Rainfall recharge control method and device

Abstract

The invention relates to the technical field of foundation pit dewatering, in particular to a dewatering recharge control method and device.The method comprises the steps that a first control instruction is issued to edge control equipment according to a preset frequency, so that the edge control equipment controls imaging equipment to shoot an underground image based on the first control instruction; performing identification processing on the underground image by using an image identification model to extract water level information corresponding to the underground image; the water level information corresponding to two adjacent underground images is compared, and the real-time precipitation rate is determined; and under the condition that the real-time precipitation rate is larger than the first rate threshold value, a second control instruction is issued to the edge control equipment, so that the edge control equipment adjusts the operation frequencies of the precipitation pump and the recharge pump based on the second control instruction. According to the method, the visual features of the underground image can be extracted based on the image recognition model, control is carried out based on the visual features, the control process is not affected by the underground complex environment, and the accuracy of water level monitoring and adjusting can be effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of foundation pit dewatering, and in particular to a dewatering recharge control method and device. Background Art

[0002] In modern urban construction, underground space development, and geological engineering, precipitation recharge technology is a key technology for ensuring the smooth progress of projects and the rational utilization of groundwater resources. During the precipitation recharge process, accurate and real-time monitoring of water level changes within precipitation wells is crucial for properly regulating the operation of precipitation and recharge pumps, maintaining groundwater stability, and preventing precipitation-induced land subsidence and groundwater resource imbalances.

[0003] Currently, monitoring and control of precipitation and recharge systems primarily relies on manual labor or sensor equipment. Manual monitoring requires manually monitoring water levels and adjusting the operating frequency of precipitation and recharge pumps based on experience. This is not only inefficient, but also has limited measurement frequency and accuracy, making it difficult to meet the needs of real-time monitoring.

[0004] Sensor-based monitoring solutions require the installation of multiple water level and flow sensors within the well. These sensors collect real-time water level and flow data, transmit this data to the control system, and then adjust the operating frequency of the dewatering and recharge pumps. However, the underground environment is complex, and sensors are susceptible to environmental factors such as silt clogging and corrosion, resulting in inaccurate data collection or transmission interruptions, which in turn hinders monitoring and control, resulting in poor results. Summary of the Invention

[0005] The embodiments of the present application provide a precipitation recharge control method and device to solve the problem of inaccurate monitoring and control caused by environmental influences in traditional control methods.

[0006] In a first aspect, an embodiment of the present application provides a precipitation and recharge control method, which is applied to a cloud processing platform, and the method includes: during the operation of the precipitation pump, issuing a first control instruction to the edge control device at a preset frequency, so that the edge control device controls the imaging device to capture downhole images based on the first control instruction, and the downhole image includes at least a background area corresponding to the well wall in the precipitation well and a water level area corresponding to the groundwater; using an image recognition model to identify and process the downhole image to extract water level information corresponding to the downhole image; comparing the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate; when the real-time precipitation rate is greater than the first rate threshold, issuing a second control instruction to the edge control device, so that the edge control device adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and issuing a third control instruction to the edge control device, so that the edge control device magnifies the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0007] In one possible implementation, an image recognition model is used to identify and process downhole images to extract water level information corresponding to the downhole images, including: determining target model parameters based on the corresponding imaging focal length when the downhole image is taken; wherein the image recognition model includes multiple groups of preset model parameters, different groups of preset model parameters correspond to different imaging focal lengths, and the target model parameters are one group of the multiple groups of preset model parameters; the downhole image is identified and processed based on the image recognition model and the target model parameters to determine the edge contour line of the water level area; the water level information is determined based on the edge contour line, and the water level information includes at least target distance, target length and target size, the target distance is the distance between at least one feature point on the edge contour line and the boundary of the downhole image, the target length is the length of the edge contour line, and the target size is the pixel size occupied by the area surrounded by the edge contour line.

[0008] In one possible implementation, water level information corresponding to two adjacent downhole images is compared to determine a real-time precipitation rate, including: constructing a feature vector based on target distance, target length, and target size for the two adjacent downhole images; calculating the Euclidean distance between the feature vectors of the two adjacent downhole images; inputting the Euclidean distance into a pre-trained regression model to obtain a water level change; and calculating the real-time precipitation rate based on the water level change and the timestamps of when the two adjacent downhole images were taken.

[0009] In one possible implementation, before calculating the Euclidean distance between the feature vectors of two adjacent downhole images, the method further includes: when the imaging focal lengths corresponding to the two adjacent downhole images are different, calculating a magnification ratio based on the imaging focal length corresponding to the first downhole image and the imaging focal length corresponding to the second downhole image; wherein the first downhole image is the first of the two adjacent downhole images, and the second downhole image is the second of the two adjacent downhole images; and adjusting the feature vector corresponding to the second downhole image based on the magnification ratio.

[0010] In one possible implementation, before issuing the second control instruction to the edge control device, the method further includes: when the real-time precipitation rate is greater than the first rate threshold, aggregating the real-time precipitation rates and performing curve fitting to obtain a precipitation rate fitting curve; obtaining multiple historical precipitation rate curves, where the historical precipitation rate curves are generated based on target precipitation rate data stored in a cloud database, where the target precipitation rate data is data generated when a precipitation operation is performed on an area that matches the geological characteristics of the area where the precipitation well is located; determining a target precipitation rate for the precipitation well based on the historical precipitation rate curves; substituting the target precipitation rate into a frequency formula to obtain a first target operating frequency of the precipitation pump; determining a second target operating frequency of the recharge pump based on the first target operating frequency; and generating a second control instruction based on the first target operating frequency and the second target operating frequency; the frequency formula is: ; in, Indicates the first target operating frequency, Indicates the current operating frequency of the precipitation pump. is the target precipitation rate, is the real-time precipitation rate, is the adjustment coefficient, 0< <1.

[0011] In one possible implementation, the target precipitation rate of the precipitation well is determined based on the historical precipitation rate curve, including: calculating the similarity between the precipitation rate fitting curve and each historical precipitation rate curve, and determining the historical precipitation rate curve corresponding to the maximum similarity as the target rate curve; time-aligning the precipitation rate fitting curve and the target rate curve to determine the first coordinate point corresponding to the real-time precipitation rate in the target rate curve; in the target rate curve, determining the second coordinate point corresponding to the first coordinate point after a preset time span, and determining the precipitation rate corresponding to the second coordinate point as the target precipitation rate, where the preset time span is determined based on a preset frequency.

