Edge calculation data in-situ real-time analysis method based on multi-parameter Internet of Things technology

Abstract

The invention provides an on-site real-time edge calculation data analysis method based on a multi-parameter Internet of Things technology. The method comprises the following steps: generating a detection route based on each photovoltaic target, and arranging a detection unit between the photovoltaic targets with an adjacent staggered relationship; the unmanned surface vehicle is controlled to perform image acquisition on each photovoltaic target based on the detection route, and when the early warning evaluation value of any photovoltaic target is greater than a preset early warning threshold value, the light unit of the unmanned surface vehicle is triggered to be converted into a second display mode; controlling the unmanned ship to move from the current position to the nearest detection unit, and triggering the detection unit to perform video detection; determining that the unmanned ship arrives at a preset position, controlling the unmanned ship to return to the current position, and obtaining a moving direction; and marking the image part of the photovoltaic target in the sea area view, controlling the unmanned ship to perform image acquisition on the remaining photovoltaic target based on the detection route when the unmanned ship arrives at the current position, and triggering the light unit to convert into a first display mode. The image acquisition efficiency is at least improved.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to an on-site real-time analysis method for edge computing data based on multi-parameter Internet of Things technology. Background Art

[0002] Since buildings and roads on land have occupied a lot of land resources, building offshore photovoltaic panels has become a new trend. Building offshore photovoltaic panels not only does not occupy land resources, but also can use the cooling effect of seawater to reduce the temperature of offshore photovoltaic panels, thereby improving the photoelectric conversion efficiency.

[0003] The inventors found in their research that any object may age over time. In the more complex testing environment of the sea, how to conduct quality inspection on offshore photovoltaic panels is a major problem that needs to be solved. Summary of the invention

[0004] Based on the above problems, the present invention is proposed to provide an on-site real-time analysis method of edge computing data based on multi-parameter Internet of Things technology to overcome the above problems or at least partially solve the above problems.

[0005] According to one aspect of the present invention, a method for real-time analysis of edge computing data based on multi-parameter IoT technology is provided, comprising the following steps: Based on the photovoltaic targets arranged in arrays on the sea area, a detection route passing through each photovoltaic target in sequence is generated, and corresponding detection units are respectively pre-set between each photovoltaic target having an adjacent staggered relationship determined based on the detection route; Controlling the unmanned boat to sequentially collect images of each photovoltaic target based on the detection route, and in response to determining based on the collection result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, triggering a light unit pre-installed on the unmanned boat to switch from a first display mode to a second display mode; Controlling the unmanned boat to move from a current position toward a detection unit closest to the unmanned boat, and triggering the detection unit to perform video detection; In response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, controlling the unmanned boat to return from the preset position to the current position, and acquiring a moving direction of the unmanned boat based on the information collected by the detection unit; Based on the moving direction, the image portion corresponding to the photovoltaic target is marked in the sea area view corresponding to the sea area. In response to the unmanned boat arriving at the current position, the unmanned boat is controlled to continue to collect images of the remaining photovoltaic targets based on the detection route, and the light unit is triggered to switch from the second display mode to the first display mode.

[0006] Optionally, in the method according to the present invention, based on each photovoltaic target arranged in an array and located in the sea area, a detection route passing through each photovoltaic target in sequence is generated, and corresponding detection units are respectively set in advance between each photovoltaic target having an adjacent staggered relationship determined based on the detection route, including: Determine a first arrangement number of photovoltaic targets in an array arrangement in a first arrangement direction and a second arrangement number in a second arrangement direction, wherein the first arrangement direction is perpendicular to the second arrangement direction; Compare the number of the first arrangement with the number of the second arrangement, and based on the comparison result, determine the arrangement direction with the larger corresponding number as the primary moving direction, and determine the arrangement direction with the smaller corresponding number as the secondary moving direction; Grouping the photovoltaic targets based on the secondary moving direction to obtain primary groups, and sorting the photovoltaic targets in the same primary group based on the primary moving direction to obtain a first target sequence; Among the photovoltaic targets, any photovoltaic target at a corner position is determined as a starting target; Taking the main level group including the starting target as the starting point, sorting the main level groups to obtain a group sequence, and based on the group sequence, reversely modifying the first target sequence of all main level groups corresponding to even numbers to obtain a second target sequence; Based on the grouping sequence and each target sequence, a detection route is determined that takes the starting area as the detection starting point and passes through each photovoltaic target in sequence, and each primary grouping at an adjacent position corresponding to the detection route is determined as the same staggered determination set; Based on a preset interleaving determination strategy, each photovoltaic target in the same interleaving determination set is interleaved and determined, and corresponding detection units are pre-arranged between each photovoltaic target that has the same adjacent interleaving relationship.

[0007] Optionally, in the method according to the present invention, interleaving determination is performed on photovoltaic targets corresponding to the same interleaving determination set based on a preset interleaving determination strategy, and corresponding detection units are preliminarily set between photovoltaic targets having the same adjacent interleaving relationship, including: Determine two primary groups in the same interleaved determination set as a first group and a second group respectively; Along the main-stage moving direction, the photovoltaic targets in the first group and the photovoltaic targets in the second group are respectively interlaced and connected one by one, so as to obtain target connection lines respectively connecting the photovoltaic targets having different adjacent interlaced relationships; Determine the intersection of each target connecting line, and obtain the corresponding intersection points of each line segment based on the determination result; The corresponding detection units are set based on the intersection points of each line segment.

[0008] Optionally, in the method according to the present invention, the photovoltaic size of the photovoltaic target is smaller than the frame size of the image acquisition frame corresponding to the unmanned boat; Controlling the unmanned boat to sequentially collect images of each photovoltaic target based on the detection route includes: Controlling the unmanned boat to collect images based on the detection route, and performing binarization processing on the acquired collected images to obtain a binarized image, wherein the binarized image includes photovoltaic pixel points corresponding to the first pixel value and / or noise pixel points corresponding to the second pixel value; Performing pixel recognition on the binary image to obtain each image pixel constituting the binary image; In response to at least one image pixel point corresponding to the first pixel value being included in each image pixel point, determining whether the at least one image pixel point is located in an image frame corresponding to the binary image; In response to the at least one image pixel point not being located in the image frame of the binary image, determining a pixel region composed of the at least one image pixel point in the binary image, and performing image cutting on a collected image corresponding to the binary image based on the pixel region to obtain a cut image corresponding to the photovoltaic target; Performing image recognition on the cutting image to determine a bolt image and a cable image respectively corresponding to a bolt structure and a cable structure in the cutting image; Performing a bolt evaluation based on the bolt dimension based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image; Performing a cable evaluation based on the cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image; The bolt evaluation value and the cable evaluation value are summed and calculated to obtain an early warning evaluation value corresponding to the photovoltaic target.

