Safety warning method and device for power distribution network operation and electronic equipment

Abstract

The application discloses a kind of safety warning method, device and electronic equipment of network power operation, wherein the method comprises: obtaining the first perception data of the first preset area outside network power operation site, and the second perception data of the second preset area in network power operation site;First target motion behavior is predicted in first preset area based on first perception data, and first prediction result is obtained, wherein first prediction result is used to indicate whether first target will enter network power operation site;The power operation behavior result of personnel target in multiple second targets in second preset area is predicted based on second perception data, and second prediction result is obtained, wherein second prediction result is used to indicate whether personnel target can safely execute power operation;Safety warning operation is executed based on first prediction result and second prediction result.The present application solves the technical problem that the efficiency of safety warning for network power operation in the related art is relatively low.

Description

Technical Field

[0001] This invention relates to the field of power operation safety, and more specifically, to a safety warning method, device, and electronic equipment for power distribution network operations. Background Technology

[0002] When carrying out power distribution network work, warnings and reminders can be given about potential safety risks and dangers to prevent accidents from occurring. In other words, safety warnings for power distribution network work can ensure the safety of personnel and equipment during the work process. By giving safety warnings for power distribution network work, work risks can be reduced and the safety of people's lives and property can be protected.

[0003] Currently, the safety warnings for power distribution network operations in related technologies usually involve setting up warning signs and notices at the work site. This approach to safety warnings is relatively passive and untimely, resulting in low efficiency.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, and electronic device for safety warnings during power distribution network operations, in order to at least solve the technical problem of low efficiency in providing safety warnings for power distribution network operations in related technologies.

[0006] According to one aspect of the present invention, a safety warning method for power distribution network operations is provided, comprising: acquiring first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site; predicting the movement behavior of a first target within the first preset area based on the first sensing data to obtain a first prediction result, wherein the first prediction result indicates whether the first target will enter the power distribution network operation site; predicting the power operation behavior of a personnel target among a plurality of second targets within the second preset area based on the second sensing data to obtain a second prediction result, wherein the second prediction result indicates whether the personnel target can safely perform power operations, the plurality of second targets including at least a personnel target, power equipment within the second preset area, and a tool target used by the personnel target, the tool target indicating the power tool used by the personnel target to perform power operations; and performing a safety warning operation based on the first prediction result and the second prediction result.

[0007] Optionally, predicting the motion behavior of the first target within the first preset area based on the first sensing data to obtain a first prediction result includes: determining the position, type, and motion state of the first target based on the first sensing data; and determining the first prediction result based on the position, type, and motion state.

[0008] Optionally, a first prediction result is determined based on location, type, and motion status, including: determining a first risk prediction value based on location, type, and motion status; determining the on-site risk level value of the power distribution network operation site based on the type information of the power operation; fusing the first risk prediction value and the on-site risk level value to obtain a target risk value; and determining the first prediction result based on the comparison result between the target risk value and a preset risk threshold.

[0009] Optionally, based on the second sensing data, the power operation behavior of personnel targets among multiple second targets in the second preset area is predicted to obtain a second prediction result, including: determining the type matching relationship and location spatial relationship among multiple second targets based on the second sensing data; matching the type matching relationship and location spatial relationship with preset safety rules to obtain a matching result; and determining the second prediction result based on the matching result.

[0010] Optionally, the type matching relationship includes at least the target type matching relationship between the type of personnel target and the type of tool target, and the location spatial relationship includes at least the target location spatial relationship between the location of personnel target and the location of power equipment; determining the type matching relationship and location spatial relationship between multiple second targets based on the second sensing data includes: acquiring the first type of personnel target, the first location of personnel target, the second type of tool target and the second location of power equipment from the second sensing data; determining the target type matching relationship based on the first type and the second type; determining the target location spatial relationship based on the first location and the second location; and determining the type matching relationship and location spatial relationship based on the target type matching relationship and the target location spatial relationship.

[0011] Optionally, determining the second prediction result based on the matching result includes: constructing a dangerous behavior pattern feature library, wherein the dangerous behavior pattern feature library includes at least feature data of mismatch between personnel and the electrical tools used, and feature data of personnel not maintaining a safe distance from the electrical equipment being operated; and matching the matching result with the dangerous behavior pattern feature library to determine the second prediction result.

[0012] According to another aspect of the present invention, a safety warning device for power distribution network operations is also provided, comprising: an acquisition module, configured to acquire first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site; a first prediction module, configured to predict the movement behavior of a first target within the first preset area based on the first sensing data, and obtain a first prediction result, wherein the first prediction result indicates whether the first target will enter the power distribution network operation site; a second prediction module, configured to predict the power operation behavior of a personnel target among a plurality of second targets within the second preset area based on the second sensing data, and obtain a second prediction result, wherein the second prediction result indicates whether the personnel target can safely perform power operations, the plurality of second targets including at least a personnel target, power equipment within the second preset area, and a tool target used by the personnel target, the tool target indicating the power tool used by the personnel target to perform power operations; and an execution module, configured to execute a safety warning operation based on the first prediction result and the second prediction result.