[0012] In one possible implementation, the method further includes: performing curve fitting on the operating frequency of the dewatering pump to obtain the dewatering pump operating curve; performing curve fitting on the operating frequency of the recharge pump to obtain the recharge pump operating curve; sending a display instruction to the display terminal so that the display terminal displays the downhole image, the dewatering rate fitting curve, the dewatering pump operating curve and / or the recharge pump operating curve based on the display instruction; the display terminal is a digital smart large screen, a computer terminal and / or a mobile device.

[0013] In one possible implementation, after comparing the water level information corresponding to two adjacent downhole images and determining the real-time precipitation rate, it also includes: when the real-time precipitation rate is greater than the first rate threshold and less than the second rate threshold, determining that M is equal to the first amplification coefficient; when the real-time precipitation rate is greater than or equal to the second rate threshold, determining that M is equal to the second amplification coefficient; wherein the second amplification coefficient is greater than the first amplification coefficient.

[0014] In second aspect, an embodiment of the present application provides a precipitation and recharge control method, which is applied to an edge control device, and the method includes: responding to a first control instruction issued by a cloud processing platform, controlling an imaging device to capture a downhole image based on the first control instruction, wherein the downhole image includes at least a background area corresponding to the well wall in the precipitation well and a water level area corresponding to the groundwater; responding to a second control instruction issued by the cloud processing platform, adjusting the operating frequency of the precipitation pump and the recharge pump based on the second control instruction, and responding to a third control instruction issued by the cloud processing platform, magnifying the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0015] In a third aspect, an embodiment of the present application provides a precipitation and recharge control device, which includes: an acquisition module, configured to: during the operation of the precipitation pump, send a first control instruction to the edge control device at a preset frequency, so that the edge control device controls the imaging device to capture downhole images based on the first control instruction, and the downhole image includes at least a background area corresponding to the well wall in the precipitation well and a water level area corresponding to the groundwater; an identification module, configured to: use an image recognition model to identify and process the downhole image to extract water level information corresponding to the downhole image; a comparison module, configured to: compare the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate; an adjustment module, configured to: when the real-time precipitation rate is greater than the first rate threshold, send a second control instruction to the edge control device, so that the edge control device adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and, send a third control instruction to the edge control device, so that the edge control device magnifies the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0016] From the above content, it can be seen that the embodiment of the present application provides a precipitation and recharge control method and device, which can be applied to a cloud processing platform, and specifically may include: during the operation of the precipitation pump, issuing a first control instruction to the edge control device at a preset frequency, so that the edge control device controls the imaging device to capture downhole images based on the first control instruction, and the downhole image includes at least the background area corresponding to the well wall in the precipitation well and the water level area corresponding to the groundwater; using the image recognition model to identify and process the downhole image to extract the water level information corresponding to the downhole image; comparing the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate; when the real-time precipitation rate is greater than the first rate threshold, issuing a second control instruction to the edge control device, so that the edge control device adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and issuing a third control instruction to the edge control device, so that the edge control device magnifies the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate. From the above, it can be seen that compared with the previous single perception method that relies on static sensors, the embodiment of the present application can transform static water level data monitoring into dynamic visual perception by continuously shooting downhole images. By analyzing the water level area in the downhole image through the image recognition model, not only can the water level information be obtained in real time, but the real-time precipitation rate can also be determined by comparing adjacent images, thereby realizing dynamic monitoring of the water level change trend. In addition, the method provided by the embodiment of the present application can also make full use of the significant difference in optical reflection characteristics between water and the well wall, and fully extract visual features based on the image recognition model. In this way, the monitoring means provided by the embodiment of the present application is not affected by the complex downhole environment and can effectively improve the accuracy of water level monitoring and adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of the structure of a precipitation recharge control system provided in an embodiment of the present application; Figure 2 A schematic flow chart of a first precipitation recharge control method provided in an embodiment of the present application; Figure 3 A schematic diagram of a process for determining water level information provided in an embodiment of the present application; Figure 4 A schematic diagram of an edge contour line provided in an embodiment of the present application; Figure 5 A schematic diagram of a flow chart for determining the operating frequency of a dewatering pump and a recharging pump provided in an embodiment of the present application; Figure 6 A schematic flow chart of a second precipitation recharge control method provided in an embodiment of the present application; Figure 7 A schematic structural diagram of a first precipitation recharge control device provided in an embodiment of the present application; Figure 8 This is a schematic structural diagram of the second precipitation recharge control device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0019] Figure 1 This is a schematic diagram of the structure of the precipitation recharge control system provided in an embodiment of the present application.

[0020] like Figure 1 As shown, an embodiment of the present application provides a precipitation recharge control system, which may include a cloud processing platform 100, an edge control device 200, a precipitation device 300, a recharge device 400, an imaging device 500 and a display terminal 600.

[0021] Among them, the dewatering device 300 is located in an area where dewatering operations are required, such as a foundation pit construction site, an underground engineering construction area, etc. The dewatering device 300 may include dewatering wells, dewatering pumps, drainage pipes and other components, which are used to extract and discharge groundwater in the operating area to meet the needs of engineering construction for groundwater level control.

[0022] The location of the recharge device 400 can be determined based on recharge requirements and geological conditions, specifically in an appropriate groundwater recharge area. The recharge device 400 can include components such as a recharge well, a recharge pump, and a drainage pipe. Its primary function is to reinject treated water that meets recharge standards back into the ground, maintaining groundwater level equilibrium and preventing problems such as land subsidence and groundwater depletion caused by excessive rainfall.

[0023] The edge control devices 200 are distributed near the area where the precipitation devices 300 and / or the recharge devices 400 are located, and are used to control the precipitation devices 300 and / or the recharge devices 400 in real time. In some implementations, the number of edge control devices 200 is not limited to one, and can be flexibly configured based on the distribution of the precipitation devices 300 and the recharge devices 400 and the requirements for control accuracy. For example, in a large-scale foundation pit precipitation and recharge project, the precipitation devices 300 and the recharge devices 400 can be partitioned, and an edge control device 200 can be configured for the precipitation devices 300 and the recharge devices 400 in a partition. In other words, each edge control device 200 is responsible for controlling the precipitation or recharge devices within a certain range to achieve refined regional control, or an edge control device 200 can be configured for each precipitation device 300 and recharge device 400.

[0024] The cloud processing platform 100 can interact with the edge control device 200 through a preset communication method, receive monitoring data transmitted from the edge control device 200, and then use big data analysis, artificial intelligence algorithms and other technologies to deeply process and analyze the data, formulate a more optimized overall precipitation recharge control plan, and send control instructions to the edge control device 200 to achieve intelligent and precise management of the entire precipitation recharge system, ensuring that the precipitation and recharge process can not only meet the needs of engineering construction, but also effectively protect the surrounding water environment and geological environment.

