033] The present invention discloses a system for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau, comprising:

[0034] Detection module, image generation module, first detection radar module, processing module, model building module, detection module, display module, second radar detection module, control module and positioning module;

[0035] The detection module is used to obtain the resistance information of underground ice in the permafrost area;

[0036] The processing module is used to obtain spatial distribution information of underground ice in the permafrost area based on the resistance information;

[0037] The image generation module is used to generate a detection image based on the spatial distribution information of underground ice in the permafrost area;

[0038] The first radar detection module is used to obtain an underground distribution image of the permafrost area;

[0039] The processing module is further configured to generate a spatial distribution report of underground ice in the permafrost region based on the detection image and the underground distribution image; and to establish a data set based on the spatial distribution report, and to divide the data set into a training set and a test set;

[0040] The model construction module is used to construct a permafrost area underground distribution detection model, and train and test the permafrost area underground distribution detection model based on the training set and the test set; it includes: a convolutional neural sub-model, a stacked autoencoder sub-model, a splicing unit and a classification optimization unit, the convolutional neural sub-model is connected to the stacked autoencoder sub-model through the splicing unit; the splicing unit is connected to the classification optimization unit.

[0041] The detection module is used to input the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model in the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0042] The display module is used to display the spatial distribution report of underground ice in the permafrost area in real time;

[0043] The second detection radar module is used to collect road environment information in the permafrost area of ​​the Qinghai-Tibet Plateau;

[0044] The control module is used to collect and record the spatial distribution information of underground ice in the permafrost area;

[0045] The positioning module is used to collect the real-time position of the detection system.

[0046] The display module is also used to display the real-time position of the detection system.

[0047] The detection module includes but is not limited to an intelligent ground resistance meter.

[0048] The control module is also used to record the road environment information in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0049] Example 2

[0050] like Figure 1 As shown, the present invention discloses a method for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau, comprising:

[0051] Calibration of the detector parameters specifically includes:

[0052] Adjust the impedance bridge with an accuracy of 1% in the range of 10.3-10.12, adjust the correction regulator so that the detector can maintain self-equilibrium in the permafrost underground ice environment for at least 0.5 hours, and start the test.

[0053] S1, using a detector to obtain the electrical resistance information of underground ice in permafrost areas;

[0054] Place the detector horizontally and check that the galvanometer pointer is pointing toward the centerline. If not, adjust the zero adjuster to bring the pointer toward the centerline. Set the magnification scale to the maximum setting and slowly turn the generator handle while simultaneously rotating the measuring dial to bring the galvanometer pointer toward the centerline. When the galvanometer pointer is near equilibrium (the pointer is close to the centerline), increase the handle speed to reach a speed of at least 120 rpm and adjust the measuring dial to bring the pointer toward the centerline. If the reading on the measuring dial is too low (less than 1) and difficult to read accurately, the magnification scale is too high. In this case, set the magnification scale to a lower setting and readjust the measuring dial to bring the pointer toward the centerline for an accurate reading. Within the pre-set start and end times of the detection task, calculate the measurement result: resistance information Rground = magnification scale reading × measuring dial reading.

[0055] In this embodiment, the upper soil layer of the permafrost region of the Qinghai-Tibet Plateau is hard and contains ice crystals, i.e., underground ice. The lower soil layer is soft and has melted ice crystals. Deeper soil is sandy. As soil depth increases, the ice crystals decrease and the soil texture begins to soften. Therefore, the internal resistance information of the permafrost region of the Qinghai-Tibet Plateau also gradually changes with depth. During field measurements, in the morning, the ambient temperature is low, the sunlight intensity is weak, the surface temperature is low, and the soil surface is still frozen. At this time, the temperature increases with increasing soil depth. The temperature is higher at deeper soil depths. The situation is different in the afternoon. Due to strong sunlight, the ambient temperature rises, the surface soil temperature also rises, and the surface permafrost begins to melt.

[0056] S2, obtaining spatial distribution information of underground ice in the permafrost region based on the resistance information; generating a detection image based on the spatial distribution information of underground ice in the permafrost region;

[0057] Based on the historical data of the topography and landforms in the permafrost areas of the Qinghai-Tibet Plateau, the permafrost area topography and landform data in the historical data are fitted with a non-uniform rational B-spline model to generate a stratigraphic curve. During the fitting, an inverse algorithm is used to obtain the control vertices to fit the curve surface, and the weight coefficient of each vertex on the surface is limited to 1.

