An intelligent monitoring system and method for distribution network equipment based on artificial intelligence

Abstract

The present invention discloses an intelligent monitoring system and method for distribution network equipment based on artificial intelligence, belonging to the technical field of equipment monitoring. The system includes a data acquisition module, an intelligent planning module, a monitoring and analysis module and an early warning prompt module; the data acquisition module is used to collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information and the status of each component under the distribution network equipment; the intelligent planning module is used to build a scene model and use an artificial intelligence algorithm to perform target recognition on the image; the recognition component is associated with the component in the scene model, the current tool is identified and the safety distance is calculated through the data in the historical record; the monitoring and analysis module is used to establish different execution order sets, mark the spatial position of the camera in the scene model, and mark the component code that disables the current tool, and calculate the weight coefficient and safety index of each component corresponding to the marking code; the early warning prompt module automatically warns according to the change of the safety index.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment monitoring, and in particular to an intelligent monitoring system and method for distribution network equipment based on artificial intelligence. Background Art

[0002] The power distribution network is the key link in delivering electric energy from the transmission network to end users. Its safety and reliability are crucial to the normal operation of the power system. The operating status of the distribution network equipment directly affects the stability and reliability of power supply, so maintenance monitoring and operation management are required.

[0003] At present, the maintenance and monitoring of distribution network equipment usually uses various sensors to monitor and record operating behaviors. This method has certain disadvantages. On the one hand, the monitoring process often uses a fixed-angle image analysis method, and the occlusion of moving objects during the operation and the three-dimensional distribution of the operating positions of the distribution network equipment will lead to monitoring blind spots. In addition, due to the fixed monitoring distance, the operating tools cannot be identified in a timely and accurate manner, and the degree of standardization of the use of operating tools cannot be monitored according to the real-time status of the components and the specifications, which leads to safety hazards. On the other hand, the existing technology often has a certain lag in the process of operating specifications, and can only provide prompts after non-standard operating behaviors occur. In the field of distribution equipment operations, the lagging prompts often cannot play a good preventive role. Therefore, at this stage, a more efficient and intelligent distribution network equipment operation specification monitoring technology solution is needed to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent monitoring system and method for distribution network equipment based on artificial intelligence to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent monitoring method for distribution network equipment based on artificial intelligence, the method comprising the following steps:

[0006] S100, collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information during the operation and the status of each component under the distribution network equipment.

[0007] S200, build a scene model through a structural three-dimensional map, use an artificial intelligence algorithm to identify the target in the image, associate the identified components with the components in the scene model, identify the current tool and calculate the safety distance through the data in the historical record.

[0008] S300, analyzing the historical logs and establishing different execution order sets, marking the spatial position of the camera in the scene model, and marking the component code that disables the current tool. By analyzing the execution order set, the weight coefficient of the component corresponding to each marked code is calculated, and the safety index of the component corresponding to each marked code is calculated according to the spatial position of the camera in the scene model.

[0009] S400: When the safety index is less than the index threshold, the personnel are reminded to change the current tool or keep a certain distance.

[0010] In S100, the historical log refers to the operation records of all personnel, and each operation record includes the personnel identifier, component code, tool code, operation score and operation time. The three-dimensional structural diagram refers to the three-dimensional diagram of the internal structure of the power distribution equipment, specifically including the spatial position distribution of all components in the power distribution equipment. The tool comparison table refers to the tool list of each component in different states, specifically including the component code, status, available tool list and disabled tool list. The status includes power on and power off, and the list includes the codes of different tools. Image information refers to real-time video, which is collected by the camera on the wearable device of the personnel. The status of the component includes power on and power off, and is fed back in real time by the distribution network equipment.

[0011] The operation score indicates the degree of standardization of personnel's use of tools. The camera collects real-time video of personnel using tools in the process of operation, and uses a deep learning model to detect the key points of hands and tools in the video. The action features are extracted and compared with standard actions to calculate the action standardization score. The score ranges from 1 to 10.

[0012] The operation record is a record of each time a person uses a tool to operate a component. If a person changes the tool or adjusts the component during the operation, a new operation record will be generated. The operation score is limited to the degree of standardization of the person's actions within a certain period of time without changing the tool or adjusting the component.

[0013] The tool list is set in advance by the staff, with specific reference to the available tools and prohibited tools of the components in different states. Available tools refer to tools that can be used safely in this state and will not cause equipment damage or casualties; prohibited tools refer to tools that cannot be used safely in this state and may cause equipment damage or casualties.