[0025] The imaging device 500 can be placed at the wellhead or inside the dewatering well. Specifically, each dewatering well can be equipped with an imaging device 500 to capture images of the well interior and obtain downhole images. Furthermore, the imaging device 500 can communicate with the edge control device 200 to take photos in response to control by the edge control device 200 and transmit the downhole images to the edge control device 200. The imaging device 500 can be an infrared camera, a low-light camera, and / or an industrial camera.

[0026] The display terminal 600 is used to display the real-time operating status, monitoring data and control instruction execution status of the precipitation device 300 and the recharge device 400, providing an intuitive data display and operation interface for on-site staff and remote operators.

[0027] Furthermore, in order to achieve precise control of precipitation recharge, an embodiment of the present application provides a precipitation recharge control method, and the cloud processing platform 100 can be used to execute the precipitation recharge control method.

[0028] Figure 2 A flow chart of the first precipitation recharge control method provided in an embodiment of the present application.

[0029] like Figure 2 As shown, the precipitation recharge control method provided in the embodiment of the present application may include the following steps S100-S400.

[0030] S100: During the operation of the dewatering pump, a first control instruction is sent to the edge control device 200 at a preset frequency, so that the edge control device 200 controls the imaging device 500 to capture downhole images based on the first control instruction. The downhole images include at least the background area corresponding to the well wall in the dewatering well and the water level area corresponding to the groundwater.

[0031] During the operation of the dewatering pump, to ensure efficient operation of the groundwater dewatering and recharge process and accurate monitoring of the water level, the embodiment of the present application can issue a first control instruction to the edge control device 200 at a preset frequency. The preset frequency can be 10 minutes / time, 15 minutes / time, 30 minutes / time, or 1 hour / time, and can be flexibly set according to actual project needs. For example, in the early stages of dewatering, a higher instruction issuance frequency can be set to timely grasp the underground dynamics; during the stable dewatering stage, the frequency can be appropriately reduced to reduce data processing pressure.

[0032] After receiving the first control command, the edge control device 200 controls the imaging device 500 installed at the wellhead or inside the well to capture downhole images based on the command. The imaging device 500's installation position and angle must be precisely designed to ensure that the captured downhole images include at least the background area corresponding to the well wall and the area corresponding to the groundwater level. By using the well wall as a stable background reference, water level fluctuations can be more accurately identified.

[0033] In some implementations, before step S100, step S100a may be further included: issuing a fourth control instruction to the edge control device 200, so that the edge control device 200 activates the precipitation pump, the recirculation pump, and the imaging device 500 based on the fourth control instruction. Furthermore, the fourth control instruction may include startup configuration information, which includes at least the initial frequency of the precipitation pump, the initial frequency of the recirculation pump, and the initial imaging focal length of the imaging device 500, so that the precipitation pump, the recirculation pump, and the imaging device 500 are activated based on the startup configuration information.

[0034] S200: Using an image recognition model to perform recognition processing on the downhole image to extract water level information corresponding to the downhole image.

[0035] Among them, the image recognition model can be a machine learning model for processing image data. The image recognition model provided in the embodiment of the present application can be built based on an infrastructure such as a U-Net architecture, and the image recognition model is trained with a large amount of annotated downhole image data. After pre-training, the image recognition model can be used to extract key features of downhole images, such as the edge of the water surface and the well wall, the texture of the downhole water body, and the shape of the well wall. These features can be used to determine water level information, thereby achieving accurate water level prediction. The specific composition of the water level information will be described in detail below and will not be repeated here.

[0036] It can be understood that the water surface has a mixed characteristic of specular reflection and diffuse reflection, which will produce dynamic light spots and reflections as the water level fluctuates, while the reflection of the well wall material (such as concrete and metal) is relatively stable and presents a fixed texture. Based on this, the embodiment of the present application can provide rich visual features for the image recognition model through downhole image acquisition operations.

[0037] S300: Compare water level information corresponding to two adjacent downhole images to determine a real-time precipitation rate.

[0038] It can be understood that as the dewatering pump works, the underground water level gradually decreases, and the water level information corresponding to the two adjacent underground images will also change. Therefore, the embodiment of the present application can compare the water level information corresponding to the two adjacent underground images to determine the real-time dewatering rate. The real-time dewatering rate can intuitively and accurately reflect the drainage efficiency of the dewatering pump under the current working state, and can determine the dynamic change trend of the underground water level, providing key data support for subsequent control decisions.

[0039] S400: When the real-time precipitation rate is greater than the first rate threshold, a second control instruction is issued to the edge control device 200, so that the edge control device 200 adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and a third control instruction is issued to the edge control device 200, so that the edge control device 200 magnifies the imaging focal length of the imaging device 500 by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0040] If the real-time precipitation rate exceeds the first rate threshold, it indicates that the water level is dropping relatively quickly. Therefore, if the imaging device 500 is fixed, the imaging focal length of the imaging device 500 can be magnified by M times to ensure that the imaging device 500 can clearly capture downhole images. Furthermore, the magnification factor M of the imaging focal length can be dynamically adjusted based on the real-time precipitation rate. For example, when the precipitation rate is high, the magnification factor M of the imaging focal length is correspondingly increased to improve the imaging effect of the imaging device 500.

[0041] It is understandable that the operating frequency of the dewatering pump will directly affect the size of the real-time dewatering rate. The faster the operating frequency of the dewatering pump, the greater the real-time dewatering rate in the dewatering well. The slower the operating frequency of the dewatering pump, the smaller the real-time dewatering rate in the dewatering well. The operating efficiency of the recharge pump will also directly affect the real-time recharge efficiency. The faster the operating frequency of the recharge pump, the greater the real-time recharge rate in the recharge well. Therefore, in order to avoid potential risks caused by too fast or too slow dewatering rates, the embodiment of the present application can adjust the operating frequency of the dewatering pump and the recharge pump in real time based on the second control instruction, so that the operating frequency of the dewatering pump and the recharge pump can adapt to water level changes. During adjustment, the operating frequency of the dewatering pump and the recharge pump can be specifically reduced to avoid safety problems caused by too fast dewatering. The steps for determining the operating frequency of the dewatering pump and the recharge pump will be described in detail below and will not be repeated here.