[0009] Optionally, in the method according to the present invention, performing bolt evaluation based on bolt dimensions based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image includes: Acquire each bolt structure included in the bolt image, and determine whether there is a nut structure having a threaded connection relationship with each bolt structure based on the bolt image; In response to any of the bolt structures not having a threaded connection relationship, determining the retrieved preset bolt threshold as a bolt evaluation value corresponding to the bolt image, wherein a warning evaluation value obtained by fusion calculation based on the preset bolt threshold and the cable evaluation value is less than the preset warning threshold; In response to each bolt structure having a nut structure in threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure in threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance; The bolt looseness values ​​corresponding to each bolt structure are averaged to obtain a bolt evaluation value corresponding to the bolt image.

[0010] Optionally, in the method according to the present invention, in response to each bolt structure having a nut structure with a threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure with a threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance, including: In response to any of the bolt structures having a threaded connection relationship with a nut structure, obtaining a thread distance between the bolt structure and the nut structure, and determining a bolt looseness value corresponding to the bolt structure based on the thread distance; In response to any of the bolt structures and the plurality of nut structures having a threaded connection relationship at the same time, the plurality of nut structures are assigned different level coefficients from large to small based on the respective proximity between the plurality of nut structures and the bolt structure; Obtaining thread distances between each nut structure and the bolt structure respectively, and determining original looseness values ​​corresponding to each nut structure respectively based on the thread distances; Each original looseness value is multiplied by the corresponding coefficients of different levels to obtain the looseness values ​​of the bolts corresponding to the bolt structures.

[0011] Optionally, in the method according to the present invention, performing a cable evaluation based on a cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image includes: Acquire the cable structure included in the cable image, and divide the cable structure into equal intervals based on the extension direction of the cable structure to obtain each division point; Controlling the elevation unit installed on the unmanned boat to perform elevation detection on each divided point respectively, and obtaining the elevation value of each point corresponding to each divided point; The elevation value of each point is calculated based on the absolute value with the retrieved reference elevation value to obtain each absolute elevation difference; Determine each division point whose corresponding absolute elevation difference is within the retrieved benchmark difference interval as having a flat attribute, and determine each division point whose corresponding absolute elevation difference is outside the retrieved benchmark difference interval as having a drooping attribute; Determining independently existing drooping sections formed by continuously arranged dividing points having drooping properties along the extension direction of the cable structure; Based on all the divided points included in each droop section, the elevation value of the point with the smallest corresponding value is determined as the droop minimum value, and the obtained droop minimum values ​​corresponding to each droop section are numerically compared with the retrieved preset minimum threshold value; In response to any sag minimum value being less than the preset minimum threshold, determining the retrieved preset cable threshold as the cable evaluation value corresponding to the cable image, wherein the early warning evaluation value obtained by fusion calculation based on the bolt evaluation value and the preset cable threshold is less than the preset early warning threshold; In response to all droop minimum values ​​being greater than or equal to the preset minimum threshold, respectively calculating the mean of each droop minimum value and determining the number of all divided points with droop attributes to obtain the minimum elevation mean and the number of points; The minimum elevation mean and the number of points are respectively multiplied by the retrieved elevation normalization value and the number normalization value, and the obtained first cable value and the second cable value are summed to obtain: Optionally, in the method according to the present invention, In response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, controlling the unmanned boat to return from the preset position to the current position, and obtaining a moving direction based on the detection unit, including: Acquire a collection center point based on a video collection frame corresponding to the detection unit, and determine a display pixel value corresponding to a second display mode; In response to determining, based on the video detection, that a video pixel value corresponding to a video pixel point at the acquisition center point is the same as the display pixel value, determining that the unmanned boat corresponding to the second display mode has arrived at a preset position; Controlling the unmanned boat to return from the preset position to the current position, and obtaining a corresponding return time period; Acquire a video segment corresponding to the return period based on the detection unit, and split the video segment to obtain video frames arranged based on time sequence; Determine any video frame as a reference frame, and perform pixel extraction on all remaining video frames to obtain video pixel points corresponding to display pixel values; Fill each video pixel point to the corresponding pixel position of the reference frame, and establish a first image coordinate system with the acquisition center point as the origin; The number of pixels of each video pixel point located in different quadrants is determined based on the first image coordinate system, and the quadrant with the largest number of corresponding pixels is determined as the moving direction collected by the detection unit.

[0012] Optionally, in the method according to the present invention, image marking is performed on a photovoltaic image corresponding to the photovoltaic target in a sea area upper view corresponding to the sea area based on the moving direction, comprising: Acquire a sea area top view corresponding to the sea area, and determine a unit image corresponding to the detection unit based on the sea area top view; Establishing a second image coordinate system corresponding to the upper view of the sea area with the image center point of the unit image as the origin, and determining a mapping quadrant having the same quadrant position as the moving direction based on the second image coordinate system; Determine each photovoltaic image corresponding to each photovoltaic target based on the sea area upper view, and determine all photovoltaic images located in the mapping quadrant as a positioning mark group; The image distances between each photovoltaic image in the positioning mark group and the image center point are respectively obtained, and the image contour of the photovoltaic image corresponding to the smallest distance is image marked with a preset pixel value.

[0013] According to another aspect of the present invention, there is provided an edge computing data on-site real-time analysis system based on multi-parameter IoT technology, comprising: The unit setting module is configured to generate a detection route that passes through each photovoltaic target in sequence based on each photovoltaic target that is arranged in an array and is located in the sea area, and to pre-set corresponding detection units between each photovoltaic target that has an adjacent staggered relationship determined based on the detection route; An image acquisition module is configured to control the unmanned boat to sequentially acquire images of each photovoltaic target based on the detection route, and in response to determining based on the acquisition result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, trigger the light unit pre-installed on the unmanned boat to switch from the first display mode to the second display mode; A trigger detection module is configured to control the unmanned boat to move from a current position toward a detection unit closest to the unmanned boat, and trigger the detection unit to perform video detection; a control return module, configured to control the unmanned boat to return from the preset position to the current position in response to determining that the unmanned boat corresponding to the second display mode reaches the preset position based on the video detection, and obtain the moving direction of the unmanned boat collected by the detection unit; The image marking module is configured to mark the image portion corresponding to the photovoltaic target in the sea area upper view corresponding to the sea area based on the moving direction, control the unmanned boat to continue to collect images of the remaining photovoltaic targets based on the detection route in response to the unmanned boat arriving at the current position, and trigger the light unit to switch from the second display mode to the first display mode.