[0014] The wearable device refers to a smart helmet, which has real-time shooting and micro-projection functions. It can shoot video within the field of vision as the person moves his body or turns his head. The micro-projection can project the set pattern to the designated location.

[0015] In S200, the specific steps are as follows:

[0016] S201: When a person performs operations on a distribution network device, the wearable device automatically interacts with the distribution network device to obtain the person's identifier stored in the wearable device. As well as the three-dimensional structure diagram saved in the distribution network equipment, the scene model is built according to the three-dimensional structure diagram using three-dimensional modeling software. The scene model includes a three-dimensional diagram of the distribution equipment, which can obtain the spatial position coordinates of each component.

[0017] S202: Establish a component set, use artificial intelligence algorithms to detect targets in real-time video, and analyze and determine whether each identified target is a tool or a component. If the target is a tool, obtain the code If the target is a component, the code is obtained and put into the component collection. After all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model.

[0018] The code of the component in the component collection changes with the video screen. When other component targets are added or the original component targets are lost in the video screen, the code of the corresponding component is added or deleted in the component collection. According to the component corresponding to the code in the component collection, the corresponding component is found in the scene model for association.

[0019] S203, searching the historical log for the identifier and coded as The operation scores of these operation records are summed up and the average value is calculated. . Retrieve the code in the history log operation records, sum up the operation scores in these operation records and calculate the average value .Will and Substitute the formula to calculate the safety distance, the formula is as follows:

[0020]

[0021] In the formula, For a safe distance, is the standard distance, is a constant, Score maximum action.

[0022] The safety distance is related to the average operation score of personnel using the tool. The lower the average operation score, the lower the proficiency of the personnel in using the tool. They are unable to master the tool's operating skills or safety precautions. Therefore, a larger safety distance needs to be set to ensure the safety of equipment or personnel.

[0023] In S300, the specific steps are as follows:

[0024] S301. All operation records in the history log are classified according to the identifiers. Operation records with the same identifier are first classified into the same category. Then, operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group. Each group contains at least two operation records. Count the number of all groups. and establish An execution order set is created, and the codes in the operation records of the same group are placed into the execution order set in the order of operation time.

[0025] Each execution sequence set represents the work component trajectory of the same person within a period of time, and the work component trajectory is related to the circuit design logic of the distribution equipment or the standard work requirement sequence.

[0026] S302, obtain the current status of each component of the distribution network device in real time feedback, retrieve the list of disabled tools corresponding to each code in the component set in the current state through the comparison table, and the disabled tool list contains Analyze the components and distances between components in the video in real time, compare them with the three-dimensional structure diagram, use visual SLAM algorithm to analyze the position and posture of the camera, so as to determine the spatial position of the camera relative to the distribution network equipment, and mark the spatial position coordinates of the camera in the scene model in real time. .

[0027] S303: Obtaining distance space position coordinates in the scene model The code of the component with the closest 3D distance . Filter out all the codes that contain And then contains at least one execution order set of marked code, these execution order sets Delete all previous codes. Determine whether each marked code exists in at least one of the execution order sets. If the result is no, set the weight coefficient of the corresponding component of the marked code to zero. If the result is yes, analyze the marked code in each execution order set. The number of codes in the interval is substituted into the formula to calculate the weight coefficient of the component corresponding to the marking code.

[0028]

[0029] In the formula, is the weight coefficient, is the number of all execution order sets containing the marked code, For the The execution order set contains the marked code and The number of codes in the interval, is a constant.

[0030] Code The corresponding part indicates the part closest to the worker's space, and it is used as the part that the worker is currently working on. The weight coefficient indicates the possibility of disabling the part of the current tool as the next part that the worker is working on. The higher the weight coefficient, the greater the possibility.

[0031] S304, obtain the spatial position coordinates of the component corresponding to each marking code in the scene model, and calculate the spatial position coordinates respectively The three-dimensional distance between , together with the weight coefficient of the component corresponding to the marking code, are substituted into the formula to calculate the safety index of each component corresponding to the marking code in real time. The formula is as follows:

[0032]

[0033] In the formula, is the safety index, is a constant, is the number of marking codes, For the The weight coefficient of the component corresponding to each marking code.

[0034] In S400, the changes in the spatial position coordinates of the camera in the scene model are monitored in real time, and the safety index of the component corresponding to each marking code is calculated. When the safety index is less than the index threshold, the corresponding component is used as a warning component, and the micro-projector on the wearable device is used to project and display at the position of the warning component, prompting personnel to keep a certain distance from the warning component or replace the currently used tool.