[0042] It should also be noted that the specific value of the first rate threshold can be determined by comprehensively considering factors such as the power of the dewatering pump, the downhole volume, and the permeability of the formation. In some areas with complex geological conditions and poor permeability, in order to avoid the risk of rapid precipitation causing formation collapse, the first rate threshold can be set to a lower value, such as 30 cm per hour; in scenarios with good permeability and high construction requirements, in order to speed up the dewatering process, the threshold can be appropriately increased to 50 cm per hour. In addition, historical precipitation data can be referred to, combined with construction safety requirements and equipment operating performance, and through hydrogeological analysis and simulation calculations, the first rate threshold that meets actual working conditions can be accurately set. This can not only ensure the efficient implementation of dewatering operations, but also ensure the safety and stability of the entire construction process.

[0043] In some implementations, the cloud processing platform 100 and the edge control device 200 can communicate based on 4G / 5G wireless communication technology to ensure the real-time and stability of data transmission, or they can communicate based on narrowband Internet of Things (NB-IoT) technology or long range (LoRa) communication technology. The embodiments of the present application do not specifically limit this.

[0044] From the above content, it can be seen that the embodiment of the present application provides a precipitation and recharge control method, which can be applied to the cloud processing platform 100, and specifically may include: during the operation of the precipitation pump, issuing a first control instruction to the edge control device 200 at a preset frequency, so that the edge control device 200 controls the imaging device 500 to capture a downhole image based on the first control instruction, and the downhole image includes at least a background area corresponding to the well wall in the precipitation well and a water level area corresponding to the groundwater; using an image recognition model to identify and process the downhole image to extract the water level information corresponding to the downhole image; comparing the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate; when the real-time precipitation rate is greater than the first rate threshold, issuing a second control instruction to the edge control device 200, so that the edge control device 200 adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and issuing a third control instruction to the edge control device 200, so that the edge control device 200 magnifies the imaging focal length of the imaging device 500 by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate. From the above, it can be seen that compared with the previous single perception method that relies on static sensors, the embodiment of the present application can transform static water level data monitoring into dynamic visual perception by continuously shooting downhole images. By analyzing the water level area in the downhole image through the image recognition model, not only can the water level information be obtained in real time, but the real-time precipitation rate can also be determined by comparing adjacent images, thereby realizing dynamic monitoring of the water level change trend. In addition, the method provided by the embodiment of the present application can also make full use of the significant difference in optical reflection characteristics between water and the well wall, and fully extract visual features based on the image recognition model. In this way, the monitoring means provided by the embodiment of the present application is not affected by the complex downhole environment and can effectively improve the accuracy of water level monitoring and adjustment.

[0045] The steps for determining water level information are described in detail below with reference to the accompanying drawings.

[0046] Figure 3 A schematic diagram of the process of determining water level information provided in an embodiment of the present application.

[0047] like Figure 3 As shown, step S200 may include the following steps S201-S203.

[0048] S201: Determine target model parameters based on the corresponding imaging focal length when capturing downhole images; wherein the image recognition model includes multiple groups of preset model parameters, different groups of preset model parameters correspond to different imaging focal lengths, and the target model parameters are one group among the multiple groups of preset model parameters.

[0049] The imaging focal length may refer to the distance from the optical center of the lens of the imaging device 500 to the imaging plane, where the imaging plane is, for example, the plane where the photosensitive element is located. It is understood that the imaging focal length of the imaging device 500 is adjustable.

[0050] The preset model parameters are determined during the model training phase. Specifically, the image recognition model includes multiple sets of preset model parameters, each for a different focal length. The target model parameters are then selected from these preset model parameters to match the current imaging focal length. This process ensures that the image recognition model can accurately extract and analyze water level areas under various shooting conditions.

[0051] During the training phase, in order to enable the image recognition model to adapt to downhole images at different imaging focal lengths, the embodiment of the present application can obtain a large amount of downhole image data taken at different focal lengths and annotate the images, and the annotation content includes at least key information such as the edge contour line of the water level area, the target distance, the target length and the target size. Afterwards, these annotated image data can be input into the image recognition model for training. During the training process, the model can learn the feature representation of the water level area at different focal lengths, and learn how to accurately extract these key water level information from the image. By continuously adjusting the parameters of the model, the difference between the predicted value and the actual annotation value can be minimized. When the training is completed, the model can generate multiple sets of preset model parameters, each set of parameters corresponding to a specific imaging focal length. In this way, no matter how the imaging focal length of the imaging device 500 is adjusted, the image recognition model can maintain a high accuracy in extracting water level information, thereby providing reliable data support for precipitation recharge control.

[0052] S202: performing recognition processing on the downhole image based on the image recognition model and the target model parameters to determine the edge contour line of the water level area.

[0053] In the embodiment of the present application, the edge contour line can clearly reflect the boundary of the water level area in the downhole image. Figure 4 Schematic diagram of downhole image is shown, including background area A, water level area B and edge contour line C. Figure 4 This is a schematic diagram of an ideal state in which the background area corresponding to the well wall and the water level area corresponding to the groundwater are clearly demarcated and the water level area has a regular shape. The shape of the edge contour line in the downhole image depends on the actual precipitation situation in the downhole.

[0054] S203: Determine water level information based on the edge contour line.

[0055] Water level information includes at least target distance, target length, and target size. The target distance is the distance between at least one feature point on the edge contour line and the boundary of the downhole image; this distance reflects the position of the water level relative to the wellbore wall. The target length is the length of the edge contour line and can be used to determine the extent of the water level area. The target size is the pixel size of the area enclosed by the edge contour line and can be used to estimate the actual area of the water level area.

[0056] In an embodiment of the present application, the image recognition model can identify feature points on the edge contour line based on the Harris corner detection algorithm, and the number of feature points is not limited to one.

[0057] Furthermore, step S300 may include the following steps S301-S306.

[0058] S301: For two adjacent downhole images, construct a feature vector based on target distance, target length and target size.

[0059] In this embodiment, a feature vector can be constructed for each downhole image. A feature vector is a multidimensional data structure, with each dimension corresponding to a specific water level feature. For example, a feature vector can be represented as [target distance, target length, target size]. The specific dimensions of the feature vector can be adjusted based on actual conditions and are not specifically limited in this embodiment.

[0060] S302: When the imaging focal lengths corresponding to two adjacent downhole images are different, the magnification ratio is calculated based on the imaging focal length corresponding to the first downhole image and the imaging focal length corresponding to the second downhole image; wherein the first downhole image is the first one of the two adjacent downhole images, and the second downhole image is the second one of the two adjacent downhole images.

[0061] It is understood that since the imaging focal length of the imaging device 500 is adjusted according to actual shooting requirements, the imaging ratios of two adjacent downhole images may be different, which will directly affect the imaging area of the water level area in the downhole image. In this case, the embodiment of the present application can calculate a magnification ratio to adjust the feature vector. The magnification ratio can be determined by comparing the imaging focal lengths of the two images.