[0014] According to the solution of the present invention, the server can generate a detection route that passes through each photovoltaic target in sequence according to each photovoltaic target arranged in an array on the sea area. In order to facilitate the subsequent more accurate detection of each photovoltaic target, the server will pre-set the corresponding detection unit between each photovoltaic target with an adjacent staggered relationship according to the detection route. Then, the server will control the unmanned boat to collect images of each photovoltaic target in sequence according to the detection route. When the server determines that the warning evaluation value of any photovoltaic target is greater than the preset warning threshold according to the collection result, it will trigger the light unit pre-installed on the unmanned boat to be converted from the first display mode to the second display mode, so that the subsequent server can monitor the unmanned boat accordingly. Then, the server will control the unmanned boat to move from the current position toward the detection unit closest to the unmanned boat, and trigger the detection unit to perform video detection. At this time, the server will mark the image portion corresponding to the photovoltaic target in the sea area view corresponding to the sea area according to the acquired moving direction, so that the subsequent corresponding personnel can quickly locate each photovoltaic target whose warning evaluation value is greater than the preset warning threshold by viewing the sea area view after image marking. When the unmanned boat reaches the current position, the server will control the unmanned boat to continue to collect images of the remaining photovoltaic targets according to the detection route, and trigger the light unit to switch from the second display mode to the first display mode. The present invention can improve the image acquisition efficiency to a certain extent while ensuring the accuracy of calculating the early warning evaluation value. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flow chart of a method for intelligently controlling a microgrid device according to an embodiment of the present invention is shown; Figure 2 A schematic diagram of a detection unit according to an embodiment of the present invention is shown; Figure 3 A schematic diagram of a binarized image according to an embodiment of the present invention is shown; Figure 4 A structural block diagram of an intelligent control system for microgrid equipment according to another embodiment of the present invention is shown. DETAILED DESCRIPTION

[0016] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0017] In order to solve the problems existing in the above background technology, the inventor proposes the solution of the present invention. One embodiment of the present invention provides a method for real-time analysis of edge computing data based on multi-parameter Internet of Things technology, which can be executed in a computing device.

[0018] Figure 1 A flowchart of an on-site real-time analysis method of edge computing data based on multi-parameter Internet of Things technology according to an embodiment of the present invention is shown, and the method is suitable for execution in a computing device.

[0019] like Figure 1 As shown, the edge computing data on-site real-time analysis method based on multi-parameter IoT technology proposed in this embodiment begins with step S102. In step S102, the following contents are included: Based on the photovoltaic targets arranged in an array on the sea area, a detection route passing through the photovoltaic targets in sequence is generated, and corresponding detection units are pre-arranged between the photovoltaic targets with adjacent interlaced relationships determined based on the detection route.

[0020] For example, in this embodiment, a plurality of photovoltaic targets arranged in an array are arranged in the sea area, and therefore, the server generates a detection route that can pass through each photovoltaic target in sequence.

[0021] In order to facilitate more accurate detection of each photovoltaic target in the future, the server will pre-set corresponding detection units between each photovoltaic target with adjacent staggered relationship according to the detection route.

[0022] Furthermore, the above-mentioned “generating a detection route that passes through each photovoltaic target in sequence based on each photovoltaic target arranged in an array and located in the sea area, and pre-setting corresponding detection units between each photovoltaic target having an adjacent staggered relationship determined based on the detection route” also includes the following steps: Determine a first arrangement number of photovoltaic targets in an array arrangement in a first arrangement direction and a second arrangement number in a second arrangement direction, wherein the first arrangement direction is perpendicular to the second arrangement direction; Compare the number of the first arrangement with the number of the second arrangement, and based on the comparison result, determine the arrangement direction with the larger corresponding number as the primary moving direction, and determine the arrangement direction with the smaller corresponding number as the secondary moving direction; Grouping the photovoltaic targets based on the secondary moving direction to obtain primary groups, and sorting the photovoltaic targets in the same primary group based on the primary moving direction to obtain a first target sequence; Among the photovoltaic targets, any photovoltaic target at a corner position is determined as a starting target; Taking the main level group including the starting target as the starting point, sorting the main level groups to obtain a group sequence, and based on the group sequence, reversely modifying the first target sequence of all main level groups corresponding to even numbers to obtain a second target sequence; Based on the grouping sequence and each target sequence, a detection route is determined that takes the starting area as the detection starting point and passes through each photovoltaic target in sequence, and each primary grouping at an adjacent position corresponding to the detection route is determined as the same staggered determination set; Based on a preset interleaving determination strategy, each photovoltaic target in the same interleaving determination set is interleaved and determined, and corresponding detection units are pre-arranged between each photovoltaic target that has the same adjacent interleaving relationship.

[0023] For example, in this embodiment, the server first determines the first arrangement number of each photovoltaic target in the array arrangement in the first arrangement direction and the second arrangement number in the second arrangement direction, and the first arrangement direction is perpendicular to the second arrangement direction. For example, the first arrangement direction is east-west and the second arrangement direction is north-south.

[0024] Next, the server will compare the quantity of the first arrangement with the quantity of the second arrangement, and determine the arrangement direction with the larger corresponding quantity as the primary moving direction and the arrangement direction with the smaller corresponding quantity as the secondary moving direction according to the comparison result, and then group each photovoltaic target according to the secondary moving direction to obtain each primary grouping.

[0025] For example, if the first arrangement number in the first arrangement direction is greater than the second arrangement number in the second arrangement direction, the server will determine the first arrangement direction as the primary moving direction and the second arrangement direction as the secondary moving direction, and group each photovoltaic target according to the secondary moving direction, thereby obtaining each primary grouping arranged according to the secondary moving direction.

[0026] Next, the server will sort the photovoltaic targets in the same main-level group according to the main-level moving direction to obtain a first target sequence, and then determine any photovoltaic target at a corner position among the photovoltaic targets as the starting target.

[0027] At this time, the server will sort each main level group starting from the main level group including the starting target to obtain the corresponding group sequence, and then reversely modify the first target sequence of all main level groups corresponding to even numbers according to the group sequence to obtain the second target sequence.

[0028] Then, the server will determine a detection route that takes the starting area as the detection starting point and passes through each photovoltaic target in sequence according to the group sequence and each target sequence, and determine each primary group in adjacent positions as the same staggered determination set according to the detection route, and then stagger the photovoltaic targets in the same staggered determination set based on the preset staggered determination strategy. Then, the server will pre-set the corresponding detection units between each photovoltaic target that has the same adjacent staggered relationship.

[0029] This embodiment can determine the main moving direction by comparing the first arrangement number with the second arrangement number, and further determine the detection route, so that the number of turns of the subsequent unmanned boat when collecting each photovoltaic target according to the detection route can be minimized, thereby improving the collection efficiency of the unmanned boat.

[0030] Furthermore, the above-mentioned “interleaving the photovoltaic targets corresponding to the same interleaving determination set based on a preset interleaving determination strategy, and pre-setting corresponding detection units between the photovoltaic targets having the same adjacent interleaving relationship” also includes the following steps: Determine two primary groups in the same interleaved determination set as a first group and a second group respectively; Along the main-stage moving direction, the photovoltaic targets in the first group and the photovoltaic targets in the second group are respectively interlaced and connected one by one, so as to obtain target connection lines respectively connecting the photovoltaic targets having different adjacent interlaced relationships; Determine the intersection of each target connecting line, and obtain the corresponding intersection points of each line segment based on the determination result; The corresponding detection units are set based on the intersection points of each line segment.

[0031] For example, in this embodiment, the server determines two primary groups in the same interleaved determination set as the first group and the second group respectively.

[0032] In order to ensure the detection accuracy of the set detection unit, the server will interlace the photovoltaic targets in the first group with the photovoltaic targets in the second group one by one along the main-level moving direction, thereby obtaining target connection lines that connect the photovoltaic targets with different adjacent interlace relationships.

