Power grid fault prediction method and system

Abstract

The invention provides a power grid fault prediction method and system. The method comprises the following steps: firstly, determining each power grid device meeting a fault prediction condition, and when detecting that a thermal load device value of any power grid device is lower than a thermal load reference value of the power grid device in a normal working state, determining an indirect acquisition surface having a heat source transfer relationship with the power grid device; the heat source transfer coefficient is determined through the spatial position relationship between the power grid equipment and the indirect acquisition surface, and then the value of the thermal load equipment is updated based on the heat source transfer coefficient, so that the actual thermal load state of the power grid equipment is determined more accurately. The accuracy and efficiency of thermal load collection are at least improved.

Description

Technical Field

[0001] The present invention relates to data processing technologies, and in particular, to a power grid fault prediction method and system. Background Art

[0002] In a modern industrial power environment, various power grid devices generate different degrees of thermal loads during continuous operation. To ensure power efficiency and product quality, and at the same time ensure the safe operation of the devices, it is particularly important to accurately collect and manage the thermal loads of power grid devices. Especially in large workshops or factories, there are a wide variety of power grid devices with a wide distribution. Their thermal load states not only directly affect the performance and lifespan of the devices, but also are related to the temperature control and energy consumption management of the entire power environment.

[0003] Currently, generally, the thermal load values of each power grid device are generally obtained by corresponding temperature sensors through one-to-one collection. When a problem occurs with the temperature sensor, the obtained thermal load value may deviate accordingly, and often cannot accurately reflect its actual thermal load state. Summary of the Invention

[0004] Based on the above problems, the present invention is proposed to provide a power grid fault prediction method and system that overcomes the above problems or at least partially solves the above problems.

[0005] According to one aspect of the present invention, a power grid fault prediction method is provided, including the following steps: Determine each power grid device that meets the fault prediction conditions; When the thermal load device value of any power grid device is less than the thermal load reference value in the corresponding normal working state, determine the indirect collection surface that has a heat source transfer relationship with the power grid device; Determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect collection surface, and numerically update the thermal load device value based on the heat source transfer coefficient.

[0006] Optionally, in the method according to the present invention, determining each power grid device that meets the fault prediction conditions includes: Perform image stitching on each acquisition image obtained by each fixed acquisition device from different perspectives, and compare the obtained multi-perspective image with the original image of the corresponding power workshop; In response to the existence of a missing area, control the mobile acquisition device to move to the missing area for acquisition, and update the multi-perspective image based on the obtained acquisition image; In response to the non-existence of a missing area, determine each power grid device located in the power workshop based on the multi-perspective image, and determine the power grid device with key detection attributes as the one that meets the fault prediction conditions; Perform extended processing on each acquired image including the same power grid device with edge detection attributes based on the power grid device, and determine the device status of the power grid device based on the processing results; In response to the heat dissipation state, determine that the fault prediction condition is met.

[0007] Optionally, in the method according to the present invention, performing extended processing on each acquired image including the same power grid device with edge detection attributes based on the power grid device, and determining the device status of the power grid device based on the processing results includes: Obtain each device contour corresponding to different perspectives based on each acquired image including the same power grid device with edge detection attributes, and determine each heat dissipation area corresponding to a preset magnification factor with each device contour as a reference; In response to any heat dissipation device existing in any heat dissipation area, determine the blade arrangement surface corresponding to the heat dissipation blades included in the heat dissipation device; Multiply the arrangement diameter of the blade arrangement surface by the preset heat dissipation coefficient to obtain the current heat dissipation distance; Generate a heat dissipation extension area with the current heat dissipation distance extending from the blade arrangement surface as the center to both sides based on the arrangement diameter; Determine the device status corresponding to the power grid device based on the positional relationship between the power grid device and the heat dissipation extension area.

[0008] Optionally, in the method according to the present invention, determining the device status corresponding to the power grid device based on the positional relationship between the power grid device and the heat dissipation extension area includes: In response to the heat dissipation extension area only having an overlapping relationship with this power grid device, determine that this power grid device is in a heat dissipation state; In response to the heat dissipation extension area having an overlapping relationship with this power grid device and other power grid devices simultaneously, determine the device distances between each power grid device and the heat dissipation device, and obtain each distance evaluation value based on the distance evaluation performed on each device distance; Determine the device ratios of each device located in the heat dissipation extension area based on the device contours of each power grid device, and obtain each ratio evaluation value based on the ratio evaluation performed on each device ratio; Sum up the distance evaluation value and the ratio evaluation value corresponding to the same power grid device, and determine the power grid device with the largest corresponding heat dissipation evaluation value as being in a heat dissipation state.