[0062] The specific steps for determining the magnification ratio may be: first, clarify the imaging focal lengths corresponding to the first downhole image and the second downhole image. For example, the imaging focal length of the first downhole image is M1, and the imaging focal length of the second downhole image is M2. Then, the magnification ratio calculation formula may be k=M1 / M2, where k represents the magnification ratio.

[0063] S303: Adjusting the feature vector corresponding to the second downhole image based on the magnification ratio.

[0064] Exemplarily, the feature vector corresponding to the second downhole image is ,in, represents the target distance corresponding to the second downhole image, represents the target length corresponding to the second downhole image, represents the target size corresponding to the second downhole image. Then, the adjusted target distance can be , the adjusted target length can be , the adjusted target size can be , the adjusted eigenvector can be .

[0065] In this way, the influence of imaging focal length change on the eigenvector can be eliminated.

[0066] S304: Calculate the Euclidean distance between the feature vectors of two adjacent downhole images.

[0067] For example, the feature vector corresponding to the first downhole image can be ,in, represents the target distance corresponding to the first downhole image, represents the target length corresponding to the first downhole image, Indicates the target size corresponding to the first downhole image.

[0068] The Euclidean distance calculation formula can be , where D represents the Euclidean distance, represents the target distance corresponding to the first downhole image, represents the target length corresponding to the first downhole image, represents the target size corresponding to the first downhole image, represents the adjusted target distance corresponding to the second downhole image, represents the adjusted target length corresponding to the second downhole image, Indicates the adjusted target size corresponding to the second downhole image.

[0069] S305: Input the Euclidean distance into the pre-trained regression model to obtain the water level change.

[0070] The pre-trained regression model is a machine learning model obtained by training based on historical data and labeled samples. Specifically, it can be a linear regression model. Its core goal is to establish a mapping relationship between the Euclidean distance and the actual water level change. In this embodiment, the pre-trained regression model training process includes: collecting a large number of historical downhole images and manually labeling the actual water level value corresponding to each image. Afterwards, the feature vectors of the historical image pairs are calculated and adjusted to obtain the Euclidean distance as the input feature. Furthermore, the regression model is trained using algorithms such as gradient descent and random forest to minimize the error between the predicted water level change and the labeled value. Finally, the model parameters are optimized through cross-validation, regularization and other methods to improve the generalization ability.

[0071] Furthermore, the pre-trained regression model can be deployed in the cloud processing platform 100 and incremental training can be performed regularly using newly collected data to adapt to the impact of changes in geological conditions or equipment parameter drift.

[0072] S306: Calculate the real-time precipitation rate based on the water level change and the timestamps of the two adjacent downhole images.

[0073] In the embodiment of the present application, the calculation formula for the real-time precipitation rate can be: .

[0074] in, Indicates the real-time precipitation rate, Indicates the change in water level, the unit can be meter. Indicates the time interval between two adjacent downhole images, which can be calculated from the timestamp.

[0075] Through the above steps, the real-time precipitation rate can be accurately determined, providing a reliable basis for the effective control of precipitation recharge.

[0076] It should also be noted that due to the field of view limitation of the imaging device 500, water level fluctuations, obstruction of the drain pipe or other interference factors, the edge contour line of the water level area in the actual downhole image may be discontinuous. This discontinuity may cause the extracted water level information to be inaccurate, thereby affecting the construction of the feature vector and the subsequent calculation of the precipitation rate. After acquiring the downhole image, the embodiment of the present application can first perform edge detection on the downhole image to identify whether there are discontinuous areas in the water level area of the image. If there are discontinuous areas, these discontinuous areas can be supplemented and optimized by image restoration technology to ensure that the edge contour of the water level area can be fully presented. Specifically, the image interpolation method can be used to estimate the missing edges, and the complete edge contour can be restored in combination with the features of the area surrounding the discontinuous area. The embodiment of the present application does not make specific limitations on this.

[0077] The following is a detailed introduction to the steps for determining the operating frequency of the dewatering pump and the recharge pump with reference to the accompanying drawings.

[0078] Figure 5 A schematic diagram of a flow chart for determining the operating frequency of a dewatering pump and a recharge pump provided in an embodiment of the present application.

[0079] like Figure 5 As shown, in the method provided in the embodiment of the present application, the following steps S501-S506 may be included before step S400.

[0080] S501: When the real-time precipitation rate is greater than a first rate threshold, aggregate the real-time precipitation rates and perform curve fitting to obtain a precipitation rate fitting curve.

[0081] It is understood that each real-time precipitation rate is determined based on interval-captured downhole images. In this embodiment, a curve fitting algorithm can be used to fit each real-time precipitation rate to produce a curve showing the real-time precipitation rate over time. This curve clearly demonstrates the temporal trend of the precipitation rate, providing a basis for subsequent analysis.

[0082] S502: Obtain multiple historical precipitation rate curves.

[0083] The historical precipitation rate curve is generated based on target precipitation rate data stored in a cloud database. It is a curve showing the target precipitation rate changing over time. The target precipitation rate data is generated when a precipitation operation is performed in an area that matches the geological characteristics of the area where the precipitation well is located. Geological characteristics include geological structure, stratigraphic distribution, and / or rock type.

[0084] S503: Determine the target precipitation rate of the precipitation well based on the historical precipitation rate curve.

[0085] Specifically, step S503 may include the following steps S5031-S5033.

[0086] S5031: Calculate the similarity between the precipitation rate fitting curve and each historical precipitation rate curve, and determine the historical precipitation rate curve corresponding to the largest similarity value as the target rate curve.

[0087] In an embodiment of the present application, the similarity between the two curves can be quantified using methods such as a dynamic time warping (DTW) algorithm or calculation of the Pearson correlation coefficient to obtain the similarity between the precipitation rate fitting curve and each historical precipitation rate curve. It is understood that the historical precipitation rate curve with the greatest similarity is most similar to the precipitation rate fitting curve, and in this embodiment of the present application, this historical precipitation rate curve can be determined as the target rate curve. When there is more than one historical precipitation rate curve corresponding to the greatest numerical similarity, one of them can be randomly selected as the target rate curve, and this embodiment of the present application does not specifically limit this.

[0088] S5032: Time-align the precipitation rate fitting curve and the target rate curve to determine the first coordinate point corresponding to the real-time precipitation rate in the target rate curve.