[0035] The larger the weight coefficient, the smaller the base of the safety index. As the three-dimensional distance between the personnel and the components corresponding to the marking codes gradually decreases, early warning will be issued when the distance is less than the index threshold, thus avoiding simultaneous warnings for multiple components corresponding to marking codes in a short period of time.

[0036] An artificial intelligence-based intelligent monitoring system for distribution network equipment, the system comprises a data acquisition module, an intelligent planning module, a monitoring and analysis module and an early warning prompt module.

[0037] The data acquisition module is used to collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information and the status of each component under the distribution network equipment. The intelligent planning module is used to build a scene model, use artificial intelligence algorithms to identify targets in images, associate identified components with components in the scene model, identify the current tool, and calculate the safety distance through data in the historical records. The monitoring and analysis module is used to establish different execution order sets, mark the spatial position of the camera in the scene model, mark the component code that disables the current tool, and calculate the weight coefficient and safety index of each component corresponding to the marked code. The early warning prompt module automatically warns according to changes in the safety index.

[0038] The data collection module includes an environmental information collection unit, a chart information collection unit and a historical log collection unit.

[0039] The environmental information acquisition unit is used to collect image information during the operation and the status of each component of the distribution network equipment. Image information refers to real-time video, and the status of components includes power on and power off.

[0040] The chart information collection unit is used to collect the three-dimensional structure diagram and the tool comparison table. The three-dimensional structure diagram refers to the three-dimensional diagram of the internal structure of the power distribution equipment, including the spatial position distribution of all components in the power distribution equipment. The tool comparison table refers to the tool list of each component in different states, including the component code, state, available tool list and disabled tool list, and the list includes the codes of different tools.

[0041] The history log collection unit is used to collect history logs. History logs refer to the operation records of all personnel, and each operation record includes the personnel identifier, component code, tool code, operation score and operation time.

[0042] The intelligent planning module includes a scene building unit and a distance calculation unit.

[0043] The scene building unit is used to build the scene model. When a person performs operations on the distribution network equipment, the person's identifier is obtained. As well as the structural 3D graph, the scene model is built according to the structural 3D graph. A component set is established, and an artificial intelligence algorithm is used to detect targets in real-time video. If the target is a tool, the encoding is obtained. If the target is a component, the code is obtained and put into the component collection. After all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model.

[0044] The distance calculation unit is used to calculate the safety distance; all identifiers are and coded as The operation scores in the operation records are summed up and the average value is calculated to get , and then encode all The operation scores in the operation records are summed up and the average value is calculated to get ; Substitute into the formula: Calculate the safe distance .in, is the standard distance, is a constant, Score maximum action.

[0045] The monitoring and analysis module includes a component analysis unit and a safety monitoring unit.

[0046] The component analysis unit is used to mark the component code that disables the current tool.

[0047] First, the operation records with the same identifier in the historical log are classified into the same category, and then the operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group. Second, the number of all groups is counted. and establish The codes in the same group of operation records are put into the execution order set in the order of operation time. Finally, the current status of each component is obtained, and the disabled tool list of each code corresponding to the component in the component set in the current state is retrieved from the comparison table. The disabled tool list contains Then mark the corresponding code.

[0048] The safety monitoring unit is used to calculate the safety index of each component corresponding to the marking code.

[0049] First, compare the components appearing in the video with the three-dimensional structure diagram to determine the spatial position of the camera relative to the distribution network equipment and mark the spatial position coordinates in the scene model. , get the distance space position coordinates in the scene model The code of the component with the closest 3D distance .

[0050] Second, filter out all the code containing And then contains at least one execution order set of marked code, these execution order sets If the marked code does not exist in any of the execution order sets, the weight coefficient of the corresponding component of the marked code is set to zero. In other cases, the marked code is analyzed in each execution order set. The number of codes in the interval, substitute into the formula: Calculate the weight coefficient of the component corresponding to the marking code .in, is the number of all execution order sets containing the marked code, For the The execution order set contains the marked code and The number of codes in the interval, is a constant.

[0051] Finally, the spatial position coordinates and The three-dimensional distance between , substitute into the formula: , calculate the safety index of each component corresponding to the marking code. is a constant, is the number of marking codes, For the The weight coefficient of the component corresponding to each marking code.

[0052] The early warning prompt module is used for automatic early warning. The safety index of each component corresponding to the marking code is calculated in real time, and an early warning prompt is automatically issued when the safety index is less than the index threshold.

[0053] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0054] 1. Accurate tool analysis: This application only analyzes the components that appear in the real-time video, retrieves the disabled tool list based on the status of these components, determines whether the current tool is included, and marks the component code. Compared with the fixed number of component analysis and fixed tool range judgment in traditional technologies, it is more accurate and reliable.