[0009] Optionally, in the method according to the present invention, determining the indirect acquisition surface having a heat source transfer relationship with this power grid device includes: Based on the multi-perspective images, determine the device accommodation medium for accommodating the power grid device, and determine each indirect perspective surface corresponding to different perspectives that make up the device accommodation medium, as well as each heat load medium value corresponding to each indirect perspective surface, based on each acquired image; Sort the values of each heat load medium from largest to smallest to obtain a numerical sequence; Based on the numerical sequence, determine that there is a corresponding largest heat load medium value, and determine the indirect perspective plane corresponding to this heat load medium value as the indirect acquisition plane; or, Based on the numerical sequence, determine that there are multiple corresponding largest heat load medium values, and determine the indirect perspective plane with the largest corresponding indirect area among them as the indirect acquisition plane.

[0010] Optionally, in the method according to the present invention, the method further includes: When it is determined that multiple power grid devices are accommodated by the same device accommodating medium, based on the indirect acquisition plane, obtain the device contour corresponding to each power grid device, and map the device contour to the indirect acquisition plane to obtain each mapped contour; Determine that there is an overlap between the mapped contours, and based on the numerical sequence, re-determine one of the other indirect perspective planes as the indirect acquisition plane; Perform surface shape division on the indirect perspective plane based on the mapped contour, and determine each obtained divided sub-plane as the indirect acquisition plane corresponding to each power grid device respectively.

[0011] Optionally, in the method according to the present invention, determining the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition plane, and numerically updating the heat load device value based on the heat source transfer coefficient, includes: In the trained heat source transfer determination model, based on the spatial position relationship between the power grid device and the indirect acquisition plane, given different environmental conditions, determine the heat source transfer curve of the power grid device under different environmental conditions; Obtain real-time environmental data, and based on the heat source transfer curve corresponding to the environmental condition having a data matching relationship with the real-time environmental data, determine the heat source transfer coefficient corresponding to the power grid device; Numerically modify the heat load medium value corresponding to the indirect acquisition plane based on the heat source transfer coefficient to obtain the updated heat load device value corresponding to the power grid device.

[0012] Optionally, in the method according to the present invention, the trained heat source transfer determination model is obtained through the following method: Obtain the standard position relationship, historical environmental conditions, and historical heat source transfer curve between the standard device and the standard acquisition plane; Input the standard position relationship into the benchmark determination model to be trained, and based on the historical environmental conditions, obtain the training heat source transfer curve; Based on the training heat source transfer curve and the historical heat source transfer curve, train the benchmark determination model to obtain the trained heat source transfer determination model.

[0013] Optionally, in the method according to the present invention, the real-time environmental data includes real-time humidity data and real-time temperature data.

[0014] Optionally, in the method according to the present invention, the method further includes: Performing three-dimensional modeling on the multi-view images based on the digital twin space to obtain a workshop model; Generating corresponding marking slots for each power grid device in the marking space located above the workshop model, where the marking slots include primary slots and secondary slots; Filling the primary slots of the power grid devices whose corresponding heat load device values are numerically updated; Comparing each heat load device value with the heat load reference value corresponding to the same power grid device, and filling the secondary slots of the power grid devices whose corresponding heat load device values are greater than or equal to the heat load reference value.

[0015] Optionally, in the method according to the present invention, the method further includes: Obtaining each device contour of each power grid device for secondary marking based on the workshop model, and determining each density region with a corresponding preset area centered on each device contour; Determining each density coefficient corresponding to each power grid device based on the number of devices in each density region; Determining each volume coefficient and each grade coefficient based on the device volume and importance level corresponding to each power grid device for secondary marking; Performing a summation calculation on the density coefficient, volume coefficient, and grade coefficient corresponding to the same power grid device to obtain each acquisition coefficient, and performing a product calculation on each acquisition coefficient and the preset time to obtain each acquisition time; Performing heat load acquisition on each power grid device for secondary marking based on the acquisition time; Performing a summation calculation on each heat load reference value and the preset heat load threshold to obtain each heat load warning value; In response to the heat load device value corresponding to any power grid device at any acquisition moment during the acquisition time being greater than the heat load warning value and having an upward trend, performing a power-off operation on the power grid device and sending the warning information corresponding to the power grid device to the management end.

[0016] According to another aspect of the present invention, there is provided a power grid fault prediction system, including: A device determination module configured to determine each power grid device that meets the fault prediction conditions; An acquisition determination module configured to determine an indirect acquisition surface having a heat source transfer relationship with a power grid device when the heat load device value of any power grid device is less than the heat load reference value corresponding to the normal working state; A numerical update module, configured to determine a heat source transfer coefficient based on the spatial position relationship between a power grid device and an indirect acquisition surface, and numerically update the heat load device value based on the heat source transfer coefficient.