[0089] Time alignment can also be referred to as time axis calibration. Because the sampling frequencies (the frequency used to determine the precipitation rate) of the target rate curve and the precipitation rate fitting curve may differ, embodiments of the present application can utilize linear interpolation or other methods to unify the time axes and ensure time scale alignment.

[0090] Furthermore, on the aligned time axis, based on the real-time precipitation rate The first coordinate point corresponding to the target rate curve can be located ( , ), which reflects the current precipitation status in the historical trend.

[0091] S5033: In the target rate curve, determine a second coordinate point corresponding to the first coordinate point after a preset time span, and determine the precipitation rate corresponding to the second coordinate point as the target precipitation rate, where the preset time span is determined based on a preset frequency.

[0092] Among them, the preset time span It can be dynamically adjusted according to the frequency of image acquisition. For example, if the preset frequency is once every 10 minutes, then the preset time span is It can be 10 minutes. Then, in the target rate curve, the first coordinate point ( , ) Moves the preset time span backward along the time axis , determine the second coordinate point ( + , ),in, Indicates the target precipitation rate.

[0093] S504: Substitute the target precipitation rate into the frequency formula to obtain a first target operating frequency of the precipitation pump.

[0094] The frequency formula can be: ; in, Indicates the first target operating frequency, Indicates the current operating frequency of the precipitation pump. is the target precipitation rate, is the real-time precipitation rate, is the adjustment coefficient, 0< <1.

[0095] It should be noted that the current operating frequency of the dewatering pump can be collected by the edge control device 200 at a preset frequency and reported to the cloud processing platform 100, or it can be collected and reported in response to the first control instruction. The embodiment of the present application does not make specific limitations on this.

[0096] It can be understood that based on the above frequency formula, if the real-time precipitation rate Greater than target precipitation rate , then the first target operating frequency It must be less than the current operating frequency of the precipitation pump Therefore, adjusting the precipitation pump based on the first target operating frequency can have the effect of slowing down the precipitation rate.

[0097] In some implementations, if the real-time precipitation rate No more than the target precipitation rate , then the first target operating frequency Must be greater than or equal to the current operating frequency of the precipitation pump However, since the real-time precipitation rate is greater than the first rate threshold, that is, the current precipitation rate is already in a relatively fast state, if the first target operating frequency is used Speeding up the operating frequency of the precipitation pump or maintaining the operating frequency of the precipitation pump unchanged may easily bring about safety hazards. In the embodiment of the present application, the first target operating frequency may be set to Greater than or equal to the current operating frequency of the precipitation pump In the case of times, 0< <1, exemplary, It can be equal to 0.8 or 0.9. In this way, the operating frequency of the precipitation pump can be appropriately reduced to avoid the occurrence of safety accidents.

[0098] S505: Determine a second target operating frequency of the recharge pump based on the first target operating frequency.

[0099] It can be understood that the operating frequency of the recharge pump is related to the operating frequency of the dewatering pump. The embodiment of the present application can determine the second target operating frequency of the recharge pump based on the first target operating frequency of the dewatering pump to maintain the stability of the groundwater level in the dewatering well and the surrounding area.

[0100] The embodiment of the present application can calculate the adjustment ratio of the first target operating frequency to the current operating frequency of the precipitation pump , and, can obtain the current operating frequency of the recharge pump , then the second target operating frequency of the recharge pump can be equal to the adjustment ratio The current operating frequency of the recharging pump The product of .

[0101] S506: Generate a second control instruction based on the first target operating frequency and the second target operating frequency.

[0102] In this way, the edge control device 200 can respond to the second control instruction and adjust the operating frequency of the precipitation pump and the recharge pump to achieve monitoring and adjustment of precipitation recharge.

[0103] It should also be noted that, in the embodiment of the present application, the imaging focal length of the imaging device 500 can be adjusted based on the real-time precipitation rate, and the specific adjustment steps may include the following S601-S602.

[0104] S601: When the real-time precipitation rate is greater than a first rate threshold and less than a second rate threshold, determine that M is equal to a first amplification factor.

[0105] The first magnification factor is, for example, 1.2 or 1.5. For example, if the current imaging focal length is 20 mm, the imaging focal length after magnification can be changed to 24 mm or 30 mm. Specifically, when the first magnification factor is 1.2, the imaging focal length after magnification is equal to 20 mm × 1.2 = 24 mm. When the first magnification factor is 1.5, the imaging focal length after magnification is equal to 20 mm × 1.5 = 30 mm. In this way, the imaging device 500 can enhance its ability to capture image details.

[0106] S602: When the real-time precipitation rate is greater than or equal to a second rate threshold, determine that M is equal to a second amplification factor; wherein the second amplification factor is greater than the first amplification factor.

[0107] Among them, the second rate threshold is greater than the first rate threshold. The second rate threshold is, for example, equal to 80 centimeters per hour. This embodiment of the present application does not make any specific limitation on this.

[0108] The second magnification factor is, for example, 1.8 or 2. For example, assuming the current focal length is 20mm, the zoomed focal length can be 36mm or 40mm. Specifically, when the second magnification factor is 1.8, the zoomed focal length is 20mm x 1.8 = 36mm, and when the second magnification factor is 2, the zoomed focal length is 20mm x 2 = 40mm. This significantly improves the ability to capture image detail and meets the needs of accurate monitoring under high precipitation rates.

[0109] In summary, the faster the real-time precipitation rate, the greater the magnification factor of the imaging focal length.

[0110] In some implementations, after step S300, the following step S701 may also be included: when the real-time precipitation rate is less than the third rate threshold, a fifth control instruction is issued to the edge control device 200, so that the edge control device 200 adjusts the operating frequency of the precipitation pump and the recharge pump based on the fifth control instruction.

[0111] In this step, the fifth adjustment instruction can be specifically used to increase the operating frequency of the dewatering pump and the recharge pump. For example, the fifth adjustment instruction can include the third target operating frequency of the dewatering pump and the fourth target operating frequency of the recharge pump. The third target operating frequency can be equal to p times the current operating frequency of the dewatering pump, p>1, and the fourth operating frequency can be equal to q times the current operating frequency of the recharge pump, q>1. p and q can be equal or unequal.

[0112] In addition, the third rate threshold is less than the first rate threshold. The third rate threshold is, for example, equal to 10 centimeters per hour. The specific value can be determined based on actual needs, and the embodiment of the present application does not make any specific limitations on this.

[0113] In some implementations, the precipitation recharge control method provided in the embodiment of the present application may further include the following steps S801-S803.

[0114] S801: Perform curve fitting on the operating frequency of the precipitation pump to obtain a precipitation pump operating curve.

