A method and system for unmanned aerial vehicle inspection of a pumped storage power station

Abstract

The present invention proposes a method and system for unmanned aerial vehicle inspection of a pumped storage power station, which belongs to the technical field of unmanned aerial vehicle intelligent inspection. The method comprises the following steps: determining the current power load level of the power system; determining the inspection area of ​​the unmanned aerial vehicle; the unmanned aerial vehicle obtains multi-resolution image frames and multi-density laser point cloud data of the inspection area; and judging whether there is an abnormality in the inspection area. The system comprises a power load prediction unit, an inspection scheduling unit, and an inspection risk identification unit. The technical solution of the present invention determines the key inspection area of ​​the unmanned aerial vehicle based on the current power load level of the power system, and comprehensively judges whether there is an abnormality in the inspection area of ​​the pumped storage power station based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area obtained by the unmanned aerial vehicle. The inspection method is more targeted and the accuracy of abnormality judgment is higher.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) intelligent inspection, and in particular relates to a pumped storage power station UAV inspection method and system, a computer-readable storage medium for implementing the method, a computer program product, and an electronic device. Background Art

[0002] Pumped storage power stations generally include an upper reservoir, a lower reservoir, a waterway system, a power generation system, a switch station, and an outgoing line field, etc. The working principle of a pumped storage power station is to use the power during the low load period of the power grid to pump water from the lower reservoir to the upper reservoir for energy storage, and when the power grid reaches the peak load period, release the water back to the lower reservoir for power generation, mainly playing the role of peak load regulation, frequency regulation, phase regulation, and emergency standby.

[0003] Due to geographical restrictions, the areas where pumped-storage power stations are located are usually densely vegetated, with poor visibility, steep slopes and basically no roads. Due to the construction needs of various areas of pumped-storage power stations, it is inevitable that certain changes in the mountain / water structure will occur. The rock mass is sometimes loose, suspended and broken, and the water level fluctuates greatly during the different operation stages of the power station. We should always pay attention to these dangerous factors. The traditional manual inspection method is inefficient and has low accuracy. The inspection effect will be worse in fog, ice, high temperature and low temperature weather, and it will also affect the safety of the inspectors.

[0004] Chinese invention patent application CN202410073860.1 proposes a pumped-storage power station inspection system based on space remote sensing, including a smart hangar subsystem, an inspection path subsystem, a data acquisition and transmission subsystem, a dynamic verification subsystem and a visualization control subsystem, which realizes remote inspection and supervision management of the pumped-storage power station construction site and improves the level of intelligent control of the construction site.

[0005] The related prior arts retrieved all involve engineering supervision and inspection during the construction phase of a pumped-storage power station. There are no related reports on how to carry out inspections during the daily operation phase after the construction of the pumped-storage power station is completed. Summary of the invention

[0006] In response to the above technical problems, the present invention proposes a pumped-storage power station drone inspection method and system, a computer-readable storage medium, a computer program product and an electronic device for implementing the method.

[0007] In the first aspect of the present invention, a method for drone inspection of a pumped-storage power station is proposed. The method is executed based on the automatic scheduling of control equipment and specifically includes the following steps: determining the current power load level of the power system; determining the inspection area of ​​the drone; the drone obtains multi-resolution image frames and multi-density laser point cloud data of the inspection area; and determining whether there is any abnormality in the inspection area.

[0008] Specifically, the above steps of the method are implemented as follows:

[0009] Determine the current electricity load level of the power system;

[0010] When the power load level is lower than a first preset value, starting the pumping device of the pumped-storage power station, and determining that the inspection area of ​​the drone is the upper reservoir area;

[0011] When the power load level is higher than a second preset value, the power generation device of the pumped storage power station is started, and the inspection area of ​​the drone is determined to be the lower reservoir area; the second preset value is greater than the first preset value;

[0012] When the power load level is greater than a first preset value and less than a second preset value, the inspection area of ​​the drone is determined to be other areas connecting the upper reservoir area and the lower reservoir area.

[0013] Wherein, the drone is equipped with an image capture device and a laser radar system;

[0014] The image capturing device is used to obtain multi-resolution image frames of the patrol area;

[0015] The laser radar system is used to collect multi-density laser point cloud data of the inspection area;

[0016] Based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone, it is determined whether there is an abnormality in the inspection area.