[0033] Then, the server will determine the intersection of each target connection line, thereby obtaining the corresponding line segment intersection points according to the determination results, and then set the corresponding detection unit based on each line segment intersection point, such as Figure 2 as shown in the circle.

[0034] In step S104, the following contents are included: The unmanned boat is controlled to collect images of each photovoltaic target in sequence based on the detection route, and in response to determining based on the collection result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, the light unit pre-installed on the unmanned boat is triggered to switch from the first display mode to the second display mode.

[0035] For example, in this embodiment, the server will control the unmanned boat to collect images of each photovoltaic target in turn along the detection route. When the server determines that the warning evaluation value of any photovoltaic target is greater than the preset warning threshold based on the collection results, it will trigger the light unit pre-installed on the unmanned boat to switch from the first display mode to the second display mode, for example, the light unit switches from turning on the red light to turning on the blue light, so as to facilitate the subsequent server to perform corresponding monitoring of the unmanned boat.

[0036] Furthermore, the above-mentioned “controlling the unmanned boat to sequentially collect images of each photovoltaic target based on the detection route” further includes the following steps: Controlling the unmanned boat to collect images based on the detection route, and performing binarization processing on the acquired collected images to obtain a binarized image, wherein the binarized image includes photovoltaic pixel points corresponding to the first pixel value and / or noise pixel points corresponding to the second pixel value; Performing pixel recognition on the binary image to obtain each image pixel constituting the binary image; In response to at least one image pixel point corresponding to the first pixel value being included in each image pixel point, determining whether the at least one image pixel point is located in an image frame corresponding to the binary image; In response to the at least one image pixel point not being located in the image frame of the binary image, determining a pixel region composed of the at least one image pixel point in the binary image, and performing image cutting on a collected image corresponding to the binary image based on the pixel region to obtain a cut image corresponding to the photovoltaic target; Performing image recognition on the cutting image to determine a bolt image and a cable image respectively corresponding to a bolt structure and a cable structure in the cutting image; Performing a bolt evaluation based on the bolt dimension based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image; Performing a cable evaluation based on the cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image; The bolt evaluation value and the cable evaluation value are summed and calculated to obtain an early warning evaluation value corresponding to the photovoltaic target.

[0037] For example, in this embodiment, in order to ensure that the unmanned boat can capture a complete captured image when capturing images of the photovoltaic target, the photovoltaic size of the photovoltaic target is smaller than the frame size of the image capture frame of the corresponding unmanned boat.

[0038] The server will control the unmanned boat to collect images according to the detection route, and perform binarization processing on the acquired collected images to obtain a binary image. Since the acquired image is collected in real time when the unmanned boat moves along the detection route, it is possible that a complete photovoltaic target or a partial photovoltaic target appears in the collected image. Therefore, the binary image may include photovoltaic pixels corresponding to the first pixel value and noise pixels corresponding to the second pixel value.

[0039] Then, the server will perform pixel recognition on the binary image to obtain the individual image pixels that make up the binary image. Since the photovoltaic size of the photovoltaic target is smaller than the frame size of the image acquisition frame of the corresponding unmanned boat, when the unmanned boat fully acquires the photovoltaic target, the individual image pixels of the binary image will not be located on the image frame of the binary image. That is, if there are image pixels located on the image frame of the binary image, the unmanned boat has only acquired part of the image of the photovoltaic target. Figure 3-1 shown.

[0040] Therefore, when at least one image pixel corresponding to the first pixel value is included in each image pixel point, the server will determine whether the at least one image pixel point is located on the image frame of the binary image. When the at least one image pixel point corresponding to the first pixel value is not located in the image frame of the binary image, it means that the unmanned boat has collected a complete image of the photovoltaic target, such as Figure 3-2 Therefore, the server determines a pixel region composed of at least one image pixel point in the binary image, and performs image segmentation on the collected image corresponding to the binary image according to the pixel region, thereby obtaining a segmentation image corresponding to the photovoltaic target.

[0041] Since the loosening or falling off of the nut structure of the bolt structure on the photovoltaic target and the sagging of the cable structure indicate that the photovoltaic target has certain quality problems, the server will determine the bolt image of the corresponding bolt structure and the cable image of the corresponding cable structure in the cut image by performing image recognition on the cut image. Then, a bolt evaluation based on the bolt dimension is performed based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image, and a cable evaluation based on the cable dimension is performed based on the cable image to obtain a cable evaluation value corresponding to the cable image. Finally, the server will sum the bolt evaluation value and the cable evaluation value to obtain a warning evaluation value corresponding to the photovoltaic target.

[0042] In this embodiment, the captured image is binarized to obtain photovoltaic pixel points corresponding to the first pixel value and noise pixel points corresponding to the second pixel value, which can not only simplify the image processing process but also improve the recognition of the photovoltaic target in the captured image.

[0043] Furthermore, the above-mentioned “performing a bolt evaluation based on the bolt dimension based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image” further includes the following steps: Acquire each bolt structure included in the bolt image, and determine whether there is a nut structure having a threaded connection relationship with each bolt structure based on the bolt image; In response to any of the bolt structures not having a threaded connection relationship, determining the retrieved preset bolt threshold as a bolt evaluation value corresponding to the bolt image, wherein a warning evaluation value obtained by fusion calculation based on the preset bolt threshold and the cable evaluation value is less than the preset warning threshold; In response to each bolt structure having a nut structure in threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure in threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance; The bolt looseness values ​​corresponding to each bolt structure are averaged to obtain a bolt evaluation value corresponding to the bolt image.

[0044] For example, in this embodiment, the server first obtains each bolt structure in the bolt image, and then determines whether there is a nut structure having a threaded connection relationship with each bolt structure according to the bolt image.

[0045] When any bolt structure does not have a threaded connection relationship, it means that the nut structure that was previously threaded with the bolt structure has fallen off the bolt structure, which means that a major quality problem has occurred. Therefore, the server will determine the retrieved preset bolt threshold as the bolt evaluation value corresponding to the bolt image. Since the preset bolt threshold indicates that the quality problem is relatively large, the warning evaluation value obtained by subsequently merging the preset bolt threshold with the cable evaluation value will be smaller than the preset warning threshold, which is convenient for the subsequent management end to conduct corresponding checks on photovoltaic targets with major quality problems.

[0046] When each bolt structure has a nut structure with a threaded connection relationship, the server will respectively obtain the thread distances between each thread structure and the corresponding nut structure with a threaded connection relationship. Since the longer the thread distance, the more serious the looseness of the nut structure, that is, the greater the quality problem, the server will determine the looseness values ​​of each bolt corresponding to each bolt structure based on each thread distance.

[0047] Finally, the server will average the looseness values ​​of the bolts corresponding to the bolt structures to obtain the bolt evaluation value corresponding to the bolt image.