[0017] According to the solution of the present invention, the server will first determine each power grid device located in the power workshop that meets the fault prediction conditions, so that the server only collects the heat load of the relevant power grid devices, which can effectively avoid resource waste. Furthermore, when the server detects that the heat load device value of any one power grid device is lower than the heat load reference value in its normal working state, it can quickly determine the indirect acquisition surface that has a heat source transfer relationship with the power grid device, and by considering the spatial position relationship between the power grid device and the indirect acquisition surface, scientifically and reasonably determine the heat source transfer coefficient, making the numerical update of the heat load device value more accurate and reliable. Then, the server can correct the original heat load device value based on the heat source transfer coefficient, thereby more accurately determining the actual heat load state of the power grid device. That is to say, the present invention not only improves the accuracy and efficiency of heat load collection, but also provides strong support for the power management and equipment maintenance of the power grid. Brief Description of the Drawings

[0018] Figure 1 Shows a flowchart of a power grid fault prediction method according to an embodiment of the present invention; Figure 2 Shows a schematic diagram of a mapping profile according to an embodiment of the present invention; Figure 3 Shows a schematic diagram of dividing sub - surfaces according to an embodiment of the present invention; Figure 4 Shows a block diagram of a power grid fault prediction system according to another embodiment of the present invention. Detailed Embodiments

[0019] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the 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 so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully communicated to those skilled in the art.

[0020] To solve the problems existing in the above - mentioned background technology, the inventor proposed the solution of the present invention. An embodiment of the present invention provides a power grid fault prediction method, which can be executed in a computing device.

[0021] Figure 1 Shows a flowchart of a power grid fault prediction method according to an embodiment of the present invention, which is suitable for execution in a computing device.

[0022] As Figure 1 shown, a power grid fault prediction method proposed in this embodiment starts from step S101, and in step S101, it includes the following contents: Determine each power grid device that meets the fault prediction conditions.

[0023] For example, in this embodiment, there may be multiple power grid devices in a vehicle production workshop. Among them, some power grid devices are more important or some power grid devices generate higher heat loads during operation. Therefore, more refined heat load collection needs to be carried out for each power grid device with key collection attributes to ensure the corresponding power safety.

[0024] Therefore, the server will first determine each power grid device that meets the fault prediction conditions to facilitate subsequent heat load collection for each power grid device.

[0025] Furthermore, the above "determine each power grid device that meets the fault prediction conditions" also includes the following steps: Perform image stitching on each acquisition image obtained by each fixed acquisition device from different perspectives, and compare the obtained multi-perspective image with the original image of the corresponding power workshop; In response to the existence of a missing area, control the mobile acquisition device to move to the missing area for acquisition, and update the multi-perspective image based on the obtained acquisition image; In response to the non-existence of a missing area, determine each power grid device located in the power workshop based on the multi-perspective image, and determine the power grid device with key detection attributes as meeting the fault prediction conditions; Perform extension processing based on the power grid device on each acquisition image including the same power grid device with edge detection attributes, and determine the device state of the power grid device based on the processing result; In response to the heat dissipation state, determine that the fault prediction conditions are met.

[0026] For example, in this embodiment, first, the server will use fixed acquisition devices set at different positions to obtain acquisition images from their respective perspectives. For example, the fixed acquisition devices can be set on the ceiling of the power workshop.

[0027] Then, the server will stitch these images taken from different angles to obtain a multi-perspective image. Next, the server will compare this multi-perspective image with the original image of the corresponding power workshop to check whether the multi-perspective image is complete, that is, whether there is an information missing area.

[0028] If there is a missing area, the server will control the mobile acquisition device to move, so as to collect images of the missing area, and then fill the missing area according to the newly acquired acquisition images, thus completing the update of the multi-view images.

[0029] If the comparison result shows that there is no missing area, then the server will identify all the grid devices located in the power workshop according to the multi-view images. Among these grid devices, the server will determine the grid devices with key detection attributes as meeting the fault prediction conditions.

[0030] Since the heat dissipation effect of some grid devices with edge detection attributes may be poor due to equipment aging, it is necessary to use heat dissipation devices such as fans to dissipate heat for them. In order to ensure the power safety of these aging grid devices, the server will also collect their heat loads.

[0031] First, the server will perform extension processing based on the grid device on all the acquisition images containing the same grid device with edge detection attributes, that is, judge whether the grid device is dissipating heat with the help of a heat dissipation device, so as to determine the device state of the grid device.

[0032] When the server determines that the device state of the grid device is the heat dissipation state, that is, it determines that the device meets the fault prediction conditions.

[0033] Furthermore, the above-mentioned "performing extension processing based on the grid device on each acquisition image including the same grid device with edge detection attributes, and determining the device state of the grid device based on the processing result" further includes the following steps: Obtaining the device contours corresponding to different perspectives based on each acquisition image including the same grid device with edge detection attributes, and determining the heat dissipation areas corresponding to the preset magnification factors based on each device contour; In response to the existence of any heat dissipation device in any heat dissipation area, determining the blade arrangement plane corresponding to the heat dissipation blades included in the heat dissipation device; Calculating the product of the arrangement diameter of the blade arrangement plane and the preset heat dissipation coefficient to obtain the current heat dissipation distance; Generating a heat dissipation extension area with the current heat dissipation distance extending from the blade arrangement plane as the center to both sides based on the arrangement diameter; Determining the device state corresponding to the grid device based on the positional relationship between the grid device and the heat dissipation extension area.

[0034] For example, in this embodiment, the server will obtain the device contours corresponding to different perspectives through each acquisition image including the same grid device with edge detection attributes.