[0017] The image capture device includes at least a first resolution capture mode and a second resolution capture mode; the first resolution is smaller than the second resolution;

[0018] Based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone, determining whether there is an abnormality in the inspection area includes the following steps:

[0019] S310: Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0020] S320: When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference value, determining that there is no abnormality in the inspection area;

[0021] S330: When the absolute value of the frame difference between the first image frame and the second image frame is less than a first preset difference, controlling the drone to enter the second resolution shooting mode;

[0022] S340: Acquire a third image frame and a fourth image frame acquired by the drone in the second resolution shooting mode at a predetermined interval;

[0023] S350: When the absolute value of the frame difference between the third image frame and the fourth image frame is less than a first preset difference, it is determined that an abnormality exists in the inspection area.

[0024] The laser radar system includes at least a first density acquisition mode and a second density acquisition mode; the first density is less than the second density;

[0025] When the step S320 determines that there is no abnormality in the inspection area, the method further includes:

[0026] After acquiring the first laser point cloud data acquired by the drone in the first density acquisition mode, controlling the drone to enter the second density acquisition mode;

[0027] Acquire second laser point cloud data acquired by the UAV in the second density acquisition mode;

[0028] The first laser point cloud data and the second laser point cloud data are analyzed to further confirm whether there is any abnormality in the inspection area.

[0029] The drone acquires the first image frame and the second image frame at a first shooting position, and acquires the third image frame and the fourth image frame at a second shooting position.

[0030] In the second aspect of the present invention, in order to execute the method described in the first aspect, a pumped-storage power station drone inspection system is proposed, wherein the pumped-storage power station is equipped with an upper reservoir area, a lower reservoir area and other areas connecting the upper reservoir area and the lower reservoir area; the drone is equipped with an image shooting device and a lidar system; the image shooting device is used to obtain multi-resolution image frames of the inspection area; the lidar system is used to collect multi-density laser point cloud data of the inspection area.

[0031] The system further comprises:

[0032] An electricity load forecasting unit, used to forecast the current electricity load level of the power system;

[0033] A patrol dispatching unit, when the power load level is lower than a first preset value, the patrol dispatching unit starts the pumping device of the pumped storage power station and determines that the patrol area of ​​the drone is the upper reservoir area;

[0034] When the power load level is higher than a second preset value, the inspection dispatching unit starts the power generation device of the pumped storage power station and determines that the inspection area of ​​the drone is the lower reservoir area; the second preset value is greater than the first preset value;

[0035] When the power load level is greater than a first preset value and less than a second preset value, the inspection area of ​​the drone is determined to be other areas connecting the upper reservoir area and the lower reservoir area.

[0036] The inspection risk identification unit determines whether there is any abnormality in the inspection area based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone.

[0037] The laser radar system includes at least a first density acquisition mode and a second density acquisition mode; the first density is less than the second density.

[0038] The image capture device includes at least a first resolution capture mode and a second resolution capture mode; the first resolution is smaller than the second resolution.

[0039] The inspection risk identification unit determines whether there is an abnormality in the inspection area based on the multi-resolution image frame and multi-density laser point cloud data of the inspection area acquired by the drone, including:

[0040] Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0041] When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference value, it is determined that there is no abnormality in the inspection area.

[0042] The inspection risk identification unit determines whether there is an abnormality in the inspection area based on the multi-resolution image frame and multi-density laser point cloud data of the inspection area acquired by the drone, including:

[0043] Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0044] When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference, after acquiring the first laser point cloud data acquired by the drone in the first density acquisition mode, controlling the drone to enter the second density acquisition mode;

[0045] Acquire second laser point cloud data acquired by the UAV in the second density acquisition mode;

[0046] The first laser point cloud data and the second laser point cloud data are analyzed to further confirm whether there is any abnormality in the inspection area.

[0047] The aforementioned pumped-storage power station drone inspection method can be automatically implemented through various forms of electronic devices and computer program instructions; the computer program instructions can be stored in different forms of storage media and loaded into computer electronic devices for execution.

[0048] Therefore, in the third aspect of the present invention, a computer-readable storage medium is also provided for storing computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes all or part of the steps of the aforementioned pumped-storage power station drone inspection method.

[0049] In the fourth aspect of the present invention, a computer device is also proposed, which includes a processor and a memory, the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the computer device executes the aforementioned pumped-storage power station drone inspection method.

[0050] In the fifth aspect of the present invention, a computer program product is also proposed, which includes a computer program. When the computer program is executed, all or part of the steps of the above-mentioned pumped-storage power station drone inspection method are implemented.