[0013] According to another aspect of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods of various embodiments of the present invention during runtime.

[0014] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0015] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0016] According to another aspect of the present invention, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0017] According to another aspect of the present invention, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of the present invention.

[0018] In this embodiment of the invention, a safety warning method for power distribution network operations is provided, comprising: acquiring first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site; predicting the movement behavior of a first target within the first preset area based on the first sensing data to obtain a first prediction result, wherein the first prediction result indicates whether the first target will enter the power distribution network operation site; predicting the power operation behavior of a personnel target among a plurality of second targets within the second preset area based on the second sensing data to obtain a second prediction result, wherein the second prediction result indicates whether the personnel target can safely perform power operations, the plurality of second targets including at least a personnel target, power equipment within the second preset area, and a tool target used by the personnel target, the tool target indicating the power tool used by the personnel target to perform power operations; and performing a safety warning operation based on the first prediction result and the second prediction result. It is noteworthy that this application comprehensively acquires first perception data of a first preset area outside the power distribution network operation site and second perception data of a second preset area inside the power distribution network operation site. Based on the first prediction result, it can pre-screen and restrict moving objects allowed to enter the power distribution network operation site. Based on the second prediction result, it can help determine whether personnel targets can safely perform power operations. By comprehensively considering the first and second prediction results and performing safety warning operations, it can effectively improve the safety of power distribution network operations. By predicting potential safety risks in advance, it can reduce the possibility of accidents and ensure the safety of personnel and equipment, thereby solving the technical problem of low efficiency in providing safety warnings for power distribution network operations in related technologies. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0020] Figure 1 This is a flowchart of a safety warning method for power distribution network operations according to an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of a safety warning device for power distribution network operations according to an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] According to one aspect of the present invention, a safety warning method for power distribution network operations is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0025] Figure 1 This is a flowchart of a safety warning method for power distribution network operations according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0026] Step S102: Obtain first sensing data of the first preset area outside the power distribution network operation site and second sensing data of the second preset area inside the power distribution network operation site.

[0027] The aforementioned power distribution network operation site refers to the power distribution network operation site within the power grid system, which may include, but is not limited to, substations, distribution rooms, distribution cabinets, cable lines, etc. At the power distribution network operation site, the following work can be performed: inspection and maintenance, i.e., regularly inspecting and maintaining power equipment to ensure normal operation and extend equipment life; safety inspection, i.e. regularly conducting safety inspections of power distribution equipment to identify and promptly eliminate potential safety hazards, ensuring site safety; emergency repair, i.e. promptly carrying out emergency repairs to restore power supply and ensure user electricity needs are met in the event of a fault or emergency; and on-site construction, i.e. performing line repairs, equipment replacements, and other construction work to meet new power demands or improve power grid facilities. The specific power distribution network operation site can be determined according to actual needs and is not limited here.

[0028] The aforementioned first preset area can refer to a preset area outside the power distribution network operation site, such as an area close to the power distribution network operation site or the entrance area of ​​the power distribution network operation site. The first preset area can be determined according to actual needs, and is not limited here.

[0029] The aforementioned first sensing data can refer to data such as images, videos, radar, sound, and vibration of the first prediction area collected or sensed by various devices. The first sensing data can be determined according to actual needs and is not limited here.

[0030] The aforementioned second preset area can refer to a preset area within the power distribution network operation site, or it can be the entire area or a designated area within the power distribution network operation site. The second preset area can be determined according to actual needs, and there is no limitation here.

[0031] The aforementioned second sensing data can refer to data such as images, videos, radar, sound, and vibration of the second prediction area collected or sensed by various devices. The second sensing data can be determined according to actual needs and is not limited here.

[0032] In one optional embodiment, first sensing data from a first preset area outside the power distribution network operation site and second sensing data from a second preset area inside the power distribution network operation site can be acquired. This helps to promptly grasp the on-site situation and prevent accidents. Before conducting power distribution network operations, various sensing devices, such as cameras, sound sensors, and gas detectors, can be used to set up sensing devices around the first preset area, operate the sensing devices, and monitor changes in the surrounding environment, such as personnel activities, sounds, and odors. The sensing data is then transmitted to a monitoring center or on-site command center for real-time monitoring and recording to obtain the first sensing data. Similarly, various sensing devices, such as surveillance cameras, sound sensors, and vibration sensors, can be set up in the second preset area, operate the sensing devices, and monitor changes in environmental parameters within the work area, such as temperature, humidity, and vibration. The sensing data is then transmitted to a monitoring center or on-site command center for real-time monitoring and recording to obtain the second sensing data. By acquiring sensing data from the first and second preset areas, the on-site situation can be understood in a timely manner, potential safety hazards can be detected early, accidents can be effectively prevented, and the safe conduct of power distribution network operations can be ensured.