[0058] Extract the upper and lower surface vertex sets of the permafrost region from the stratigraphic curve. Determine a vertex in each of the upper and lower surface vertex sets as the starting point. Adjust the order of the vertices in each vertex set to ensure that the vertices in both sets are arranged in the same direction. Select two vertices from either vertex set, then select another vertex from the other vertex set. Connect the three selected vertices to construct a 3D model of the permafrost region.

[0059] The resistance information of underground ice in the permafrost area obtained in S1 is marked on the three-dimensional model of the permafrost area to obtain the spatial distribution information of underground ice in the permafrost area.

[0060] The process of generating a detection image based on the spatial distribution information of underground ice in the permafrost region specifically includes:

[0061] S301, constructing a real image database based on the obtained spatial distribution information of underground ice in the permafrost region in the three-dimensional model of the permafrost region, where the real image database includes multiple real images;

[0062] S302: Perform smoothing filtering on all real images in the real image database to generate a detection image.

[0063] S3, using radar to obtain an underground distribution image of the permafrost area;

[0064] In this embodiment, a ground-penetrating radar is used to obtain a distribution image of underground ice in the permafrost area. The ground-penetrating radar collects data from the underground of the permafrost area and transmits the data to the receiving antenna of the hub, which converts the data into a digital signal. A pulse sequence is generated based on the control command signal of the ground-penetrating radar, and a transmission signal is generated under the triggering of the pulse sequence and sent through the transmitting antenna. The receiving antenna receives the transmission signal, samples the received signal, and then converts the sampled signal into a low-speed, low-voltage pulse signal. The A / D converter in the ground-penetrating radar packages the obtained data, and then converts the packaged data into an electrical pulse signal, modulates the electrical pulse signal, and sends the electrical pulse signal through the receiving antenna. The processing module receives the signal and displays the underground data of the permafrost area. The operator judges the distribution of objects underground in the permafrost area based on the waveform and its color, and obtains an underground distribution image of the permafrost area.

[0065] S4, generating a spatial distribution report of underground ice in the permafrost area based on the detection image and the underground distribution image;

[0066] The control module matches the detection image obtained by S2 and the underground distribution image of the permafrost area obtained by S3 in the constructed three-dimensional model of the permafrost area, and matches the resistance values ​​of the underground soil, underground ice, and sand in the permafrost area obtained by the resistance detector to the distribution image to obtain a spatial distribution image.

[0067] The detection radar module collects road surface environment information in the permafrost zone of the Qinghai-Tibet Plateau, the control module records the road surface environment information in the permafrost zone of the Qinghai-Tibet Plateau collected by the detection radar module, and the positioning module is used to collect the real-time position of the detection system. Furthermore, based on the road surface environment information, the real-time position and the spatial distribution image, a spatial distribution report of underground ice in the permafrost zone is generated.

[0068] S5, establishing a data set based on the spatial distribution report, dividing the data set into a training set and a test set, and constructing a permafrost area underground distribution detection model using a deep learning method;

[0069] A dataset was established based on the spatial distribution report, with 80% of the dataset used as a training set and the remaining 20% ​​as a test set;

[0070] The convolutional neural sub-model consists of ten layers: five convolutional layers, three pooling layers, one dropout layer, and three softmax output layers. The first, second, and third convolutional layers have kernel sizes of 5×5, 5×5, and 6×6, respectively, corresponding to 16, 32, and 64 kernels. The first three convolutional layers are each followed by a pooling layer using max pooling with a pooling size of 2 and a stride of 2. The fourth convolutional layer has 128 kernels of 5×5 size, followed by a dropout layer. The fifth convolutional layer has five kernels of 3×3 size, outputting five neurons, and finally a softmax layer to generate the classification probabilities. Reinforced Luminance (ReLU) activation function is used, with a stride of 1 and no edge augmentation. To reduce network parameters, no fully connected layers are used, and a dropout operation is added after the penultimate convolutional layer.

[0071] The stacked autoencoder sub-model includes a dimensionality reduction unit and a feature extraction unit. The dimensionality reduction unit uses principal component analysis (PCA) to reduce the dimensionality of the input sample image. PCA is a mathematical transformation and dimensionality reduction method that uses a set of feature vectors with fewer dimensions to maximize the information in the original data. In this embodiment, the input sample image is reduced in dimensionality using PCA, the resulting two-dimensional reduced matrix is ​​stretched into a one-dimensional vector, and then the feature extraction unit uses an autoencoder to extract features from the reduced dimensionality sample image.