[0051] The technical solution of the present invention determines the key inspection area of ​​the UAV based on the current power load level of the power system, and comprehensively judges whether there is any abnormality in the inspection area of ​​the pumped-storage power station based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the UAV. The inspection method is more targeted and the accuracy of abnormality judgment is higher, which effectively solves the technical problem of how to carry out inspections in the daily operation stage after the construction of the pumped-storage power station is completed.

[0052] Further advantages of the present invention will be further reflected in detail in the specific embodiments section in conjunction with the drawings of the specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. 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 creative work.

[0054] Figure 1 This is a common regional layout diagram of a pumped storage power station;

[0055] Figure 2 It is a main flow chart of a method for inspecting a pumped storage power station by a drone according to an embodiment of the present invention;

[0056] Figure 3 yes Figure 2 A step diagram of a computer process implementation of the pumped storage power station drone inspection method;

[0057] Figure 4 It is a schematic diagram of the appearance structure of the UAV used in the technical solution of the present invention;

[0058] Figure 5 is based on Figure 4 A schematic diagram of the drone performing multi-density laser point cloud data collection;

[0059] Figure 6 It is for Figure 5 A schematic diagram of data processing of the multi-density laser point cloud data;

[0060] Figure 7 The present invention is a schematic diagram of the functional module unit composition of a pumped storage power station drone inspection system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] First of all, it should be pointed out that the embodiment of the pumped-storage power station drone inspection method mentioned in this section can be implemented through a computer program on an electronic device or system configured with a memory and a processor. The electronic device or system can be in the form of a physical machine, a virtual machine, a server, a cluster, or any combination thereof.

[0062] Preferably, the specific form of the electronic device may also be a human-computer interaction terminal, and the human-computer interaction terminal may be a desktop terminal, a smart handheld terminal, a mobile terminal, etc. with a human-computer interaction interface.

[0063] See first Figure 1 , Figure 1 This is a common regional layout diagram of a pumped-storage power station.

[0064] The working principle of a pumped-storage power station is to utilize electricity from the grid during low load periods, pump water from the lower reservoir to the upper reservoir for energy storage, and then release water back to the lower reservoir to generate electricity when the grid is at peak load. It mainly serves the functions of peak-loading, frequency regulation, phase regulation, and emergency standby.

[0065] Pumped storage power stations generally include upper reservoirs, lower reservoirs, waterway systems, power generation systems, switch stations and outgoing line yards, etc. Figure 1The upper reservoir, lower reservoir and underground powerhouse are shown. Pumped storage power stations use the terrain advantages to convert energy, and their core power equipment is mostly located in underground powerhouses (such as underground caverns).

[0066] When the power load in the power system is at a low point, the power station uses the excess power in the system to pump water from the lower reservoir to the upper reservoir and store it in the form of hydropower; when the power load is at a peak, the water in the upper reservoir is released to the lower reservoir, and the hydropower is converted into electricity for users. Compared with conventional power stations, pumped storage power stations have frequent operating conditions. In order to meet the peak and frequency regulation needs of the system, the pumped storage units are frequently turned on and off and the operating conditions are frequently changed within a regulation cycle. Taking a large domestic daily regulation pumped storage unit as an example, the typical operation mode is: "two generation and one pumping", that is, generating electricity during the morning peak of electricity consumption from 8:00 to 12:00, generating electricity during the evening peak from 17:00 to 22:00, and pumping water during the valley from 23:00 to 6:00 the next day. During the peak summer (winter) and power supply shortage periods, the operation mode is adjusted to "two generation and two pumping per day" or "two generation and three pumping per day", that is, increasing pumping when the load is relatively low before the noon or evening peak, thereby greatly increasing the power generation load during the evening peak period.

[0067] During the early construction of each area of ​​the pumped storage power station, there are inevitably certain changes in the mountain / water structure. The rock mass is sometimes loose, suspended, and broken, and the water level fluctuates greatly in different operating stages of the power station. We should always pay attention to these dangerous factors, especially at different nodes where the working conditions of the pumped storage power station change, and we need to inspect key areas. The inspection mainly focuses on whether some abnormal phenomena occur, such as cracks, new leakage points, concrete scouring and freeze-thaw, precipitates from the dam foundation, local deformation, landslide collapse, etc.