[0035] Under normal circumstances, in order to ensure a certain heat dissipation effect, the heat dissipation device is placed in an area at a certain distance from the power grid equipment. Therefore, the server will, based on each device contour, enlarge the device contour according to a preset magnification factor to obtain the corresponding heat dissipation areas.

[0036] Then, the server determines whether there is a heat dissipation device in the heat dissipation area. When there is a heat dissipation device in any heat dissipation area, in order to facilitate the determination of the heat dissipation direction, the server first determines the blade arrangement plane formed by the heat dissipation blades located in the heat dissipation device.

[0037] Since the larger the blade arrangement plane, the larger the corresponding heat dissipation range, the server multiplies the arrangement diameter of the blade arrangement plane by the preset heat dissipation coefficient to calculate the current heat dissipation distance. Thus, a heat dissipation extension area with the current heat dissipation distance is constructed based on the arrangement diameter and extending from the blade arrangement plane to both sides.

[0038] Finally, the server determines the device state of the power grid equipment by analyzing the relative position relationship between the power grid equipment and the heat dissipation extension area.

[0039] Furthermore, the above "determining the device state corresponding to the power grid equipment based on the position relationship between the power grid equipment and the heat dissipation extension area" further includes the following steps: In response to the heat dissipation extension area only overlapping with this power grid equipment, determining that this power grid equipment is in a heat dissipation state; In response to the heat dissipation extension area overlapping with this power grid equipment and other power grid equipment simultaneously, determining the device distances between each power grid equipment and the heat dissipation device, and obtaining each distance evaluation value based on the distance evaluation performed on each device distance; Based on the device contours of each power grid equipment, determining the proportion of each device located in the heat dissipation extension area, and obtaining each proportion evaluation value based on the proportion evaluation performed on each device proportion; Performing a summation calculation on the distance evaluation value and the proportion evaluation value corresponding to the same power grid equipment, and determining the power grid equipment with the largest corresponding heat dissipation evaluation value as being in a heat dissipation state.

[0040] For example, in this embodiment, when the heat dissipation extension area only overlaps with this power grid equipment, it means that the heat dissipation device is performing a heat dissipation operation on this power grid equipment. Therefore, the server determines that this power grid equipment is in a heat dissipation state.

[0041] When the heat dissipation extension area overlaps with multiple power grid equipment simultaneously, the server further determines which power grid equipment the heat dissipation device is dissipating heat to.

[0042] First, the server determines the device distance between each power grid device and the heat dissipation device, and then conducts a distance assessment based on the device distance to obtain the corresponding distance assessment value. Next, the server determines the device proportion within the heat dissipation extension area according to the device profile of each power grid device, and conducts a proportion assessment based on the device proportion to obtain each proportion assessment value.

[0043] Finally, the server adds the distance assessment value and the proportion assessment value corresponding to the same power grid device, calculates the total heat dissipation assessment value, and determines the power grid device with the maximum heat dissipation assessment value as the power grid device in the heat dissipation state.

[0044] This embodiment can obtain the heat dissipation assessment value from two aspects of the distance assessment value and the proportion assessment value, thereby determining the power grid device in the heat dissipation state, which has a certain degree of accuracy.

[0045] In step S102, the following content is included: When the heat load device value of any power grid device is less than the heat load reference value in the corresponding normal working state, determine the indirect acquisition surface that has a heat source transfer relationship with the power grid device.

[0046] For example, in this embodiment, usually, the heat load device value of each power grid device is collected by a temperature sensor set beside the power grid device. However, the temperature sensor may malfunction during the collection process due to reasons such as sensor aging, resulting in the output heat load device value being less than the actual heat load device value.

[0047] Therefore, when the heat load device value of any power grid device is less than the heat load reference value in the corresponding normal working state, the server first determines whether the power grid device is located within the device accommodation medium, such as within the iron sheet accommodation medium, through multi - perspective images.

[0048] When the power grid device is not located within the device accommodation medium, the server obtains the heat load device value corresponding to the device through multi - perspective images to replace the heat load device value output by the temperature sensor; when the power grid device is located within the device accommodation medium, in order to more accurately determine the heat load device value of the power grid device, the server determines the indirect acquisition surface that has a heat source transfer relationship with the power grid device on the device accommodation, and thus conducts heat load collection on the power grid device based on the indirect acquisition surface.

[0049] Furthermore, the above "when the heat load device value of any power grid device is less than the heat load reference value in the corresponding normal working state, determine the indirect acquisition surface that has a heat source transfer relationship with the power grid device" further includes the following steps: Determine a device accommodation medium for accommodating power grid devices based on multi-view images, and determine each indirect view surface corresponding to different views that make up the device accommodation medium and each heat load medium value corresponding to each indirect view surface based on each acquired image; Perform a numerical sorting of the heat load medium values from largest to smallest to obtain a numerical sequence; Based on the numerical sequence, determine that there is a corresponding maximum heat load medium value, and determine the indirect view surface corresponding to this heat load medium value as the indirect acquisition surface; or, Based on the numerical sequence, determine that there are multiple corresponding maximum heat load medium values, and determine the indirect view surface with the largest corresponding indirect area among them as the indirect acquisition surface.