[0115] S802: Perform curve fitting on the operating frequency of the recharge pump to obtain an operating curve of the recharge pump.

[0116] In practical applications, fitting methods such as linear regression, polynomial regression or line interpolation can be used to generate the dewatering pump operation curve and the recharge pump operation curve, which is not specifically limited in the embodiment of the present application.

[0117] S803: Sending a display instruction to the display terminal 600 so that the display terminal 600 displays the downhole image, the dewatering rate fitting curve, the dewatering pump operation curve and / or the recharge pump operation curve based on the display instruction.

[0118] By observing downhole images, operators can visually observe water level changes, water flow patterns, and wellbore conditions within the dewatering and recharge wells. This helps promptly detect downhole anomalies such as blockages and leaks. By observing the dewatering rate fitting curve, operators can quickly understand the dynamics of the dewatering process and determine whether the dewatering effect is meeting expectations. By observing the dewatering and recharge pump operating curves, operators can monitor the pumps' operating status in real time.

[0119] In some implementations, such as when the precipitation rate is too high, a pump malfunctions, or an image recognition error occurs, the display terminal 600 can also be used to display corresponding alarm information, which is not specifically limited in the embodiments of the present application.

[0120] In practical applications, the display terminal 600 can be a touch-enabled operation panel or smart device, etc. Through the display terminal 600, operators can not only monitor the precipitation and recharge process in real time, but also adjust parameters and optimize the operation process based on the analysis results provided by the cloud processing platform 100.

[0121] Exemplarily, the display terminal 600 can be a digital smart screen, a computer terminal, and / or a mobile device. The digital smart screen is suitable for use in field monitoring rooms or command centers, displaying detailed information over a large area and facilitating simultaneous data viewing and analysis by multiple operators. A computer terminal can provide field engineers and managers with more detailed system operations and data management capabilities, supporting complex data analysis and chart presentation. Mobile devices, such as smartphones or tablets, allow field personnel to view real-time precipitation recharge status, receive alarm information, and perform simple remote operations at different locations.

[0122] In addition, the communication between the cloud processing platform 100 and the display terminal 600 can be achieved using a wireless network (such as Wi-Fi, 5G, etc.) or a wired network (such as optical fiber, Ethernet, etc.), and the embodiments of the present application do not specifically limit this.

[0123] In this way, the control logic of "one cloud and three ends" can be realized, where "one cloud" refers to the cloud processing platform 100, and "three ends" refer to the digital smart large screen, computer terminal and mobile device end.

[0124] Figure 6 A flow chart of the second precipitation recharge control method provided in an embodiment of the present application.

[0125] like Figure 6 As shown, the embodiment of the present application further provides a precipitation recharge control method, which can be applied to the edge control device 200 and can specifically include the following steps S901-S902: S901: In response to a first control instruction issued by the cloud processing platform 100, the imaging device 500 is controlled based on the first control instruction to capture a downhole image, where the downhole image includes at least a background area corresponding to the well wall in the dewatering well and a water level area corresponding to the groundwater; S902: In response to the second control instruction issued by the cloud processing platform 100, the operating frequency of the precipitation pump and the recharge pump is adjusted based on the second control instruction; and, in response to the third control instruction issued by the cloud processing platform 100, the imaging focal length of the imaging device 500 is magnified by M times based on the third control instruction, where the value of M is determined by the real-time precipitation rate.

[0126] In some implementations, the edge control device 200 may also respond to other instructions from the cloud processing platform 100 and perform corresponding actions. For example, the edge control device 200 may respond to a fourth control instruction and, based on the fourth control instruction, start the dewatering pump, the recharge pump, and the imaging device 500. Alternatively, the edge control device 200 may respond to a fifth control instruction and, based on the fifth control instruction, adjust the operating frequency of the dewatering pump and the recharge pump.

[0127] Figure 7This is a structural diagram of the first precipitation recharge control device provided in an embodiment of the present application.

[0128] like Figure 7 As shown, the precipitation recharge control device 1000 provided in the embodiment of the present application may include: The acquisition module 1001 is configured to: during the operation of the dewatering pump, issue a first control instruction to the edge control device 200 at a preset frequency, so that the edge control device 200 controls the imaging device 500 to capture downhole images based on the first control instruction, where the downhole images include at least a background area corresponding to the well wall and a water level area corresponding to the groundwater in the dewatering well; The recognition module 1002 is configured to: perform recognition processing on the downhole image using an image recognition model to extract water level information corresponding to the downhole image; The comparison module 1003 is configured to: compare the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate; The adjustment module 1004 is configured to: when the real-time precipitation rate is greater than the first rate threshold, issue a second control instruction to the edge control device 200, so that the edge control device 200 adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and issue a third control instruction to the edge control device 200, so that the edge control device 200 magnifies the imaging focal length of the imaging device 500 by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0129] Figure 8 This is a structural diagram of the second precipitation recharge control device 2000 provided in an embodiment of the present application.

[0130] like Figure 8 As shown, the precipitation recharge control device provided in the embodiment of the present application may include: The first control module 2001 is configured to: respond to a first control instruction issued by the cloud processing platform 100, and control the imaging device 500 to capture a downhole image based on the first control instruction, wherein the downhole image includes at least a background area corresponding to the well wall of the dewatering well and a water level area corresponding to the groundwater; The second control module 2002 is configured to: respond to the second control instruction issued by the cloud processing platform 100, adjust the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and respond to the third control instruction issued by the cloud processing platform 100, magnify the imaging focal length of the imaging device 500 by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

[0131] In a specific implementation, the present invention further provides a computer storage medium, wherein the computer storage medium may store a program that, when executed, may include some or all of the steps of each embodiment of the precipitation recharge control method provided by the present invention. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0132] It is easy to understand that those skilled in the art can combine, split, reorganize, etc. the embodiments of the present application based on the several embodiments provided in the present application to obtain other embodiments, and these embodiments do not exceed the scope of protection of the present application.

[0133] The above specific implementation methods further explain in detail the purpose, technical solutions and beneficial effects of the embodiments of the present application. It should be understood that the above are only specific implementation methods of the embodiments of the present application and are not intended to limit the scope of protection of the embodiments of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the scope of protection of the embodiments of the present application.