[0068] With the rapid development of drone technology in recent years, drone inspection technology has been widely used in industries such as agriculture, energy, public safety, infrastructure and construction. Compared with traditional manual inspections, the efficiency and quality of drone inspections have been significantly improved.

[0069] However, there have been no reports on how to carry out inspections during the daily operation phase after the construction of pumped-storage power stations is completed.

[0070] In this regard, a technical solution of the present application is proposed.

[0071] The specific embodiments of the technical solution of this application are as follows: Figure 1 Let’s start with the introduction.

[0072] Figure 2 The main flow chart of the method for inspecting a pumped storage power station by a drone according to one embodiment of the present invention is shown, and mainly includes the following steps (for ease of reading, the step numbers are omitted in the relevant drawings and are appropriately summarized):

[0073] S1: Determine the current power load level of the power system;

[0074] S2: Based on the current power load level of the power system determined in step S1, determining the inspection area of ​​the drone;

[0075] S3: Acquire multi-resolution image frames and multi-density laser point cloud data of the inspection area through a drone;

[0076] S4: Based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area, determine whether there is an abnormality in the inspection area.

[0077] Figure 2 The method can be automatically implemented on a control device or a scheduling device, such as a terminal control platform, by a computer program. Specifically, a specific flowchart of the method implemented by a computer program can be found in Figure 3 The description is as follows:

[0078] The terminal control platform determines the current power load level of the power system;

[0079] In a specific embodiment, the terminal control platform may predict the power load level of the power system in the current period based on historical data; or, the terminal control platform may obtain the current power load level of the power system in real time according to a preset period.

[0080] When the power load level is lower than a first preset value, starting the pumping device of the pumped-storage power station, and determining that the inspection area of ​​the drone is the upper reservoir area;

[0081] When the power load level is higher than a second preset value, the power generation device of the pumped-storage power station is started, and the inspection area of ​​the drone is determined to be the lower reservoir area; the second preset value is greater than the first preset value.

[0082] In an electricity consumption scenario, the terminal control platform calculates the morning peak period from 8:00 to 12:00 based on historical rules; the evening peak period from 17:00 to 22:00; and the valley period from 23:00 to 6:00 the next day. The "peak period" and "valley period" here are determined based on the real-time power load level of the power system. When the power load level is lower than the first preset value (for example, less than 50% of the average value), it is determined as the valley period; when the power load level is higher than the second preset value (for example, greater than 30% of the average value), it is determined as the peak period.

[0083] A pumped-storage power station uses electricity during the low-load period to pump water to the upper reservoir (referred to as the pumping stage), and releases water to the lower reservoir for power generation during the peak load period (referred to as the power generation stage). In the above example, power is generated during the morning peak period of 8:00-12:00, the evening peak period of 17:00-22:00, and water is pumped during the low-load period of 23:00-6:00 the next day.

[0084] After long-term observation, the inventors noticed that during the pumping stage, the biggest risk is the idling of the pump due to insufficient pumping kinetic energy. Since the pump is located in the underground space and cannot be directly observed, whether the pumping device is operating normally is mainly reflected in whether the water level in the upper reservoir area changes significantly, and when the water level in the upper reservoir changes significantly, whether it is accompanied by changes in the mountain / water structure in the upper reservoir area due to water level changes and switching of pump operating conditions, loosening of rock, overhead, and fragmentation.

[0085] To this end, as a first improvement point of the present invention, when the power load level is lower than a first preset value, the pumping device of the pumped-storage power station is started, and the inspection area of ​​the drone is determined to be the upper reservoir area.

[0086] In other words, if the current phase is pumping water, the key inspection area of ​​the drone will be the upper reservoir area.

[0087] On the other hand, if it is in the power generation stage, the biggest risk comes from the smooth flow of water channels and whether the generators are energy efficient. Similarly, water channels and generators are usually located in underground structures and cannot be directly inspected. At this time, it is necessary to inspect the water level and other water / rock structure changes in the lower reservoir area, mainly to see whether the water level in the lower reservoir area has changed significantly, and when the lower reservoir changes significantly, whether it is accompanied by changes in the mountain / water structure in the lower reservoir area due to water level changes and generator operating mode switching, rock loosening, overhead, and fragmentation, etc.

[0088] To this end, as a second improvement of the present invention, when the power load level is higher than a second preset value, the power generation device of the pumped-storage power station is started, and the inspection area of ​​the drone is determined to be the lower reservoir area.

[0089] In other words, if the power generation stage is currently underway, the key inspection area for the drone will be the lower reservoir area.