[0050] For example, in this embodiment, the server will determine the medium for accommodating power grid devices (i.e., the device accommodation medium) through multi-view images. Then, the server will use the multi-view images to identify each indirect view surface that makes up the device accommodation medium, and calculate the corresponding heat load medium value for each indirect view surface.

[0051] Then, the server will sort these heat load medium values from largest to smallest to obtain a numerical sequence.

[0052] If there is a corresponding maximum heat load medium value in the numerical sequence, the server will determine the indirect view surface corresponding to this maximum value as the indirect acquisition surface; if there are multiple indirect view surfaces corresponding to the maximum heat load medium value in the numerical sequence, the server will select the indirect view surface with the largest corresponding indirect area among these indirect view surfaces as the indirect acquisition surface. When there are multiple indirect view surfaces with the largest corresponding indirect area, the server will randomly select one from these indirect view surfaces with the largest indirect area as the indirect acquisition surface.

[0053] Furthermore, the above method further includes the following steps: When it is determined that multiple power grid devices are accommodated by the same device accommodation medium, obtain the device contour corresponding to each power grid device based on the indirect acquisition surface, and map the device contour to the indirect acquisition surface to obtain each mapped contour; Determine that there is an overlap between the mapped contours, and re-determine one of the other indirect view surfaces as the indirect acquisition surface based on the numerical sequence; Perform a surface shape division on the indirect view surface based on the mapped contour, and determine each obtained divided sub-surface as the indirect acquisition surface corresponding to each power grid device respectively.

[0054] For example, in this embodiment, when multiple power grid devices are accommodated by the same device accommodating medium, the server will first obtain the device contour of each power grid device according to the previously determined indirect acquisition surface, and then map each device contour onto the indirect acquisition surface to form respective mapped contours.

[0055] If there is an overlap between these mapped contours (as shown in Figure 2 ), it indicates that this indirect acquisition surface is not conducive to the thermal load acquisition of all power grid devices. Therefore, the server will select from other indirect perspective surfaces as the indirect acquisition surface according to the numerical sequence, that is, the finally determined indirect acquisition surface can ensure that there is no overlap between each mapped contour.

[0056] Finally, the server will divide the indirect perspective surface using the mapped contours to form multiple divided sub - surfaces (as shown in Figure 3 ), and then determine each divided sub - surface as the indirect acquisition surface corresponding to each power grid device respectively.

[0057] In step S103, the following contents are included: Determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition surface, and numerically update the thermal load device value based on the heat source transfer coefficient.

[0058] For example, in this embodiment, different spatial position relationships between the power grid device and the indirect acquisition surface will lead to different heat source transfer effects. For example, the closer the power grid device is to the indirect acquisition surface, the better the heat source transfer effect.

[0059] Therefore, the server will determine the heat source transfer coefficient according to the spatial position relationship between the power grid device and the indirect acquisition surface, and numerically update the thermal load device value based on the heat source transfer coefficient.

[0060] Furthermore, the above "determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition surface, and numerically update the thermal load device value based on the heat source transfer coefficient" further includes the following steps: In the trained heat source transfer determination model, based on the spatial position relationship between the power grid device and the indirect acquisition surface, give different environmental conditions, and determine the heat source transfer curve of the power grid device under different environmental conditions; Obtain real - time environmental data, and determine the heat source transfer coefficient corresponding to the power grid device based on the heat source transfer curve corresponding to the environmental condition having a data matching relationship with the real - time environmental data; Numerically modify the thermal load medium value corresponding to the indirect acquisition surface based on the heat source transfer coefficient to obtain the updated thermal load device value corresponding to the power grid device.

[0061] For example, in this embodiment, since different environmental conditions also affect the heat source transfer effect, the server will determine the heat source transfer characteristic curves of the power grid equipment under various environmental conditions by giving different environmental conditions in the trained heat source transfer prediction model according to the spatial position relationship between the power grid equipment and the indirect acquisition surface.

[0062] Next, the server will obtain real-time environmental data, which includes real-time humidity data and real-time temperature data. Then, it will select the heat source transfer curve corresponding to the environmental condition that matches the real-time environmental data from the heat source transfer characteristic curves to determine the heat source transfer coefficient for the current power grid equipment.

[0063] Finally, the server will use this heat source transfer coefficient to adjust the heat load medium value corresponding to the indirect acquisition surface, thereby obtaining the updated heat load equipment value corresponding to the power grid equipment.

[0064] Furthermore, the trained heat source transfer determination model is obtained through the following method: Obtain the standard position relationship, historical environmental conditions, and historical heat source transfer curves between the standard equipment and the standard acquisition surface; Input the standard position relationship into the benchmark determination model to be trained, and based on the historical environmental conditions, obtain the heat source transfer curve for training; Based on the heat source transfer curve for training and the historical heat source transfer curves, train the benchmark determination model to obtain the trained heat source transfer determination model.