[0048] Furthermore, the above-mentioned “in response to each bolt structure having a nut structure with a threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure with a threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance” also includes the following steps: In response to any of the bolt structures having a threaded connection relationship with a nut structure, obtaining a thread distance between the bolt structure and the nut structure, and determining a bolt looseness value corresponding to the bolt structure based on the thread distance; In response to any of the bolt structures and the plurality of nut structures having a threaded connection relationship at the same time, the plurality of nut structures are assigned different level coefficients from large to small based on the respective proximity between the plurality of nut structures and the bolt structure; Obtaining thread distances between each nut structure and the bolt structure respectively, and determining original looseness values ​​corresponding to each nut structure respectively based on the thread distances; Each original looseness value is multiplied by the corresponding coefficients of different levels to obtain the looseness values ​​of the bolts corresponding to the bolt structures.

[0049] For example, in this embodiment, when any bolt structure has a threaded connection relationship with a nut structure, the server will obtain the thread distance between the bolt structure and the nut structure, and then call the preset distance interval determination table, in which there are different thread distances and various bolt looseness values ​​corresponding to different thread distances. The server will determine the bolt looseness value corresponding to the obtained thread distance in the preset distance interval determination table, that is, determine the bolt looseness value corresponding to the bolt structure.

[0050] When any bolt structure has a threaded connection with multiple nut structures at the same time, when the nut structure closest to the bolt structure is loose, it indicates that the nut structure is loosening more seriously. Therefore, the server will assign different level coefficients from large to small to the multiple nut structures according to the degree of proximity between the multiple nut structures and the bolt structure. For example, if there are three nut structures, the level of the nut structure closest to the bolt structure is 1, and its level coefficient is 2; the level of the nut structure in the middle is 2, and its level coefficient is 1.5; the level of the nut structure farthest from the bolt structure is 3, and its level coefficient is 1.2.

[0051] Then, the server will obtain the thread distances between each nut structure and the bolt structure respectively, and determine the original looseness values ​​corresponding to each nut structure according to the thread distances, and then multiply the original looseness values ​​with the corresponding different level coefficients to obtain the bolt looseness values ​​corresponding to each bolt structure.

[0052] Furthermore, the above-mentioned “performing a cable evaluation based on the cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image” further includes the following steps: Acquire the cable structure included in the cable image, and divide the cable structure into equal intervals based on the extension direction of the cable structure to obtain each division point; Controlling the elevation unit installed on the unmanned boat to perform elevation detection on each divided point respectively, and obtaining the elevation value of each point corresponding to each divided point; The elevation value of each point is calculated based on the absolute value with the retrieved reference elevation value to obtain each absolute elevation difference; Determine each division point whose corresponding absolute elevation difference is within the retrieved benchmark difference interval as having a flat attribute, and determine each division point whose corresponding absolute elevation difference is outside the retrieved benchmark difference interval as having a drooping attribute; Determining independently existing drooping sections formed by continuously arranged dividing points having drooping properties along the extension direction of the cable structure; Based on all the divided points included in each droop section, the elevation value of the point with the smallest corresponding value is determined as the droop minimum value, and the obtained droop minimum values ​​corresponding to each droop section are numerically compared with the retrieved preset minimum threshold value; In response to any sag minimum value being less than the preset minimum threshold, determining the retrieved preset cable threshold as the cable evaluation value corresponding to the cable image, wherein the early warning evaluation value obtained by fusion calculation based on the bolt evaluation value and the preset cable threshold is less than the preset early warning threshold; In response to all droop minimum values ​​being greater than or equal to the preset minimum threshold, respectively calculating the mean of each droop minimum value and determining the number of all divided points with droop attributes to obtain the minimum elevation mean and the number of points; The minimum elevation mean and the number of points are respectively multiplied by the retrieved elevation normalization value and the number normalization value, and the obtained first cable value and the second cable value are summed to obtain a cable evaluation value corresponding to the cable image.

[0053] For example, in this embodiment, the server will obtain the cable structure included in the cable image, and divide the cable structure into equal intervals according to the extension direction of the cable structure, thereby obtaining each division point. Then, the server will control the elevation unit installed on the unmanned boat to perform elevation detection on each division point, thereby obtaining the elevation value of each point corresponding to each division point, and then perform a difference calculation based on the absolute value between the elevation value of each point and the retrieved reference elevation value, thereby obtaining each absolute elevation difference.

[0054] Then, the server will compare the absolute elevation difference with the retrieved benchmark difference interval. When the absolute elevation difference is within the benchmark difference interval, it means that there is no obvious sagging at the division points corresponding to the absolute elevation difference, that is, they basically remain flat. Therefore, the server will determine that each division point whose corresponding absolute elevation difference is within the retrieved benchmark difference interval has a flat attribute.

[0055] When the absolute elevation difference is outside the benchmark difference interval, it means that each division point corresponding to the absolute elevation difference has obvious drooping. Therefore, the server will determine each division point whose corresponding absolute elevation difference is outside the retrieved benchmark difference interval as having a drooping attribute.

[0056] Since multiple sagging sections may appear in a cable structure, the server will determine the sagging sections formed by the consecutively arranged division points with sagging properties along the extension direction of the cable structure, and the determined sagging sections exist independently.

[0057] Next, the server will determine the point elevation value with the smallest corresponding value among all the divided points included in each droop section as the droop minimum value, and compare the obtained droop minimum values ​​corresponding to each droop section with the retrieved preset minimum threshold value.

[0058] When any of the minimum droop values ​​is less than the preset minimum threshold, it indicates that the cable structure has a large degree of droop, which means that a large quality problem has occurred. Therefore, the server will determine the retrieved preset cable threshold as the cable evaluation value corresponding to the cable image. Since the preset cable threshold indicates that the quality problem is large, the warning evaluation value obtained by subsequently merging the bolt evaluation value with the preset cable threshold is smaller than the preset warning threshold, which is convenient for the subsequent management end to conduct corresponding checks on photovoltaic targets with large quality problems.

[0059] When all the droop minimum values ​​are greater than or equal to the preset minimum threshold, the server will calculate the average of each droop minimum value to obtain the minimum elevation average, and then determine the number of all the points with droop attributes to obtain the number of points. The larger the minimum elevation average, the greater the droop degree; the more points, the more droop parts.

[0060] Therefore, the server will multiply the minimum elevation mean and the number of points with the retrieved elevation normalization value and the number normalization value respectively to obtain the first cable value and the second cable value. Since the higher the cable evaluation value, the better the cable quality, the corresponding cable evaluation value can be obtained by the following formula.

[0061] , Among them, p is the cable evaluation value, n is the first cable value, and m is the second cable value.

[0062] In step S106, the following contents are included: The unmanned boat is controlled to move from a current position toward a detection unit closest to the unmanned boat, and the detection unit is triggered to perform video detection.

[0063] For example, in this embodiment, the server controls the unmanned boat to move from the current position toward the detection unit closest to the unmanned boat, and triggers the detection unit to perform video detection.

[0064] In step S108, the following contents are included: In response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, the unmanned boat is controlled to return from the preset position to the current position, and the moving direction of the unmanned boat collected by the detection unit is obtained.

[0065] For example, in this embodiment, when the server determines based on video detection that the unmanned boat corresponding to the second display mode has reached a preset position, it will control the unmanned boat to return from the preset position to the current position, thereby obtaining the moving direction of the unmanned boat collected by the detection unit.