[0055] 2. Intelligent distance setting: The setting of the safety distance in this application refers to the operating score of the personnel using the tool. According to the difference between the score and the average score of other personnel, the proficiency of the personnel in using the tool is intelligently compared, excluding the influence of the operating difficulty of the tool itself. It is more intelligent than the fixed safety distance design adopted by traditional technology.

[0056] 3. Efficient early warning prompts: In this application, the possibility of the next operating part of the personnel is analyzed through historical logs and the weight coefficient is calculated. The real-time three-dimensional distance between the corresponding part of the current tool and the personnel is disabled and combined with the weight coefficient to calculate the safety index. Compared with traditional technologies, it can provide early warning more efficiently and quickly.

[0057] In summary, compared with traditional technologies, the present invention has the advantages of accurate tool analysis, intelligent distance setting and efficient early warning prompts, and can improve the efficiency of distribution equipment operation monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0059] Figure 1 It is a flow chart of an intelligent monitoring method for distribution network equipment based on artificial intelligence of the present invention;

[0060] Figure 2 It is a structural schematic diagram of an intelligent monitoring system for distribution network equipment based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] See also Figure 1 The present invention provides an intelligent monitoring method for distribution network equipment based on artificial intelligence, the method comprising the following steps:

[0063] S100, collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information during the operation and the status of each component under the distribution network equipment.

[0064] S200, build a scene model through a structural three-dimensional map, use an artificial intelligence algorithm to identify the target in the image, associate the identified components with the components in the scene model, identify the current tool and calculate the safety distance through the data in the historical record.

[0065] S300, analyzing the historical logs and establishing different execution order sets, marking the spatial position of the camera in the scene model, and marking the component code that disables the current tool. By analyzing the execution order set, the weight coefficient of the component corresponding to each marked code is calculated, and the safety index of the component corresponding to each marked code is calculated according to the spatial position of the camera in the scene model.

[0066] S400: When the safety index is less than the index threshold, the personnel are reminded to change the current tool or keep a certain distance.

[0067] In S100, the historical log refers to the operation records of all personnel, and each operation record includes the personnel identifier, component code, tool code, operation score and operation time. The three-dimensional structural diagram refers to the three-dimensional diagram of the internal structure of the power distribution equipment, specifically including the spatial position distribution of all components in the power distribution equipment. The tool comparison table refers to the tool list of each component in different states, specifically including the component code, status, available tool list and disabled tool list. The status includes power on and power off, and the list includes the codes of different tools. Image information refers to real-time video, which is collected by the camera on the wearable device of the personnel. The status of the component includes power on and power off, and is fed back in real time by the distribution network equipment.

[0068] The operation score indicates the degree of standardization of personnel's use of tools. The camera collects real-time video of personnel using tools in the process of operation, and uses a deep learning model to detect the key points of hands and tools in the video. The action features are extracted and compared with standard actions to calculate the action standardization score. The score ranges from 1 to 10.

[0069] The operation record is a record of each time a person uses a tool to operate a component. If a person changes the tool or adjusts the component during the operation, a new operation record will be generated. The operation score is limited to the degree of standardization of the person's actions within a certain period of time without changing the tool or adjusting the component.

[0070] The tool list is set in advance by the staff, with specific reference to the available tools and prohibited tools of the components in different states. Available tools refer to tools that can be used safely in this state and will not cause equipment damage or casualties; prohibited tools refer to tools that cannot be used safely in this state and may cause equipment damage or casualties.

[0071] The wearable device refers to a smart helmet, which has real-time shooting and micro-projection functions. It can shoot video within the field of vision as the person moves his body or turns his head. The micro-projection can project the set pattern to the designated location.

[0072] In S200, the specific steps are as follows:

[0073] S201: When a person performs operations on a distribution network device, the wearable device automatically interacts with the distribution network device to obtain the person's identifier stored in the wearable device. As well as the three-dimensional structure diagram saved in the distribution network equipment, the scene model is built according to the three-dimensional structure diagram using three-dimensional modeling software. The scene model includes a three-dimensional diagram of the distribution equipment, which can obtain the spatial position coordinates of each component.

[0074] S202: Establish a component set, use artificial intelligence algorithms to detect targets in real-time video, and analyze and determine whether each identified target is a tool or a component. If the target is a tool, obtain the code If the target is a component, the code is obtained and put into the component collection. After all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model.

[0075] The code of the component in the component collection changes with the video screen. When other component targets are added or the original component targets are lost in the video screen, the code of the corresponding component is added or deleted in the component collection. According to the component corresponding to the code in the component collection, the corresponding component is found in the scene model for association.