[0065] For example, in this embodiment, the server will obtain the standard position relationship, historical environmental conditions, and historical heat source transfer curves between the standard equipment and the standard acquisition surface used for model training. Then, it will input the standard position relationship into a benchmark determination model to be trained and generate the heat source transfer curve for training in combination with the historical environmental conditions.

[0066] Next, the server will train and optimize the benchmark determination model according to the heat source transfer curve for training and the historical heat source transfer curves to obtain a trained heat source transfer determination model.

[0067] Furthermore, the above method further includes the following steps: Perform three-dimensional modeling on the multi-view images based on the digital twin space to obtain the workshop model; Generate marker slots corresponding to each power grid equipment in the marker space located above the workshop model, where the marker slots include first-level slots and second-level slots; Fill the first-level slots of the power grid equipment whose corresponding heat load equipment value is numerically updated; Compare the heat load device values of each device with the heat load reference values of the corresponding grid devices on the same power grid, and fill the secondary slots of the grid devices for which the corresponding heat load device values are greater than or equal to the heat load reference values.

[0068] For example, in this embodiment, the server first uses digital twin technology to construct a three-dimensional model of the power workshop based on multi-view images.

[0069] Next, the server sets a marking space above the workshop model based on the top view perspective, and generates marking slots corresponding to each grid device in the marking space. Among them, the marking slots include primary slots and secondary slots.

[0070] Then, the server fills the primary slots of the grid devices whose corresponding heat load device values are numerically updated. For example, the primary slot corresponding to the grid device is filled with yellow.

[0071] Then, the server compares each heat load device value with the heat load reference value of the corresponding grid device on the same power grid. If the heat load device value of a certain grid device is greater than or equal to its heat load reference value, it means that the heat load generated by the grid device is relatively high. Therefore, the server fills the secondary slot corresponding to the grid device in the workshop model. For example, the secondary slot corresponding to the grid device is filled with red to enhance the identification.

[0072] Furthermore, the above method further includes the following steps: Obtain the device outlines of each grid device for secondary marking based on the workshop model, and determine density regions with a corresponding preset area centered on each device outline; Determine the density coefficients corresponding to each grid device based on the number of devices located in each density region; Determine the volume coefficients and level coefficients based on the device volume and importance level corresponding to each grid device for secondary marking; Sum up the density coefficients, volume coefficients, and level coefficients corresponding to the same grid device to obtain each collection coefficient, and multiply each collection coefficient by the preset time to obtain each collection time; Perform heat load collection based on the collection time for each grid device for secondary marking; Sum up each heat load reference value and the preset heat load threshold to obtain each heat load warning value; In response to the heat load device value corresponding to any grid device at any collection moment during the collection time being greater than the heat load warning value and having an upward trend, perform a power-off operation on the grid device and send the warning information corresponding to the grid device to the management terminal.

[0073] For example, in this embodiment, since the heat load device values generated by each power grid device with secondary marking are greater than or equal to their heat load reference values, in order to ensure the corresponding power security, the server will collect the heat loads of these power grid devices with secondary marking for corresponding time periods.

[0074] Since the higher the density of power grid devices and the larger the device volume, the higher the degree of harm caused by excessive heat generated. Therefore, the server will first obtain the device contours of each power grid device with secondary marking according to the workshop model, and then determine the density regions with corresponding preset areas centered on each device contour. Next, according to the number of devices in each density region, the density coefficients corresponding to each power grid device are determined.

[0075] Then, the server will determine the volume coefficients corresponding to each power grid device according to the device volume corresponding to each power grid device with secondary marking.

[0076] Since different power grid devices have different importance levels, the higher the level of a power grid device, the longer the heat load needs to be collected to monitor the operation of the power grid device in a timely manner. Therefore, the server will determine the level coefficients according to the importance levels corresponding to each power grid device with secondary marking.

[0077] After that, the server will sum up the density coefficient, volume coefficient, and level coefficient corresponding to the same power grid device to obtain each collection coefficient. Then, multiply each collection coefficient by the preset time to obtain the collection time corresponding to each power grid device, and thus collect the heat load of each power grid device with secondary marking according to the collection time.

[0078] The server will add the heat load reference value of each power grid device to the preset heat load threshold to calculate their respective heat load warning values. When the heat load device value of a certain power grid device collected at any collection moment within the collection time not only exceeds its corresponding heat load warning value but also shows an upward trend, the system will automatically cut off the power supply of this power grid device to prevent potential safety risks, and immediately send the warning information corresponding to this power grid device to the management terminal so that the management personnel can respond quickly and take corresponding measures.

[0079] According to the solution of the present invention, the server will first determine each power grid device located in the power workshop that meets the fault prediction conditions, so that the server only collects the thermal load of relevant power grid devices, which can effectively avoid waste of resources. Furthermore, when the server detects that the thermal load device value of any power grid device is lower than the thermal load reference value in its normal working state, it can quickly determine the indirect collection surface that has a heat source transfer relationship with the power grid device, and by considering the spatial position relationship between the power grid device and the indirect collection surface, scientifically and reasonably determine the heat source transfer coefficient, making the numerical update of the thermal load device value more accurate and reliable. Then, the server can correct the original thermal load device value based on the heat source transfer coefficient, so as to more accurately determine the actual thermal load state of the power grid device. That is to say, the present invention not only improves the accuracy and efficiency of thermal load collection, but also provides strong support for the power management and equipment maintenance of the power grid.