[0072] Furthermore, the splicing unit uses the Concat method to fuse the feature vector output by the convolutional neural sub-model with the feature vector of the stacked autoencoder sub-model to obtain the underground distribution detection model of the permafrost area.

[0073] S6, training and testing the permafrost area underground distribution detection model based on the training set and the test set;

[0074] The classification optimization unit includes a softmax classifier and an Adam optimizer. The softmax classifier calculates the cross-entropy loss based on the fusion results. The Adam optimizer trains the permafrost area underground distribution detection model based on the cross-entropy loss, obtains model parameters, and then loads and constructs the permafrost area underground distribution detection model based on the model parameters. The permafrost area underground distribution detection model is trained using the training dataset. After the training is completed, the permafrost area underground distribution detection model is tested.

[0075] S7, inputting the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model of the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

[0076] The tested underground distribution detection model in permafrost areas is directly used to detect the spatial distribution report of underground ice in permafrost areas. During the training and testing of the model, the latest image data obtained is continuously added to the dataset to expand the number of datasets, thereby ensuring the validity and continuity of the data, continuously optimizing the model training, and improving the model output accuracy.

[0077] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting the spatial distribution of underground ice in permafrost areas of the Qinghai-Tibet Plateau, characterized in that: include: S1, using a detector to obtain the electrical resistance information of underground ice in permafrost areas; S2, obtaining spatial distribution information of underground ice in the permafrost region based on the resistance information, and generating a detection image based on the spatial distribution information of underground ice in the permafrost region; S3, using radar to obtain an underground distribution image of the permafrost area; S4, generating a spatial distribution report of underground ice in the permafrost area based on the detection image and the underground distribution image; S5, establishing a data set based on the spatial distribution report, dividing the data set into a training set and a test set, and constructing a permafrost area underground distribution detection model using a deep learning method; S6, training and testing the permafrost area underground distribution detection model based on the training set and the test set; S7, inputting the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model of the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

2. The method for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau according to claim 1, characterized in that: The calculation formula of the resistance information is: R 地 = "Magnification scale" reading × "Measurement scale" reading, where R 地 is the resistance information.

3. The method for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau according to claim 1, characterized in that: Before obtaining the resistance information of underground ice in the permafrost area in S1, parameter calibration is performed on the detector.

4. The method for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau according to claim 1, characterized in that: The spatial distribution report includes location data of underground ice in the permafrost zone and road surface environment information in the permafrost zone of the Qinghai-Tibet Plateau.

5. A system for detecting the spatial distribution of underground ice in the permafrost region of the Qinghai-Tibet Plateau, characterized in that: include: a detection module, an image generation module, a first radar detection module, a processing module, a model building module, and a detection module; The detection module is used to obtain the resistance information of underground ice in the permafrost area; The processing module is used to obtain spatial distribution information of underground ice in the permafrost area based on the resistance information; The image generation module is used to generate a detection image based on the spatial distribution information of underground ice in the permafrost area; The first radar detection module is used to obtain an underground distribution image of the permafrost area; The processing module is further configured to generate a spatial distribution report of underground ice in the permafrost region based on the detection image and the underground distribution image; and to establish a data set based on the spatial distribution report, and to divide the data set into a training set and a test set; The model building module is used to build a permafrost area underground distribution detection model, and train and test the permafrost area underground distribution detection model based on the training set and the test set; The detection module is used to input the spatial distribution report of underground ice in the permafrost area into the trained and tested underground distribution detection model in the permafrost area to obtain the spatial distribution detection result of underground ice in the permafrost area of ​​the Qinghai-Tibet Plateau.

6. The underground ice spatial distribution detection system in the permafrost region of the Qinghai-Tibet Plateau according to claim 5 is characterized in that: The calculation formula for the resistance information of underground ice in the permafrost area obtained by the detection module is: 地 = "Magnification scale" reading × "Measurement scale" reading, where R 地 is the resistance information.

7. The underground ice spatial distribution detection system in the permafrost region of the Qinghai-Tibet Plateau according to claim 5 is characterized in that: The detection module includes but is not limited to an intelligent ground resistance meter.

8. The underground ice spatial distribution detection system in the permafrost region of the Qinghai-Tibet Plateau according to claim 5 is characterized in that: The spatial distribution report includes location data of underground ice in the permafrost zone and road surface environment information in the permafrost zone of the Qinghai-Tibet Plateau.

Images