[0076] S203, searching the historical log for the identifier and coded as The operation scores of these operation records are summed up and the average value is calculated. . Retrieve the code in the history log operation records, sum up the operation scores in these operation records and calculate the average value .Will and Substitute the formula to calculate the safety distance, the formula is as follows:

[0077]

[0078] In the formula, For a safe distance, is the standard distance, is a constant, Score maximum action.

[0079] The safety distance is related to the average operation score of personnel using the tool. The lower the average operation score, the lower the proficiency of the personnel in using the tool. They are unable to master the tool's operating skills or safety precautions. Therefore, a larger safety distance needs to be set to ensure the safety of equipment or personnel.

[0080] In S300, the specific steps are as follows:

[0081] S301. All operation records in the history log are classified according to the identifiers. Operation records with the same identifier are first classified into the same category. Then, operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group. Each group contains at least two operation records. Count the number of all groups. and establish An execution order set is created, and the codes in the operation records of the same group are placed into the execution order set in the order of operation time.

[0082] Each execution sequence set represents the work component trajectory of the same person within a period of time, and the work component trajectory is related to the circuit design logic of the distribution equipment or the standard work requirement sequence.

[0083] S302, obtain the current status of each component of the distribution network device in real time feedback, retrieve the list of disabled tools corresponding to each code in the component set in the current state through the comparison table, and the disabled tool list contains Analyze the components and distances between components in the video in real time, compare them with the three-dimensional structure diagram, use visual SLAM algorithm to analyze the position and posture of the camera, so as to determine the spatial position of the camera relative to the distribution network equipment, and mark the spatial position coordinates of the camera in the scene model in real time. .

[0084] S303: Obtaining distance space position coordinates in the scene model The code of the component with the closest 3D distance . Filter out all the codes that contain And then contains at least one execution order set of marked code, these execution order sets Delete all previous codes. Determine whether each marked code exists in at least one of the execution order sets. If the result is no, set the weight coefficient of the corresponding component of the marked code to zero. If the result is yes, analyze the marked code in each execution order set. The number of codes in the interval is substituted into the formula to calculate the weight coefficient of the component corresponding to the marking code.

[0085]

[0086] In the formula, is the weight coefficient, is the number of all execution order sets containing the marked code, For the The execution order set contains the marked code and The number of codes in the interval, is a constant.

[0087] Code The corresponding part indicates the part closest to the worker's space, and it is used as the part that the worker is currently working on. The weight coefficient indicates the possibility of disabling the part of the current tool as the next part that the worker is working on. The higher the weight coefficient, the greater the possibility.

[0088] S304, obtain the spatial position coordinates of the component corresponding to each marking code in the scene model, and calculate the spatial position coordinates respectively The three-dimensional distance between , together with the weight coefficient of the component corresponding to the marking code, are substituted into the formula to calculate the safety index of each component corresponding to the marking code in real time. The formula is as follows:

[0089]

[0090] In the formula, is the safety index, is a constant, is the number of marking codes, For the The weight coefficient of the component corresponding to each marking code.

[0091] In S400, the changes in the spatial position coordinates of the camera in the scene model are monitored in real time, and the safety index of the component corresponding to each marking code is calculated. When the safety index is less than the index threshold, the corresponding component is used as a warning component, and the micro-projector on the wearable device is used to project and display at the position of the warning component, prompting personnel to keep a certain distance from the warning component or replace the currently used tool.

[0092] The larger the weight coefficient, the smaller the base of the safety index. As the three-dimensional distance between the personnel and the components corresponding to the marking codes gradually decreases, early warning will be issued when the distance is less than the index threshold, thus avoiding simultaneous warnings for multiple components corresponding to marking codes in a short period of time.

[0093] See also Figure 2 The present invention provides an intelligent monitoring system for distribution network equipment based on artificial intelligence, which includes a data acquisition module, an intelligent planning module, a monitoring and analysis module, and an early warning prompt module.

[0094] The data acquisition module is used to collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information and the status of each component under the distribution network equipment. The intelligent planning module is used to build a scene model, use artificial intelligence algorithms to identify targets in images, associate identified components with components in the scene model, identify the current tool, and calculate the safety distance through data in the historical records. The monitoring and analysis module is used to establish different execution order sets, mark the spatial position of the camera in the scene model, mark the component code that disables the current tool, and calculate the weight coefficient and safety index of each component corresponding to the marked code. The early warning prompt module automatically warns according to changes in the safety index.

[0095] The data collection module includes an environmental information collection unit, a chart information collection unit and a historical log collection unit.