[0090] Preferably, if it is in other stages, for example, when the power load level is greater than the first preset value and less than the second preset value, the inspection area of ​​the drone is determined to be other areas connecting the upper reservoir area and the lower reservoir area. That is to say, when the pumped storage power station is in a dormant period, the inspection is focused on other areas.

[0091] It can be seen that the above-mentioned inspection area determination method of the embodiment of the present invention conforms to the actual operating rules of the pumped-storage power station and the characteristics of the on-site risk sources, making the inspection area determination of the drone more targeted, thereby enabling the inspection to be completed efficiently and accurately under limited conditions such as limited drone resources and the limited endurance of the drone itself.

[0092] After determining the cruising area during the corresponding period, the next step is to identify risks based on the data collected by the drone.

[0093] As a third advantage of the present invention, this embodiment comprehensively considers multi-resolution image frames and multi-density laser point cloud data of the inspection area when performing risk identification.

[0094] Specifically, the drone is equipped with an image capture device and a laser radar system; the image capture device is used to obtain multi-resolution image frames of the patrol area; and the laser radar system is used to collect multi-density laser point cloud data of the patrol area.

[0095] Figure 4 It is a schematic diagram of the appearance structure of the UAV used in the technical solution of the present invention. Figure 4 It is highlighted that the drone includes an onboard laser radar LiDAR.

[0096] Based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone, it is determined whether there is an abnormality in the inspection area.

[0097] In this embodiment, judging whether the patrol area is abnormal includes a judgment process at different levels in two dimensions, and the judgment in each dimension includes two levels at the same time.

[0098] The level of the first dimension is based on the inter-frame difference judgment of the image data frames.

[0099] Specifically, the image capture device includes at least a first resolution capture mode and a second resolution capture mode; the first resolution is smaller than the second resolution;

[0100] Based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone, determining whether there is an abnormality in the inspection area includes the following steps:

[0101] S310: Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0102] S320: When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference value, determining that there is no abnormality in the inspection area;

[0103] As mentioned above, during the pumping stage, the inspection area is the upper reservoir area. If there is no abnormality, the water level in the upper reservoir area should change to a certain extent, such as the water level rising by a certain height per second. At this time, the drone takes two images of the same water area at a predetermined interval. There must be obvious inter-frame differences between the two images, such as the water level line caused by the water level change, the edge detection line caused by the water level change, etc.; that is, when the absolute value of the inter-frame difference between the first image frame and the second image frame is greater than the first preset difference, it can be preliminarily determined that there is no abnormality in the inspection area;

[0104] On the contrary, if the drone targets the same water area and the two images taken at a preset time interval cannot detect obvious frame differences, that is, when the absolute value of the frame difference between the first image frame and the second image frame is less than the first preset difference, it may indicate the possibility of an abnormality, that is, no water level change is detected, and the state of the pumped-storage power station is likely to be abnormal.

[0105] Of course, this anomaly may also be caused by other factors such as the drone's collection position, image resolution, etc. To avoid misjudgment, this embodiment is further improved as follows:

[0106] S330: When the absolute value of the frame difference between the first image frame and the second image frame is less than a first preset difference, controlling the drone to enter the second resolution shooting mode;

[0107] S340: Acquire a third image frame and a fourth image frame acquired by the drone in the second resolution shooting mode at a predetermined interval;

[0108] S350: When the absolute value of the frame difference between the third image frame and the fourth image frame is less than a first preset difference, it is determined that an abnormality exists in the inspection area.

[0109] The drone acquires the first image frame and the second image frame at a first shooting position, and acquires the third image frame and the fourth image frame at a second shooting position.

[0110] That is to say, if the drone improves the shooting resolution and changes the shooting position, and the two images taken at a predetermined interval still cannot detect an obvious frame difference, then the judgment result is abnormal, that is, no water level change is detected and the state of the pumped-storage power station is abnormal.

[0111] The first dimension level introduced above is based on the inter-frame difference judgment of the multi-resolution image data frame, and there are two judgment levels with different resolutions and different positions.

[0112] Next, the second dimension is introduced, which is the abnormal point recognition and judgment based on multi-density laser point cloud data. It also has judgment levels of different densities and latitudes.

[0113] The activation of this dimension is mainly to further verify the result of the aforementioned step S320 that initially determined that there is no abnormality, so as to avoid possible errors in single-dimensional judgment.