[0066] Furthermore, the above-mentioned “in response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, controlling the unmanned boat to return from the preset position to the current position, and obtaining the moving direction of the unmanned boat collected by the detection unit” also includes the following steps: Acquire a collection center point based on a video collection frame corresponding to the detection unit, and determine a display pixel value corresponding to a second display mode; In response to determining, based on the video detection, that a video pixel value corresponding to a video pixel point at the acquisition center point is the same as the display pixel value, determining that the unmanned boat corresponding to the second display mode has arrived at a preset position; Controlling the unmanned boat to return from the preset position to the current position, and obtaining a corresponding return time period; Acquire a video segment corresponding to the return period based on the detection unit, and split the video segment to obtain video frames arranged based on time sequence; Determine any video frame as a reference frame, and perform pixel extraction on all remaining video frames to obtain video pixel points corresponding to display pixel values; Fill each video pixel point to the corresponding pixel position of the reference frame, and establish a first image coordinate system with the acquisition center point as the origin; The number of pixels of each video pixel point located in different quadrants is determined based on the first image coordinate system, and the quadrant with the largest number of corresponding pixels is determined as the moving direction collected by the detection unit.

[0067] For example, in this embodiment, the server will first obtain the acquisition center point of the video acquisition frame of the detection unit, and determine the display pixel value corresponding to the second display method. For example, the second display method is to turn on the blue light of the unmanned boat, that is, the display pixel value corresponding to the second display method is the display pixel value corresponding to blue.

[0068] When the server determines that the video pixel value corresponding to the video pixel point at the acquisition center point is the same as the display pixel value based on the video detection, it means that the unmanned boat has reached the preset position. At this time, the server will determine that the unmanned boat corresponding to the second display mode has reached the preset position.

[0069] Then, the server will control the unmanned boat to return from the preset position to the current position, and obtain the return time period corresponding to this return process, and then obtain the video segment corresponding to the return time period according to the detection unit, and split the video segment to obtain each video frame arranged based on time sequence.

[0070] Since the unmanned boat is in a moving state during the return process, the server will first determine any one of the video frames as the reference frame, and then extract pixels from all the remaining video frames to obtain the corresponding display pixel values ​​of each video frame.

[0071] Then, the server will fill each video pixel point to the corresponding pixel position in the reference frame, that is, obtain the return trajectory of the unmanned boat during the return process. Next, the server will establish a first image coordinate system with the acquisition center point as the origin, and then determine the number of pixels of each video pixel point located in different quadrants according to the first image coordinate system. The larger the number of pixels in a quadrant, the more the unmanned boat is moving towards that quadrant, so the server will determine the quadrant with the largest number of corresponding pixels as the moving direction of the unmanned boat collected by the detection unit.

[0072] In step S110, the following contents are included: Based on the moving direction, the image portion corresponding to the photovoltaic target is marked in the sea area view corresponding to the sea area. In response to the unmanned boat arriving at the current position, the unmanned boat is controlled to continue to collect images of the remaining photovoltaic targets based on the detection route, and the light unit is triggered to switch from the second display mode to the first display mode.

[0073] For example, in this embodiment, the server will mark the image portion corresponding to the photovoltaic target in the sea area view corresponding to the sea area according to the acquired moving direction, so that subsequent corresponding personnel can quickly locate each photovoltaic target whose warning evaluation value is greater than the preset warning threshold by checking the sea area view after image marking.

[0074] When the unmanned boat reaches the current position, the server will control the unmanned boat to continue to collect images of the remaining photovoltaic targets according to the detection route, and trigger the light unit to switch from the second display mode to the first display mode.

[0075] Furthermore, the above-mentioned “marking the image portion corresponding to the photovoltaic target in the upper view of the sea area corresponding to the sea area based on the moving direction” further includes the following steps: Acquire a sea area top view corresponding to the sea area, and determine a unit image corresponding to the detection unit based on the sea area top view; Establishing a second image coordinate system corresponding to the upper view of the sea area with the image center point of the unit image as the origin, and determining a mapping quadrant having the same quadrant position as the moving direction based on the second image coordinate system; Determine each photovoltaic image corresponding to each photovoltaic target based on the sea area upper view, and determine all photovoltaic images located in the mapping quadrant as a positioning mark group; The image distances between each photovoltaic image in the positioning mark group and the image center point are respectively obtained, and the image contour of the photovoltaic image corresponding to the smallest distance is image marked with a preset pixel value.

[0076] For example, in this embodiment, the server obtains a sea area upper view corresponding to the sea area, thereby determining a unit image corresponding to the detection unit in the sea area upper view, and then establishes a second image coordinate system corresponding to the sea area upper view with the image center point of the unit image as the origin, and determines a mapping quadrant having the same quadrant position as the moving direction in the second image coordinate system.

[0077] Then, the server will determine the photovoltaic images corresponding to each photovoltaic target in the upper view of the sea area, and then determine all photovoltaic images in the mapping quadrant as a positioning mark group. Since there may be multiple photovoltaic images in the positioning mark group, the server will obtain the image distance between each photovoltaic image in the positioning mark group and the image center point, so as to mark the image contour of the photovoltaic image with the smallest corresponding distance with a preset pixel value.

[0078] According to the solution of the present invention, the server can generate a detection route that passes through each photovoltaic target in sequence according to each photovoltaic target arranged in an array on the sea area. In order to facilitate the subsequent more accurate detection of each photovoltaic target, the server will pre-set the corresponding detection unit between each photovoltaic target with an adjacent staggered relationship according to the detection route. Then, the server will control the unmanned boat to collect images of each photovoltaic target in sequence according to the detection route. When the server determines that the warning evaluation value of any photovoltaic target is greater than the preset warning threshold according to the collection result, it will trigger the light unit pre-installed on the unmanned boat to be converted from the first display mode to the second display mode, so that the subsequent server can monitor the unmanned boat accordingly. Then, the server will control the unmanned boat to move from the current position toward the detection unit closest to the unmanned boat, and trigger the detection unit to perform video detection. At this time, the server will mark the image portion corresponding to the photovoltaic target in the sea area view corresponding to the sea area according to the acquired moving direction, so that the subsequent corresponding personnel can quickly locate each photovoltaic target whose warning evaluation value is greater than the preset warning threshold by viewing the sea area view after image marking. When the unmanned boat reaches the current position, the server will control the unmanned boat to continue to collect images of the remaining photovoltaic targets according to the detection route, and trigger the light unit to switch from the second display mode to the first display mode. The present invention can improve the image acquisition efficiency to a certain extent while ensuring the accuracy of calculating the early warning evaluation value.