[0096] The environmental information acquisition unit is used to collect image information during the operation and the status of each component of the distribution network equipment. Image information refers to real-time video, and the status of components includes power on and power off.

[0097] The chart information collection unit is used to collect the three-dimensional structure diagram and the tool comparison table. The three-dimensional structure diagram refers to the three-dimensional diagram of the internal structure of the power distribution equipment, including the spatial position distribution of all components in the power distribution equipment. The tool comparison table refers to the tool list of each component in different states, including the component code, state, available tool list and disabled tool list, and the list includes the codes of different tools.

[0098] The history log collection unit is used to collect history logs. History logs refer to the operation records of all personnel, and each operation record includes the personnel identifier, component code, tool code, operation score and operation time.

[0099] The intelligent planning module includes a scene building unit and a distance calculation unit.

[0100] The scene building unit is used to build the scene model. When a person performs operations on the distribution network equipment, the person's identifier is obtained. As well as the structural 3D graph, the scene model is built according to the structural 3D graph. A component set is established, and an artificial intelligence algorithm is used to detect targets in real-time video. If the target is a tool, the encoding is obtained. If the target is a component, the code is obtained and put into the component collection. After all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model.

[0101] The distance calculation unit is used to calculate the safety distance; all identifiers are and coded as The operation scores in the operation records are summed up and the average value is calculated to get , and then encode all The operation scores in the operation records are summed up and the average value is calculated to get ; Substitute into the formula: Calculate the safe distance .in, is the standard distance, is a constant, Score maximum action.

[0102] The monitoring and analysis module includes a component analysis unit and a safety monitoring unit.

[0103] The component analysis unit is used to mark the component code that disables the current tool.

[0104] First, the operation records with the same identifier in the historical log are classified into the same category, and then the operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group. Second, the number of all groups is counted. and establish The codes in the same group of operation records are put into the execution order set in the order of operation time. Finally, the current status of each component is obtained, and the disabled tool list of each code corresponding to the component in the component set in the current state is retrieved from the comparison table. The disabled tool list contains Then mark the corresponding code.

[0105] The safety monitoring unit is used to calculate the safety index of each component corresponding to the marking code.

[0106] First, compare the components appearing in the video with the three-dimensional structure diagram to determine the spatial position of the camera relative to the distribution network equipment and mark the spatial position coordinates in the scene model. , get the distance space position coordinates in the scene model The code of the component with the closest 3D distance .

[0107] Second, filter out all the code containing And then contains at least one execution order set of marked code, these execution order sets If the marked code does not exist in any of the execution order sets, the weight coefficient of the corresponding component of the marked code is set to zero. In other cases, the marked code is analyzed in each execution order set. The number of codes in the interval, substitute into the formula: Calculate the weight coefficient of the component corresponding to the marking code .in, is the number of all execution order sets containing the marked code, For the The execution order set contains the marked code and The number of codes in the interval, is a constant.

[0108] Finally, the spatial position coordinates and The three-dimensional distance between , substitute into the formula: , calculate the safety index of each component corresponding to the marking code. is a constant, is the number of marking codes, For the The weight coefficient of the component corresponding to each marking code.

[0109] The early warning prompt module is used for automatic early warning. The safety index of each component corresponding to the marking code is calculated in real time, and an early warning prompt is automatically issued when the safety index is less than the index threshold.

[0110] Embodiment 1:

[0111] Assume that person A1 uses tool B1 to operate on the power distribution equipment. The average value of the sum of the operation scores of all operation records with the identifier A1 and the code B1 in the historical log is 3.2, and the average value of the sum of the operation scores of all operation records with the code B1 is 5.8. The standard distance is 1m. is 1, the maximum operation score is 10, substitute it into the formula to calculate the safety distance:

[0112] Safe distance: m.