Claims

1. A precipitation recharge control method, characterized in that: Applied to a cloud processing platform, the method includes: During the operation of the dewatering pump, a first control instruction is issued to the edge control device at a preset frequency, so that the edge control device controls the imaging device to capture a downhole image based on the first control instruction, wherein the downhole image includes at least a background area corresponding to the well wall in the dewatering well and a water level area corresponding to the groundwater; Performing recognition processing on the downhole image using an image recognition model to extract water level information corresponding to the downhole image; Comparing the water level information corresponding to two adjacent downhole images to determine a real-time precipitation rate; When the real-time precipitation rate is greater than a first rate threshold, a second control instruction is issued to the edge control device, so that the edge control device adjusts the operating frequency of the precipitation pump and the recharge pump based on the second control instruction; and a third control instruction is issued to the edge control device, so that the edge control device magnifies the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

2. The precipitation recharge control method according to claim 1, characterized in that: The using an image recognition model to perform recognition processing on the downhole image to extract water level information corresponding to the downhole image includes: Determining target model parameters based on the imaging focal length corresponding to when the downhole image is captured; wherein the image recognition model includes multiple sets of preset model parameters, different sets of preset model parameters correspond to different imaging focal lengths, and the target model parameters are one set of the multiple sets of preset model parameters; Performing recognition processing on the downhole image based on the image recognition model and the target model parameters to determine the edge contour line of the water level area; The water level information is determined based on the edge contour line, and the water level information includes at least target distance, target length and target size. The target distance is the distance between at least one feature point on the edge contour line and the boundary of the downhole image, the target length is the length of the edge contour line, and the target size is the pixel size occupied by the area enclosed by the edge contour line.

3. The precipitation recharge control method according to claim 2, characterized in that: The comparing the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate includes: For two adjacent downhole images, construct a feature vector based on the target distance, the target length, and the target size; Calculating the Euclidean distance between the feature vectors of two adjacent downhole images; Inputting the Euclidean distance into a pre-trained regression model to obtain a water level change; The real-time precipitation rate is calculated based on the water level change and the timestamps of taking two adjacent downhole images.

4. The precipitation recharge control method according to claim 3, characterized in that: Before calculating the Euclidean distance between the feature vectors of two adjacent downhole images, the method further includes: When the imaging focal lengths corresponding to two adjacent downhole images are different, the magnification ratio is calculated based on the imaging focal length corresponding to the first downhole image and the imaging focal length corresponding to the second downhole image; wherein the first downhole image is the first of the two adjacent downhole images, and the second downhole image is the second of the two adjacent downhole images; The feature vector corresponding to the second downhole image is adjusted based on the magnification ratio.

5. The precipitation recharge control method according to claim 1, characterized in that: Before sending the second control instruction to the edge control device, the method further includes: When the real-time precipitation rate is greater than the first rate threshold, aggregating the real-time precipitation rates and performing curve fitting to obtain a precipitation rate fitting curve; Obtaining a plurality of historical precipitation rate curves, wherein the historical precipitation rate curves are generated based on target precipitation rate data stored in a cloud database, wherein the target precipitation rate data is data generated when a precipitation operation is performed in an area matching the geological characteristics of the area where the precipitation well is located; Determining a target precipitation rate for the precipitation well based on the historical precipitation rate curve; Substituting the target precipitation rate into the frequency formula to obtain a first target operating frequency of the precipitation pump; determining a second target operating frequency of the recharge pump based on the first target operating frequency; generating the second control instruction based on the first target operating frequency and the second target operating frequency; The frequency formula is: ； in, represents the first target operating frequency, Indicates the current operating frequency of the precipitation pump, is the target precipitation rate, is the real-time precipitation rate, is the adjustment coefficient, 0< <1.

6. The precipitation recharge control method according to claim 5, characterized in that: Determining the target precipitation rate of the precipitation well based on the historical precipitation rate curve includes: Calculating the similarity between the precipitation rate fitting curve and each of the historical precipitation rate curves, and determining the historical precipitation rate curve corresponding to the similarity with the largest value as the target rate curve; Performing time alignment on the precipitation rate fitting curve and the target rate curve to determine a first coordinate point corresponding to the real-time precipitation rate in the target rate curve; In the target rate curve, a second coordinate point corresponding to the first coordinate point after a preset time span is determined, and the precipitation rate corresponding to the second coordinate point is determined as the target precipitation rate, wherein the preset time span is determined based on the preset frequency.

7. The precipitation recharge control method according to claim 5, characterized in that: The method further comprises: Performing curve fitting on the operating frequency of the precipitation pump to obtain a precipitation pump operating curve; Performing curve fitting on the operating frequency of the recharging pump to obtain a recharging pump operating curve; issuing a display instruction to a display terminal, so that the display terminal displays the downhole image, the dewatering rate fitting curve, the dewatering pump operation curve and / or the recharge pump operation curve based on the display instruction; The display terminal is a digital smart large screen, a computer terminal and / or a mobile device.

8. The precipitation recharge control method according to claim 1, characterized in that: After comparing the water level information corresponding to two adjacent downhole images to determine the real-time precipitation rate, the method further includes: When the real-time precipitation rate is greater than the first rate threshold and less than the second rate threshold, determining M to be equal to a first amplification factor; When the real-time precipitation rate is greater than or equal to the second rate threshold, M is determined to be equal to a second amplification factor; wherein the second amplification factor is greater than the first amplification factor.

9. A precipitation recharge control method, characterized in that: Applied to an edge control device, the method includes: In response to a first control instruction issued by the cloud processing platform, controlling the imaging device to capture a downhole image based on the first control instruction, the downhole image including at least a background area corresponding to the well wall in the dewatering well and a water level area corresponding to the groundwater; In response to a second control instruction issued by the cloud processing platform, the operating frequencies of the precipitation pump and the recharge pump are adjusted based on the second control instruction; and in response to a third control instruction issued by the cloud processing platform, the imaging focal length of the imaging device is magnified M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

10. A precipitation recharge control device, characterized in that: The device comprises: The acquisition module is configured to: during the operation of the dewatering pump, issue a first control instruction to the edge control device at a preset frequency, so that the edge control device controls the imaging device to capture a downhole image based on the first control instruction, wherein the downhole image includes at least a background area corresponding to the well wall of the dewatering well and a water level area corresponding to the groundwater; a recognition module configured to: perform recognition processing on the downhole image using an image recognition model to extract water level information corresponding to the downhole image; A comparison module is configured to: compare the water level information corresponding to two adjacent downhole images to determine a real-time precipitation rate; The adjustment module is configured to: when the real-time precipitation rate is greater than a first rate threshold, issue a second control instruction to the edge control device, so that the edge control device adjusts the operating frequencies of the precipitation pump and the recharge pump based on the second control instruction; and issue a third control instruction to the edge control device, so that the edge control device magnifies the imaging focal length of the imaging device by M times based on the third control instruction, wherein the value of M is determined based on the real-time precipitation rate.

Images