[0114] Specifically, the laser radar system includes at least a first density acquisition mode and a second density acquisition mode; the first density is less than the second density;

[0115] When the step S320 preliminarily determines that there is no abnormality in the inspection area, the method further includes:

[0116] After acquiring the first laser point cloud data acquired by the drone in the first density acquisition mode, controlling the drone to enter the second density acquisition mode;

[0117] Acquire second laser point cloud data acquired by the UAV in the second density acquisition mode;

[0118] The first laser point cloud data and the second laser point cloud data are analyzed to further confirm whether there is any abnormality in the inspection area.

[0119] The first laser point cloud data and the second laser point cloud data are acquired by the UAV at two different acquisition positions.

[0120] By combining regional images with regional LiDAR technology, based on predetermined standard regional image data, it is possible to effectively identify abnormal water changes and abnormal landform changes that may exist in the target area. This technology is well known to those skilled in the art. The implementation method is briefly introduced below.

[0121] Figure 5 is based on Figure 4 Schematic diagram of the UAV performing multi-density laser point cloud data collection.

[0122] In an illustrative example, in order to meet the requirements of large height difference and high-precision laser point cloud data acquisition in dense vegetation areas, the DV-L i DAR20 laser radar system is used in combination with Pegasus UAV Manager software to design a terrain-simulating flight route; laser point cloud and optical image data are collected at the same time to ensure the consistency of image resolution and point cloud density in the area. The initial design image ground resolution is 16cm, and the design laser point cloud density is 50 points / m 2 The heading overlap is 30%, the lateral overlap is 40%, and the flight speed is 20m / s. Figure 5 The laser point cloud data collection results are shown in Figure 2.

[0123] When collecting again, the ground resolution of the image is designed to be 6 cm, and the density of the laser point cloud is designed to be 60 points / m2 The heading overlap is 80%, the lateral overlap is 70%, and the flight speed is 10m / s. Next, the acquired laser point cloud data needs to be processed. Figure 6 Shows the Figure 5 Schematic diagram of data processing of the multi-density laser point cloud data.

[0124] Specifically, the acquired laser point cloud data are subjected to POS solution, flight zone adjustment, denoising filtering, and classification processing to obtain ground point cloud and non-ground point cloud data. TerraSo lid software is used to obtain DSM and DEM results with a grid spacing of 0.2m. At the same time, DOM results are generated based on airborne synchronous optical images and POS data with an image resolution of 0.1m.

[0125] EarthSurvey software was used to build a comprehensive interpretation platform for multi-source remote sensing data. The LiDAR data results DOM, DEM, hill shade and other basic geological data were imported into the platform to create a three-dimensional geological interpretation sand table, establish karst interpretation signs, and carry out the interpretation of karst engineering geological problems.

[0126] Based on the hydrogeological and environmental geological background data of the survey area, combined with the karst geomorphological characteristics of carbonate rock formations, a technical method combining three-dimensional models and two-dimensional images was used to establish karst remote sensing interpretation signs. A preliminary interpretation was carried out based on the results of airborne LiDAR data. The interpretation signs and preliminary interpretation results were verified during the field review process, and further detailed interpretation was carried out.

[0127] Based on the EarthSurvey 3D interpretation platform, the rock structure surface is identified and measured through the "geological disaster analysis" - "structural surface measurement" module to determine whether there are risk disaster points and their specific location distribution.

[0128] Most of the above processes belong to the prior art and will not be further expanded in this embodiment. For more principles, please refer to the following documents:

[0129] Sun Tao, Xu Mingyu, Dong Xiujun, et al. Application of airborne LiDAR technology in geological disaster investigation in densely vegetated mountainous areas[J]. Bulletin of Surveying and Mapping, 2021(4):90-97.

[0130] Preferably, analyzing the first laser point cloud data and the second laser point cloud data to further confirm whether there is an abnormality in the inspection area specifically includes:

[0131] Analyze the first laser point cloud data, and when the analysis result indicates that there is an abnormality, determine that there is an abnormal area in the inspection area, and exit the method;

[0132] When the analysis result of the first laser point cloud data shows that there is no abnormality, the second laser point cloud data continues to be analyzed. When the analysis result of the second laser point cloud data shows that there is an abnormality, it is determined that there is an abnormal area in the inspection area.