[0079] Another embodiment of the present invention provides an edge computing data on-site real-time analysis system based on multi-parameter IoT technology. Figure 4 As the corresponding system block diagram, the system includes: The unit setting module is configured to generate a detection route that passes through each photovoltaic target in sequence based on each photovoltaic target that is arranged in an array and is located in the sea area, and to pre-set corresponding detection units between each photovoltaic target that has an adjacent staggered relationship determined based on the detection route; An image acquisition module is configured to control the unmanned boat to sequentially acquire images of each photovoltaic target based on the detection route, and in response to determining based on the acquisition result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, trigger the light unit pre-installed on the unmanned boat to switch from the first display mode to the second display mode; A trigger detection module is configured to control the unmanned boat to move from a current position toward a detection unit closest to the unmanned boat, and trigger the detection unit to perform video detection; a control return module, configured to control the unmanned boat to return from the preset position to the current position in response to determining that the unmanned boat corresponding to the second display mode reaches the preset position based on the video detection, and obtain the moving direction of the unmanned boat collected by the detection unit; The image marking module is configured to mark the image portion corresponding to the photovoltaic target in the sea area upper view corresponding to the sea area based on the moving direction, control the unmanned boat to continue to collect images of the remaining photovoltaic targets based on the detection route in response to the unmanned boat arriving at the current position, and trigger the light unit to switch from the second display mode to the first display mode.

[0080] In the description provided herein, algorithms and displays are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the examples of the present invention. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to implement the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the preferred embodiment of the present invention.

[0081] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0082] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof.

[0083] Those skilled in the art will appreciate that the modules or units or components of the devices in the examples disclosed herein may be arranged in the devices described in the embodiment, or alternatively may be located in one or more devices different from the devices in the examples. The modules in the foregoing examples may be combined into one module or may be divided into multiple submodules.

[0084] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore may be divided into a plurality of submodules or subunits or subcomponents.

[0085] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present invention and to form different embodiments.

[0086] In addition, some of the embodiments are described herein as methods or combinations of method elements that can be implemented by a processor of a computer system or by other devices that perform the functions. Therefore, a processor with necessary instructions for implementing the method or method elements forms a device for implementing the method or method elements. In addition, the elements described herein of the device embodiments are examples of devices for implementing the functions performed by the elements for the purpose of implementing the invention.

[0087] As used herein, unless otherwise specified, the use of ordinal numbers "first," "second," "third," etc. to describe common objects merely indicates that different instances of similar objects are involved, and is not intended to imply that the objects so described must have a given order in time, space, order, or in any other manner.

[0088] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is primarily selected for readability and instructional purposes, rather than for explaining or limiting the subject matter of the present invention.

Claims

1. A method for real-time analysis of edge computing data based on multi-parameter IoT technology, characterized in that: The following steps are involved: Based on the photovoltaic targets arranged in arrays on the sea area, a detection route passing through each photovoltaic target in sequence is generated, and corresponding detection units are respectively pre-set between each photovoltaic target having an adjacent staggered relationship determined based on the detection route; Controlling the unmanned boat to sequentially collect images of each photovoltaic target based on the detection route, and in response to determining based on the collection result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, triggering a light unit pre-installed on the unmanned boat to switch from a first display mode to a second display mode; Controlling the unmanned boat to move from a current position toward a detection unit closest to the unmanned boat, and triggering the detection unit to perform video detection; In response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, controlling the unmanned boat to return from the preset position to the current position, and acquiring a moving direction of the unmanned boat based on the information collected by the detection unit; Based on the moving direction, the image portion corresponding to the photovoltaic target is marked in the sea area view corresponding to the sea area. In response to the unmanned boat arriving at the current position, the unmanned boat is controlled to continue to collect images of the remaining photovoltaic targets based on the detection route, and the light unit is triggered to switch from the second display mode to the first display mode.

2. The method for on-site real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 1 is characterized in that: Based on the photovoltaic targets arranged in arrays on the sea area, a detection route passing through the photovoltaic targets in sequence is generated, and corresponding detection units are respectively set in advance between the photovoltaic targets having adjacent interlaced relationships determined based on the detection route, including: Determine a first arrangement number of photovoltaic targets in an array arrangement in a first arrangement direction and a second arrangement number in a second arrangement direction, wherein the first arrangement direction is perpendicular to the second arrangement direction; Compare the number of the first arrangement with the number of the second arrangement, and based on the comparison result, determine the arrangement direction with the larger corresponding number as the primary moving direction, and determine the arrangement direction with the smaller corresponding number as the secondary moving direction; Grouping the photovoltaic targets based on the secondary moving direction to obtain primary groups, and sorting the photovoltaic targets in the same primary group based on the primary moving direction to obtain a first target sequence; Among the photovoltaic targets, any photovoltaic target at a corner position is determined as a starting target; Taking the main level group including the starting target as the starting point, sorting the main level groups to obtain a group sequence, and based on the group sequence, reversely modifying the first target sequence of all main level groups corresponding to even numbers to obtain a second target sequence; Based on the grouping sequence and each target sequence, a detection route is determined that takes the starting area as the detection starting point and passes through each photovoltaic target in sequence, and each primary grouping at an adjacent position corresponding to the detection route is determined as the same staggered determination set; Based on a preset interleaving determination strategy, each photovoltaic target in the same interleaving determination set is interleaved and determined, and corresponding detection units are pre-arranged between each photovoltaic target that has the same adjacent interleaving relationship.

3. The method for real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 2 is characterized in that: Based on a preset interleaving determination strategy, interleaving determination is performed on each photovoltaic target corresponding to the same interleaving determination set, and corresponding detection units are preliminarily set between each photovoltaic target having the same adjacent interleaving relationship, including: Determine two primary groups in the same interleaved determination set as a first group and a second group respectively; Along the main-stage moving direction, the photovoltaic targets in the first group and the photovoltaic targets in the second group are respectively interlaced and connected one by one, so as to obtain target connection lines respectively connecting the photovoltaic targets having different adjacent interlaced relationships; Determine the intersection of each target connecting line, and obtain the corresponding intersection points of each line segment based on the determination result; The corresponding detection units are set based on the intersection points of each line segment.

4. The method for real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 1 is characterized in that: The photovoltaic size of the photovoltaic target is smaller than the frame size of the image acquisition frame corresponding to the unmanned boat; Controlling the unmanned boat to sequentially collect images of each photovoltaic target based on the detection route includes: Controlling the unmanned boat to collect images based on the detection route, and performing binarization processing on the acquired collected images to obtain a binarized image, wherein the binarized image includes photovoltaic pixel points corresponding to the first pixel value and / or noise pixel points corresponding to the second pixel value; Performing pixel recognition on the binary image to obtain each image pixel constituting the binary image; In response to at least one image pixel point corresponding to the first pixel value being included in each image pixel point, determining whether the at least one image pixel point is located in an image frame corresponding to the binary image; In response to the at least one image pixel point not being located in the image frame of the binary image, determining a pixel region composed of the at least one image pixel point in the binary image, and performing image segmentation on a captured image corresponding to the binary image based on the pixel region to obtain a segmentation image corresponding to the photovoltaic target; Performing image recognition on the cutting image to determine a bolt image and a cable image respectively corresponding to a bolt structure and a cable structure in the cutting image; Performing a bolt evaluation based on the bolt dimension based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image; Performing a cable evaluation based on the cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image; The bolt evaluation value and the cable evaluation value are summed and calculated to obtain an early warning evaluation value corresponding to the photovoltaic target.