[0113] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0114] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent monitoring method for distribution network equipment based on artificial intelligence, characterized in that: The method comprises the following steps: S100, collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information during the operation and the status of each component under the distribution network equipment; S200, build a scene model through a structural three-dimensional image, use an artificial intelligence algorithm to identify the target in the image, associate the identified components with the components in the scene model, identify the current tool and calculate the safety distance through the data in the historical record; S300, analyzing historical logs and establishing different execution order sets, marking the spatial position of the camera in the scene model, and marking the component code that disables the current tool; calculating the weight coefficient of the component corresponding to each marked code by analyzing the execution order set, and calculating the safety index of the component corresponding to each marked code according to the spatial position of the camera in the scene model; S400: When the safety index is less than the index threshold, remind the personnel to change the current tool or keep a certain distance; In S300, the specific steps are as follows: S301. All operation records in the history log are classified according to the identifiers. Operation records with the same identifier are first classified into the same category. Then, operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group. Each group contains at least two operation records. Count the number of all groups SL z And establish SL z An execution order set is created, and the codes in the operation records of the same group are put into the execution order set in the order of operation time; S302, obtain the current status of each component of the distribution network device in real time feedback, retrieve the list of disabled tools corresponding to each code in the component set in the current state through the comparison table, and the disabled tool list contains GJ n The corresponding code is marked; the components and the distances between the components in the video are analyzed in real time, compared with the three-dimensional structure diagram, and the position and posture of the camera are analyzed using the visual SLAM algorithm to determine the spatial position of the camera relative to the distribution network equipment, and the spatial position coordinates (x ca ,y ca , z ca ); S303, obtain the distance space position coordinates (x ca ,y ca , z ca ) Code of the component closest to the three-dimensional distance BJ u ; Filter out all the codes containing BJ u And then contain at least one set of execution order of the marked code, these execution order sets are combined into BJ u All previous codes are deleted; determine whether each marked code exists in at least one of the execution order sets, if the result is no, set the weight coefficient of the corresponding component of the marked code to zero; if the result is yes, analyze the marked code in each execution order set and BJ u The number of interval codes is substituted into the formula to calculate the weight coefficient of the component corresponding to the marking code: Where QX is the weight coefficient, j is the number of all execution order sets containing the marked code, and jg i Mark the code and BJ in the i-th execution order set u The number of codes in the interval, β is a constant; S304, obtain the spatial position coordinates of the component corresponding to each marking code in the scene model, and calculate the spatial position coordinates (x ca ,y ca , z ca ) between the three-dimensional distance SW b , together with the weight coefficient of the component corresponding to the marking code, are substituted into the formula to calculate the safety index of each component corresponding to the marking code in real time. The formula is as follows: Where AQZ is the security index, γ is a constant, p is the number of marking codes, QX e is the weight coefficient of the component corresponding to the e-th marking code.

2. According to claim 1, a distribution network equipment intelligent monitoring method based on artificial intelligence is characterized in that: In S100, the historical log refers to the operation records of all personnel, and each operation record includes the personnel identifier, component code, tool code, operation score and operation time; the structural three-dimensional diagram refers to the three-dimensional diagram of the internal structure of the distribution equipment, specifically including the spatial position distribution of all components in the distribution equipment; the tool comparison table refers to the tool list of each component in different states, specifically including the component code, status, available tool list and disabled tool list; the status includes power on and power off, and the list includes the codes of different tools; the image information refers to real-time video, which is collected by the camera on the personnel's wearable device; the status of the component includes power on and power off, and is fed back in real time by the distribution network equipment.

3. According to claim 2, a distribution network equipment intelligent monitoring method based on artificial intelligence is characterized in that: In S200, the specific steps are as follows: S201: When a person performs operations on a distribution network device, the wearable device automatically interacts with the distribution network device to obtain the person's identifier RY stored in the wearable device. k As well as the three-dimensional structural diagram stored in the distribution network equipment, a scene model is built based on the three-dimensional structural diagram using three-dimensional modeling software; S202, establish a component set, use artificial intelligence algorithm to detect the target in real-time video, analyze and determine whether each identified target is a tool or a component; if the target is a tool, obtain the code GJ n , if the target is a component, the code is obtained and put into the component collection; after all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model respectively; S203, search the historical log for the identifier RY k And the code is GJ n The operation scores of these operation records are summed up and the average is calculated to get PF ry ; Search for code GJ in the history log n The operation scores in these operation records are summed up and the average is calculated to get PF ave PF ry and PF ave Substitute the formula to calculate the safety distance, the formula is as follows: In the formula, AQJ is the safety distance, JL is sta is the standard distance, α is a constant, PF max Score maximum action.

4. The method for intelligent monitoring of distribution network equipment based on artificial intelligence according to claim 3 is characterized in that: In S400, the changes in the spatial position coordinates of the camera in the scene model are monitored in real time, and the safety index of the component corresponding to each marking code is calculated. When the safety index is less than the index threshold, the corresponding component is used as a warning component, and the micro-projector on the wearable device is used to project and display at the position of the warning component, prompting personnel to keep a certain distance from the warning component or replace the currently used tool.