[0133] Therefore, more preferably, when the step S320 determines that there is no abnormality in the inspection area, the method further includes:

[0134] Acquire first laser point cloud data acquired by the drone in the first density acquisition mode; parse the first laser point cloud data, and when the parsing result of the first laser point cloud data indicates that it is abnormal, determine that there is an abnormal area in the inspection area;

[0135] Otherwise, controlling the UAV to enter the second density collection mode;

[0136] Acquire second laser point cloud data acquired by the UAV in the second density acquisition mode;

[0137] The second laser point cloud data is analyzed, and when the analysis result of the second laser point cloud data indicates that an abnormality exists, it is determined that an abnormal area exists in the inspection area.

[0138] In order to maintain consistency with the image data, more preferably, the first laser point cloud data is acquired by the drone at the second shooting position, and the second laser point cloud data is acquired by the drone at the first shooting position.

[0139] After introducing the above method embodiments, the corresponding system embodiments of the present application are introduced next. It can be understood that the principles and advantages of the system embodiments are substantially corresponding to the method embodiments, so the system embodiments are not introduced one by one in detail, and only the basic functional units are introduced.

[0140] See also Figure 7 , Figure 7 The present invention is a schematic diagram of the functional module unit composition of a pumped storage power station drone inspection system according to an embodiment of the present invention.

[0141] Figure 7 The three functional units of the system are shown to include:

[0142] An electricity load forecasting unit, used to forecast the current electricity load level of the power system;

[0143] A patrol dispatching unit, when the power load level is lower than a first preset value, the patrol dispatching unit starts the pumping device of the pumped storage power station and determines that the patrol area of ​​the drone is the upper reservoir area;

[0144] When the power load level is higher than a second preset value, the inspection dispatching unit starts the power generation device of the pumped storage power station and determines that the inspection area of ​​the drone is the lower reservoir area; the second preset value is greater than the first preset value;

[0145] When the power load level is greater than a first preset value and less than a second preset value, the inspection area of ​​the drone is determined to be other areas connecting the upper reservoir area and the lower reservoir area.

[0146] The inspection risk identification unit determines whether there is any abnormality in the inspection area based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone.

[0147] The above method embodiments and system embodiments of the present invention are applicable to the automatic inspection process of drones in all typical pumped-storage power stations. A typical pumped-storage power station is equipped with an upper reservoir area, a lower reservoir area, and other areas connecting the upper reservoir area and the lower reservoir area; the drone is equipped with an image capture device and a laser radar system; the image capture device is used to obtain multi-resolution image frames of the inspection area; the laser radar system is used to collect multi-density laser point cloud data of the inspection area.

[0148] The laser radar system includes at least a first density acquisition mode and a second density acquisition mode; the first density is less than the second density.

[0149] The image capture device includes at least a first resolution capture mode and a second resolution capture mode; the first resolution is smaller than the second resolution.

[0150] The inspection risk identification unit determines whether there is an abnormality in the inspection area based on the multi-resolution image frame and multi-density laser point cloud data of the inspection area acquired by the drone, including:

[0151] Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0152] When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference value, it is determined that there is no abnormality in the inspection area.

[0153] The inspection risk identification unit determines whether there is an abnormality in the inspection area based on the multi-resolution image frame and multi-density laser point cloud data of the inspection area acquired by the drone, including:

[0154] Acquire a first image frame and a second image frame acquired by the drone in the first resolution shooting mode at a predetermined interval;

[0155] When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference, after acquiring the first laser point cloud data acquired by the drone in the first density acquisition mode, controlling the drone to enter the second density acquisition mode;

[0156] Acquire second laser point cloud data acquired by the UAV in the second density acquisition mode;

[0157] The first laser point cloud data and the second laser point cloud data are analyzed to further confirm whether there is any abnormality in the inspection area.

[0158] The present invention proposes for the first time an improved technical solution based on drone inspection during the daily operation stage after the construction of a pumped-storage power station is completed. The solution mainly determines the key drone inspection area based on the current power load level of the power system, and comprehensively analyzes the multi-resolution image frames and multi-density laser point cloud data of the inspection area obtained by the drone to determine whether there are any abnormalities in the inspection area of ​​the pumped-storage power station. The inspection method is more targeted and the accuracy of abnormality judgment is higher.