[0080] Another embodiment of the present invention provides a power grid fault prediction system. Figure 4 For its corresponding system block diagram, the system includes: An equipment determination module, configured to determine each power grid device that meets the fault prediction conditions; A collection determination module, configured to determine an indirect collection surface that has a heat source transfer relationship with a power grid device when the thermal load device value of any power grid device is less than the thermal load reference value corresponding to its normal working state; A numerical update module, configured to determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect collection surface, and numerically update the thermal load device value based on the heat source transfer coefficient.

[0081] In the specification provided herein, the algorithms and displays are not inherently related to any specific computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the examples of the present invention. Based on the above description, the structure required to construct such a system is obvious. In addition, the present invention is not directed to any specific programming language. It should be understood that the content of the present invention described herein can be implemented using various programming languages, and the description of a specific language above is to disclose the preferred embodiments of the present invention.

[0082] In the specification provided herein, a large number of specific details are illustrated. However, it can be understood that the embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures, and technologies have not been shown in detail so as not to obscure the understanding of this specification.

[0083] Similarly, it should be understood that, in order to streamline the present disclosure and assist in understanding one or more of the various inventive aspects, in the foregoing description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof.

[0084] Those skilled in the art should understand that the modules or units or components of the devices in the examples disclosed herein may be arranged in the devices as described in the embodiments, 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 furthermore be divided into multiple sub-modules.

[0085] Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and furthermore can be divided into multiple sub-modules or sub-units or sub-components.

[0086] In addition, those skilled in the art can understand that although some of the embodiments described herein include certain features included in other embodiments rather than other features, the combination of the features of different embodiments means that it is within the scope of the present invention and forms different embodiments.

[0087] In addition, some of the embodiments described herein are described as methods or combinations of method elements that can be implemented by a processor of a computer system or by other devices performing the functions. Therefore, a processor having the 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 in the device embodiments are examples of the following devices: the device for implementing the functions performed by the elements for the purpose of implementing the present invention.

[0088] As used herein, unless otherwise specified, the use of ordinal numbers "first", "second", "third", etc. to describe ordinary objects only indicates different instances of similar objects and does not intend to imply that the objects so described must have a given order in terms of time, space, sorting, or in any other way.

[0089] Although the present invention has been described in terms of a limited number of embodiments, those skilled in the art in this technical field will appreciate that other embodiments can be contemplated within the scope of the present invention as thus described. In addition, it should be noted that the language used in this specification has been primarily selected for readability and teaching purposes rather than for the purpose of explaining or limiting the subject matter of the present invention.

Claims

1. A power grid fault prediction method, characterized in that, Including the following steps: Determine each power grid device that meets the fault prediction conditions; When the thermal load device value of any power grid device is less than the thermal load reference value in the corresponding normal working state, determine the indirect acquisition surface that has a heat source transfer relationship with the power grid device; Determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition surface, and numerically update the thermal load device value based on the heat source transfer coefficient.

2. The power grid fault prediction method according to claim 1, wherein: Determining each power grid device that meets the fault prediction conditions includes: Perform image stitching on each acquisition image obtained by each fixed acquisition device from different perspectives, and compare the obtained multi-perspective image with the original image of the corresponding power workshop; In response to the existence of a missing area, control the mobile acquisition device to move to the missing area for acquisition, and update the multi-perspective image based on the obtained acquisition image; In response to the non-existence of a missing area, determine each power grid device located in the power workshop based on the multi-perspective image, and determine the power grid device with key detection attributes as meeting the fault prediction conditions; Perform extension processing based on the power grid device on each acquisition image including the same power grid device with edge detection attributes, and determine the device state of the power grid device based on the processing result; In response to the heat dissipation state, determine that the fault prediction conditions are met.

3. The power grid fault prediction method according to claim 2, wherein: Performing extension processing based on the power grid device on each acquisition image including the same power grid device with edge detection attributes, and determining the device state of the power grid device based on the processing result includes: Obtain each device contour corresponding to different perspectives based on each acquisition image including the same power grid device with edge detection attributes, and determine each heat dissipation area corresponding to a preset magnification factor with each device contour as a reference; In response to the existence of any heat dissipation device in any heat dissipation area, determine the blade arrangement surface corresponding to the heat dissipation blades included in the heat dissipation device; Calculate the product of the arrangement diameter of the blade arrangement surface and the preset heat dissipation coefficient to obtain the current heat dissipation distance; Generate a heat dissipation extension area with the current heat dissipation distance extending from the center of the blade arrangement surface to both sides based on the arrangement diameter; Determine the device state corresponding to the power grid device based on the position relationship between the power grid device and the heat dissipation extension area.