5. An intelligent monitoring system for distribution network equipment based on artificial intelligence, characterized in that: The system includes a data acquisition module, an intelligent planning module, a monitoring and analysis module, and an early warning module; The data acquisition module is used to collect historical logs, structural three-dimensional diagrams and tool comparison tables, as well as image information and the status of each component under the distribution network equipment; the intelligent planning module is used to build a scene model and use artificial intelligence algorithms to identify targets in images; the identification components are associated with the components in the scene model, the current tool is identified, and the safety distance is calculated through the data in the historical records; the monitoring and analysis module is used to establish different execution order sets, mark the spatial position of the camera in the scene model, and mark the component code that disables the current tool, and calculate the weight coefficient and safety index of each component corresponding to the marking code; the early warning prompt module automatically warns according to changes in the safety index; The monitoring and analysis module includes a component analysis unit and a safety monitoring unit; The component analysis unit is used to mark the component code that disables the current tool; First, the operation records with the same identifier in the historical log are classified into the same category, and then the operation records in the same category whose operation time interval is less than the duration threshold are classified into the same group; secondly, the number of all groups SL is counted z And establish SL z The codes in the same group of operation records are put into the execution order set in the order of operation time; finally, the current state of each component is obtained, and the disabled tool list of each code corresponding to the component in the component set in the current state is retrieved from the comparison table. The disabled tool list contains GJ n Then mark the corresponding code; The safety monitoring unit is used to calculate the safety index of each component corresponding to the marking code; First, the components appearing in the video are compared with the three-dimensional structure diagram to determine the spatial position of the camera relative to the distribution network equipment and mark the spatial position coordinates (x ca ,y ca , z ca ), get the distance space position coordinates (x ca ,y ca , z ca ) Code of the component closest to the three-dimensional distance BJ u ; Secondly, filter out all the files containing the code BJ u And then contain at least one set of execution order of the marked code, these execution order sets are combined into BJ u All previous codes are deleted; if the marked code does not exist in any of the execution order sets, the weight coefficient of the corresponding component of the marked code is set to zero. In other cases, the marked code is analyzed in each execution order set with BJ u The number of codes in the interval, substitute into the formula: , calculate the weight coefficient QX of the component corresponding to the marked code; where j is the number of all execution order sets containing the marked code, jg i Mark the code and BJ in the i-th execution order set u The number of codes in the interval, β is a constant; Finally, the spatial position coordinates and (x ca ,y ca , z ca ) between the three-dimensional distance SW b , substitute into the formula: , calculate the safety index of each component corresponding to the marking code; where γ is a constant, p is the number of marking codes, QX e is the weight coefficient of the component corresponding to the e-th marking code.

6. The distribution network equipment intelligent monitoring system based on artificial intelligence according to claim 5 is characterized by: The data collection module includes an environmental information collection unit, a chart information collection unit and a historical log collection unit; The environmental information acquisition unit is used to collect image information during the operation and the status of each component of the distribution network equipment; the image information refers to real-time video, and the status of the component includes power on and power off; The chart information acquisition unit is used to collect the three-dimensional structure diagram and the tool comparison table; the three-dimensional structure diagram refers to the three-dimensional diagram of the internal structure of the power distribution equipment, specifically including the spatial position distribution of all components in the power distribution equipment; the tool comparison table refers to the tool list of each component in different states, specifically including the component code, state, available tool list and disabled tool list, and the list includes the codes of different tools; The history log collection unit is used to collect history logs; history logs refer to operation records of all personnel, and each operation record includes a personnel identifier, a component code, a tool code, an operation score, and an operation time.

7. The distribution network equipment intelligent monitoring system based on artificial intelligence according to claim 6 is characterized by: The intelligent planning module includes a scene building unit and a distance calculation unit; The scene building unit is used to build a scene model; When personnel are working on distribution network equipment, obtain the personnel identifier RY k As well as the structural 3D graph, the scene model is built according to the structural 3D graph; a component set is established, and an artificial intelligence algorithm is used to detect targets in real-time video. If the target is a tool, the encoding GJ is obtained. n , if the target is a component, the code is obtained and put into the component collection; after all targets are judged, the components corresponding to all codes in the component collection are associated with the components in the scene model respectively; The distance calculation unit is used to calculate the safety distance; all identifiers are RY k And the code is GJ n The operation scores in the operation records are summed up and the average is calculated to get PF ry , and then encode all as GJ n The operation scores in the operation records are summed up and the average is calculated to get PF ave ; Substitute into the formula: , calculate the safety distance AQJ; where JL sta is the standard distance, α is a constant, PF max Score maximum action.

8. The distribution network equipment intelligent monitoring system based on artificial intelligence according to claim 7 is characterized in that: The early warning prompt module is used for automatic early warning; The safety index of the component corresponding to each marking code is calculated in real time, and an automatic warning prompt is issued when the safety index is less than the index threshold.

Images