[0159] The technical solution proposed in the present invention includes at least the following advantages:

[0160] (1) The inspection area determination method is in line with the actual operation rules of the pumped-storage power station and the characteristics of the on-site risk sources, making the inspection area determination of the drone more targeted, thereby enabling the inspection to be completed efficiently and accurately under the constraints of limited drone resources and the limited endurance of the drone itself;

[0161] (2) When conducting risk identification, the multi-resolution image frames and multi-density laser point cloud data of the inspection area are comprehensively considered; judging whether the inspection area is abnormal includes a judgment process at different levels in two dimensions, and the judgment of each dimension includes two levels at the same time to avoid possible errors in single-dimensional judgment.

[0162] It should be noted that the present invention proposes multiple embodiments, each of which can solve at least one technical problem and achieve at least one of the above-mentioned improved technical effects, but it is not required that each embodiment solves all technical problems at the same time or has all improved technical effects. Each embodiment itself may contribute to the prior art and thus constitute an independent and creative technical solution.

[0163] For other technologies, principles, algorithms or models not elaborated in detail in this application, please refer to the prior art.

[0164] At the same time, in the specific implementation of this application, if user-related data is involved, when the embodiment of this application is applied to a specific product or technology, the user's permission or consent must be obtained, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0165] The foregoing has shown and described the method embodiments and system of the present invention, but it is understandable to those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for inspecting a pumped storage power station by using a drone, characterized in that: The method comprises the following steps: Determine the current electricity load level of the power system; When the power load level is lower than the first preset value, the pumping device of the pumped-storage power station is started, and the inspection area of ​​the drone is determined to be the upper reservoir area; when the power load level is higher than the second preset value, the power generation device of the pumped-storage power station is started, and the inspection area of ​​the drone is determined to be the lower reservoir area; the second preset value is greater than the first preset value; The drone is equipped with an image capture device and a laser radar system; The image capture device includes a first resolution capture mode and a second resolution capture mode; the first resolution is smaller than the second resolution; the laser radar system includes a first density acquisition mode and a second density acquisition mode; the first density is smaller than the second density; Based on the multi-resolution image frames and multi-density laser point cloud data of the patrol area obtained by the drone, it is determined whether there are any abnormalities in the patrol area, including: Acquire a first image frame and a second image frame acquired by the drone in a first resolution shooting mode at a predetermined interval; When the absolute value of the frame difference between the first image frame and the second image frame is greater than a first preset difference value, it is determined that there is no abnormality in the inspection area; When the absolute value of the frame difference between the first image frame and the second image frame is less than the first preset difference, a third image frame and a fourth image frame acquired by the drone at a predetermined interval in the second resolution shooting mode are obtained; when the absolute value of the frame difference between the third image frame and the fourth image frame is less than the first preset difference, it is determined that an abnormality exists in the patrol area; When it is determined that there is no abnormality in the inspection area, the method further includes: After acquiring the first laser point cloud data acquired by the drone in the first density acquisition mode, continue to acquire the second laser point cloud data acquired by the drone in the second density acquisition mode; The first laser point cloud data and the second laser point cloud data are analyzed to further confirm whether there is any abnormality in the inspection area.

2. A pumped storage power station drone inspection method as claimed in claim 1, characterized in that: When the power load level is greater than a first preset value and less than a second preset value, the inspection area of ​​the drone is determined to be other areas connecting the upper reservoir area and the lower reservoir area.

3. A pumped storage power station drone inspection method as claimed in claim 1, characterized in that: The drone acquires the first image frame and the second image frame at a first shooting position, and acquires the third image frame and the fourth image frame at a second shooting position.

4. A pumped-storage power station drone inspection system for implementing the pumped-storage power station drone inspection method according to any one of claims 1 to 3, wherein the pumped-storage power station is configured with an upper reservoir area, a lower reservoir area, and other areas connecting the upper reservoir area and the lower reservoir area; It is characterized in that The system further comprises: An electricity load forecasting unit, used to forecast the current electricity load level of the power system; A patrol dispatching unit, when the power load level is lower than a first preset value, the patrol dispatching unit starts the pumping device of the pumped storage power station and determines that the patrol area of ​​the drone is the upper reservoir area; When the power load level is higher than a second preset value, the inspection dispatching unit starts the power generation device of the pumped storage power station and determines that the inspection area of ​​the drone is the lower reservoir area; the second preset value is greater than the first preset value; The inspection risk identification unit determines whether there is any abnormality in the inspection area based on the multi-resolution image frames and multi-density laser point cloud data of the inspection area acquired by the drone.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of a method for inspecting a pumped-storage power station using a drone are implemented as described in any one of claims 1 to 3.

Images