[0033] Step S104: Based on the first perception data, predict the motion behavior of the first target within the first preset area to obtain the first prediction result.

[0034] The first prediction result is used to indicate whether the first target will enter the power distribution network operation site.

[0035] The first target mentioned above can refer to movable targets such as people, animals, and vehicles that enter the first prediction area. The specific type of the first target can be determined according to the actual situation and is not limited here.

[0036] In one optional embodiment, the first sensing data can be processed to extract features related to the movement behavior of the first target. These features may include information such as the target's movement direction and speed changes. Using the extracted feature data, a prediction model or algorithm can be used to predict the movement behavior of the first target, i.e., to predict the movement behavior of the first target within a first preset area, resulting in a first prediction result. This first prediction result indicates whether the target will enter the power distribution network operation site. This facilitates subsequent risk assessment of the first target's movement behavior based on the first prediction result. If the prediction result indicates that the target may enter the power distribution network operation site, timely safety measures need to be taken. For example, the type of the first target can be determined first. If the first target is a person, the person's identity information can be further determined. If the first target is detected as an animal, measures can be taken to limit the first target's entry. If the first target is a vehicle, information such as the vehicle's license plate can be identified to achieve safety detection and screening of the first target entering the power distribution network operation site, thereby ensuring the safety of the operators and equipment.

[0037] Step S106: Based on the second perception data, predict the power operation behavior of personnel targets among multiple second targets in the second preset area to obtain the second prediction result.

[0038] The second prediction result is used to indicate whether the personnel target can safely perform electrical work. The multiple second targets include at least the personnel target, the electrical equipment in the second preset area, and the tool target used by the personnel target. The tool target is used to indicate the electrical tools used by the personnel target to perform electrical work.

[0039] The personnel targets mentioned above refer to personnel entering the power distribution network operation site. Personnel targets can be further classified according to the type of power operation. The type of personnel targets can be identified by their work clothes, safety helmets, work badges, etc.

[0040] The aforementioned power tools may include, but are not limited to, screwdrivers, wrenches, electric drills, electric saws, etc. of different types and insulation levels. The specific power tools can be determined according to actual needs and are not limited here.

[0041] In one optional embodiment, the second sensing data can be processed to determine the location, type, or movement trajectory and actions of the second target, including information on power equipment and tool targets. Appropriate machine learning algorithms or deep learning models can be selected for prediction to forecast the power operation behavior of personnel targets within the second target group. The prediction result can be a binary classification, including safe and dangerous, or a multi-class classification, including safe, general, and dangerous, to indicate whether the personnel target can safely perform power operations. This facilitates subsequent evaluation of the prediction results. Through the implementation of the above steps, the power operation behavior of personnel targets within multiple second targets in a second preset area can be predicted based on the second sensing data, resulting in a second prediction result. This effectively assesses the safety risks of personnel targets in power operations, improving operational safety and efficiency.

[0042] Step S108: Perform a safety warning operation based on the first prediction result and the second prediction result.

[0043] The aforementioned safety warning operations may include, but are not limited to, closing the entrance to the power distribution network operation site to restrict the entry of the primary target, generating audible and visual alarm information, indicating the level of power operation hazards to on-site operators, and sending safety warning record reports to the remote data center. The safety warning operations can be determined according to actual needs and are not limited here.

[0044] In one optional embodiment, based on the first prediction result, the system can predict whether a first target will enter the power distribution network operation site. If the system deems there a potential risk, it will generate a corresponding prediction result. The system can also determine the status of the power distribution network operation site and the location of the personnel target based on the second prediction result, in order to predict whether the personnel target can safely perform power operations. Based on the generated first and second prediction results, corresponding safety warning operations can be performed, which may include closing the entrance to the power distribution network operation site to restrict the entry of the first target, generating audible and visual alarm information, and prompting on-site operators about the power operation hazard level. The system can also send a safety warning record report to a remote data center for subsequent analysis and recording, which can help managers understand the safety situation of the power distribution network operation site and take necessary measures to ensure the safety of operators. Through the specific implementation process of the above steps, the safety of the power distribution network operation site can be effectively improved, the possibility of accidents can be reduced, and the safety of operators can be guaranteed.