5. The method for on-site real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 4 is characterized in that: Performing a bolt evaluation based on the bolt dimension based on the bolt image to obtain a bolt evaluation value corresponding to the bolt image includes: Acquire each bolt structure included in the bolt image, and determine whether there is a nut structure having a threaded connection relationship with each bolt structure based on the bolt image; In response to any of the bolt structures not having a threaded connection relationship, determining the retrieved preset bolt threshold as a bolt evaluation value corresponding to the bolt image, wherein a warning evaluation value obtained by fusion calculation based on the preset bolt threshold and the cable evaluation value is less than the preset warning threshold; In response to each bolt structure having a nut structure in threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure in threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance; The bolt looseness values ​​corresponding to each bolt structure are averaged to obtain a bolt evaluation value corresponding to the bolt image.

6. The method for on-site real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 5 is characterized in that: In response to each bolt structure having a nut structure with a threaded connection relationship, obtaining each thread distance between each thread structure and the corresponding nut structure with a threaded connection relationship, and determining each bolt looseness value corresponding to each bolt structure based on each thread distance, including: In response to any of the bolt structures having a threaded connection relationship with a nut structure, obtaining a thread distance between the bolt structure and the nut structure, and determining a bolt looseness value corresponding to the bolt structure based on the thread distance; In response to any of the bolt structures and the plurality of nut structures having a threaded connection relationship at the same time, the plurality of nut structures are assigned different level coefficients from large to small based on the respective proximity between the plurality of nut structures and the bolt structure; Obtaining thread distances between each nut structure and the bolt structure respectively, and determining original looseness values ​​corresponding to each nut structure respectively based on the thread distances; Each original looseness value is multiplied by the corresponding coefficients of different levels to obtain the looseness values ​​of the bolts corresponding to the bolt structures.

7. The method for on-site real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 5 is characterized in that: Performing a cable evaluation based on the cable dimension based on the cable image to obtain a cable evaluation value corresponding to the cable image includes: Acquire a cable structure included in the cable image, and divide the cable structure into equal intervals based on an extension direction of the cable structure to obtain division points; Controlling the elevation unit installed on the unmanned boat to perform elevation detection on each divided point respectively, and obtaining the elevation value of each point corresponding to each divided point; The elevation value of each point is calculated based on the absolute value with the retrieved reference elevation value to obtain each absolute elevation difference; Determine each division point whose corresponding absolute elevation difference is within the retrieved benchmark difference interval as having a flat attribute, and determine each division point whose corresponding absolute elevation difference is outside the retrieved benchmark difference interval as having a drooping attribute; Determining independently existing drooping sections formed by continuously arranged dividing points having drooping properties along the extension direction of the cable structure; Based on all the divided points included in each droop section, the elevation value of the point with the smallest corresponding value is determined as the droop minimum value, and the obtained droop minimum values ​​corresponding to each droop section are numerically compared with the retrieved preset minimum threshold value; In response to any sag minimum value being less than the preset minimum threshold, determining the retrieved preset cable threshold as the cable evaluation value corresponding to the cable image, wherein the early warning evaluation value obtained by fusion calculation based on the bolt evaluation value and the preset cable threshold is less than the preset early warning threshold; In response to all droop minimum values ​​being greater than or equal to the preset minimum threshold, respectively calculating the mean of each droop minimum value and determining the number of all divided points with droop attributes to obtain the minimum elevation mean and the number of points; The minimum elevation mean and the number of points are respectively multiplied by the retrieved normalized elevation value and the retrieved normalized number value, and the obtained first cable value and the second cable value are summed to obtain a cable evaluation value corresponding to the cable image.

8. The method for on-site real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 1 is characterized in that: In response to determining based on the video detection that the unmanned boat corresponding to the second display mode reaches a preset position, controlling the unmanned boat to return from the preset position to the current position, and obtaining a moving direction based on the detection unit, including: Acquire a collection center point based on a video collection frame corresponding to the detection unit, and determine a display pixel value corresponding to a second display mode; In response to determining, based on the video detection, that a video pixel value corresponding to a video pixel point at the acquisition center point is the same as the display pixel value, determining that the unmanned boat corresponding to the second display mode has arrived at a preset position; Controlling the unmanned boat to return from the preset position to the current position, and obtaining a corresponding return time period; Based on the detection unit, a video segment corresponding to the return period is obtained, and the video segment is split to obtain video frames arranged based on a time sequence; Determine any video frame as a reference frame, and perform pixel extraction on all remaining video frames to obtain video pixel points corresponding to display pixel values; Fill each video pixel point to the corresponding pixel position of the reference frame, and establish a first image coordinate system with the acquisition center point as the origin; The number of pixels of each video pixel point located in different quadrants is determined based on the first image coordinate system, and the quadrant with the largest number of corresponding pixels is determined as the moving direction collected by the detection unit.

9. The method for real-time analysis of edge computing data based on multi-parameter IoT technology according to claim 8 is characterized in that: Marking a photovoltaic image corresponding to the photovoltaic target in a sea area upper view corresponding to the sea area based on the moving direction includes: Acquire a sea area top view corresponding to the sea area, and determine a unit image corresponding to the detection unit based on the sea area top view; Establishing a second image coordinate system corresponding to the upper view of the sea area with the image center point of the unit image as the origin, and determining a mapping quadrant having the same quadrant position as the moving direction based on the second image coordinate system; Determine each photovoltaic image corresponding to each photovoltaic target based on the sea area upper view, and determine all photovoltaic images located in the mapping quadrant as a positioning mark group; The image distances between each photovoltaic image in the positioning mark group and the image center point are respectively obtained, and the image contour of the photovoltaic image corresponding to the smallest distance is image marked with a preset pixel value.

10. An on-site real-time analysis system for edge computing data based on multi-parameter IoT technology, characterized in that: include: The unit setting module is configured to generate a detection route that passes through each photovoltaic target in sequence based on each photovoltaic target that is arranged in an array and is located in the sea area, and to pre-set corresponding detection units between each photovoltaic target that has an adjacent staggered relationship determined based on the detection route; An image acquisition module is configured to control the unmanned boat to sequentially acquire images of each photovoltaic target based on the detection route, and in response to determining based on the acquisition result that the early warning evaluation value corresponding to any photovoltaic target is greater than a preset early warning threshold, trigger the light unit pre-installed on the unmanned boat to switch from the first display mode to the second display mode; A trigger detection module is configured to control the unmanned boat to move from a current position toward a detection unit closest to the unmanned boat, and trigger the detection unit to perform video detection; a control return module, configured to control the unmanned boat to return from the preset position to the current position in response to determining that the unmanned boat corresponding to the second display mode reaches the preset position based on the video detection, and obtain the moving direction of the unmanned boat collected by the detection unit; The image marking module is configured to mark the image portion corresponding to the photovoltaic target in the sea area upper view corresponding to the sea area based on the moving direction, control the unmanned boat to continue to collect images of the remaining photovoltaic targets based on the detection route in response to the unmanned boat arriving at the current position, and trigger the light unit to switch from the second display mode to the first display mode.

Images