4. The power grid fault prediction method according to claim 3, wherein: Determining the device state corresponding to the power grid device based on the position relationship between the power grid device and the heat dissipation extension area includes: In response to the heat dissipation extension area only having an overlapping relationship with this power grid device, determine that this power grid device is in a heat dissipation state; In response to the heat dissipation extension area having an overlapping relationship with this power grid device and other power grid devices at the same time, determine the device distances between each power grid device and the heat dissipation device, and obtain each distance evaluation value based on the distance evaluation performed on each device distance; Determine the device ratios of each device located in the heat dissipation extension area based on the device contours of each power grid device, and obtain each ratio evaluation value based on the ratio evaluation performed on each device ratio; Sum the distance evaluation value and the proportion evaluation value corresponding to the same power grid device, and determine the power grid device with the largest obtained heat dissipation evaluation value as the heat dissipation state.

5. The power grid fault prediction method according to claim 2, characterized in that Determine the indirect acquisition surface having a heat source transfer relationship with the power grid device, including: Based on multi-view images, determine the device accommodation medium that accommodates the power grid device, and based on each acquisition image, determine each indirect view surface corresponding to different perspectives that make up the device accommodation medium, and each heat load medium value corresponding to each indirect view surface; Sort the values of each heat load medium from largest to smallest to obtain a numerical sequence; Based on the numerical sequence, determine that there is a corresponding maximum heat load medium value, and determine the indirect view surface corresponding to the heat load medium value as the indirect acquisition surface; or, Based on the numerical sequence, determine that there are multiple corresponding maximum heat load medium values, and determine the indirect view surface with the largest corresponding indirect area among them as the indirect acquisition surface.

6. The power grid fault prediction method according to claim 5, characterized in that The method further includes: When it is determined that multiple power grid devices are accommodated by the same device accommodation medium, obtain the device contour corresponding to each power grid device based on the indirect acquisition surface, and map the device contour to the indirect acquisition surface to obtain each mapped contour; Determine that there is an overlap between the mapped contours, and re-determine one of the other indirect view surfaces as the indirect acquisition surface based on the numerical sequence; Divide the surface shape of the indirect view surface based on the mapped contour, and determine each divided sub-surface obtained as each indirect acquisition surface corresponding to each power grid device respectively.

7. The power grid fault prediction method according to claim 5, characterized in that Determine the heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition surface, and numerically update the heat load device value based on the heat source transfer coefficient, including: In the trained heat source transfer determination model, based on the spatial position relationship between the power grid device and the indirect acquisition surface, give different environmental conditions, and determine the heat source transfer curve of the power grid device under different environmental conditions; Obtain real-time environmental data, and determine the heat source transfer coefficient corresponding to the power grid device based on the heat source transfer curve corresponding to the environmental condition having a data matching relationship with the real-time environmental data; Numerically modify the heat load medium value corresponding to the indirect acquisition surface based on the heat source transfer coefficient to obtain the updated heat load device value corresponding to the power grid device.

8. The power grid fault prediction method according to claim 7, characterized in that The method further includes: Perform three-dimensional modeling on the multi-view images based on the digital twin space to obtain a workshop model; Generate marker slots corresponding to each power grid device in the marker space located above the workshop model, where the marker slots include primary slots and secondary slots; Fill the primary slots of the power grid device whose heat load device value is numerically updated; Compare each heat load device value with the heat load reference value corresponding to the same power grid device, and fill the secondary slots of the power grid device whose heat load device value is greater than or equal to the heat load reference value.

9. The power grid fault prediction method according to claim 8, wherein: The method further includes: Obtaining the device profiles of each power grid device for secondary marking based on the workshop model, and determining density regions with corresponding preset areas centered on each device profile; Determining the density coefficients corresponding to each power grid device based on the number of devices located in each density region; Determining the volume coefficients and grade coefficients based on the device volume and importance level corresponding to each power grid device for secondary marking; Performing a summation calculation on the density coefficient, volume coefficient, and grade coefficient corresponding to the same power grid device to obtain each acquisition coefficient, and performing a product calculation on each acquisition coefficient and the preset time to obtain each acquisition time; Performing heat load acquisition on each power grid device for secondary marking based on the acquisition time; Performing a summation calculation on each heat load reference value and the preset heat load threshold to obtain each heat load warning value; In response to the heat load device value corresponding to any power grid device at any acquisition moment during the acquisition time being greater than the heat load warning value and having an upward trend, performing a power-off operation on the power grid device and sending the warning information corresponding to the power grid device to the management terminal.

10. A power grid fault prediction system, characterized in that, Including: A device determination module configured to determine each power grid device that meets the fault prediction conditions; An acquisition determination module configured to determine an indirect acquisition surface having a heat source transfer relationship with the power grid device when the heat load device value of any power grid device is less than the heat load reference value corresponding to the normal working state; A numerical update module configured to determine a heat source transfer coefficient based on the spatial position relationship between the power grid device and the indirect acquisition surface, and perform numerical update on the heat load device value based on the heat source transfer coefficient.

Images