[0045] In this embodiment of the invention, a safety warning method for power distribution network operations is provided, comprising: acquiring first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site; predicting the movement behavior of a first target within the first preset area based on the first sensing data to obtain a first prediction result, wherein the first prediction result indicates whether the first target will enter the power distribution network operation site; predicting the power operation behavior of a personnel target among a plurality of second targets within the second preset area based on the second sensing data to obtain a second prediction result, wherein the second prediction result indicates whether the personnel target can safely perform power operations, the plurality of second targets including at least a personnel target, power equipment within the second preset area, and a tool target used by the personnel target, the tool target indicating the power tool used by the personnel target to perform power operations; and performing a safety warning operation based on the first prediction result and the second prediction result. It is noteworthy that this application comprehensively acquires first perception data of a first preset area outside the power distribution network operation site and second perception data of a second preset area inside the power distribution network operation site. Based on the first prediction result, it can pre-screen and restrict moving objects allowed to enter the power distribution network operation site. Based on the second prediction result, it can help determine whether personnel targets can safely perform power operations. By comprehensively considering the first and second prediction results and performing safety warning operations, it can effectively improve the safety of power distribution network operations. By predicting potential safety risks in advance, it can reduce the possibility of accidents and ensure the safety of personnel and equipment, thereby solving the technical problem of low efficiency in providing safety warnings for power distribution network operations in related technologies.

[0046] Optionally, predicting the motion behavior of the first target within the first preset area based on the first sensing data to obtain a first prediction result includes: determining the position, type, and motion state of the first target based on the first sensing data; and determining the first prediction result based on the position, type, and motion state.

[0047] In one optional embodiment, the location, type, and movement status of the first target can be determined based on the first sensing data. Various sensors, such as infrared sensors, radar sensors, and cameras, can be installed at the power distribution network work site to sense the surrounding environment. This allows for real-time monitoring of the location, type, and movement status of the first target, including people, animals, and vehicles. The data collected by the sensors can be sent to a data processing center for processing. The data processing center can analyze and process the data collected by the sensors, extracting information such as the location, type, and movement status of the first target. The data processing center can identify and classify the first target, categorizing it into different types such as people, animals, and vehicles. It can also identify the specific location and movement status of the first target. Based on the location, type, and movement status of the first target, predictive analysis can be performed to obtain a first prediction result indicating whether the first target will enter the power distribution network work site. If the prediction result indicates that the first target may enter the work site, the system will automatically trigger an alarm to remind the workers to pay attention to safety. Through the above steps, the power distribution network work site can obtain information about the surrounding first targets in a timely and accurate manner and perform predictive analysis, effectively preventing potential safety risks.

[0048] Optionally, a first prediction result is determined based on location, type, and motion status, including: determining a first risk prediction value based on location, type, and motion status; determining the on-site risk level value of the power distribution network operation site based on the type information of the power operation; fusing the first risk prediction value and the on-site risk level value to obtain a target risk value; and determining the first prediction result based on the comparison result between the target risk value and a preset risk threshold.

[0049] The aforementioned type information can refer to the type of power operation, which may include, but is not limited to, high-voltage line construction operations, low-voltage line construction operations, substation operation and maintenance operations, power distribution equipment installation operations, and power distribution equipment maintenance operations. The type information can be determined according to actual needs and is not limited here. Furthermore, based on the type information of power operations, different on-site risk level values ​​for power distribution network operations can be determined. The on-site risk level values ​​can be assigned or set according to the type information of power operations and are not limited here.

[0050] In one optional embodiment, a first risk prediction value can be determined based on the location, type, and motion state of a first target. Sensor technologies, such as cameras, radar, and infrared, can be used to monitor and identify the first target, acquiring information on its location, type, and motion state. Based on this information and a pre-set risk assessment model, the risk prediction value of the target is determined. The on-site risk level value of the power distribution network operation can be determined based on the type of power operation. This can be done according to the specific type of power operation, such as line inspection, equipment maintenance, or troubleshooting, combined with on-site environmental conditions, such as weather, terrain, and surrounding environment. The first risk prediction value and the on-site risk level value can be fused to obtain a target risk value. This can be achieved by weighted fusion of the first risk prediction value and the on-site risk level value to calculate the comprehensive risk value of the target. The first prediction result can be determined based on the comparison between the target risk value and a preset risk threshold. By comparing the target risk value with the preset risk threshold, measures can be taken in a timely manner to prevent potential risks and ensure the safety of the power distribution network operation site.

[0051] Optionally, based on the second sensing data, the power operation behavior of personnel targets among multiple second targets in the second preset area is predicted to obtain a second prediction result, including: determining the type matching relationship and location spatial relationship among multiple second targets based on the second sensing data; matching the type matching relationship and location spatial relationship with preset safety rules to obtain a matching result; and determining the second prediction result based on the matching result.

[0052] The aforementioned preset safety rules can refer to the pre-determined rules for users to safely perform electrical work. These rules may include rules for insulation safety protection during live work, rules for safe distances during live work, and rules for the proper use of electrical tools. The preset safety rules can be determined according to actual needs and are not limited here.

[0053] In an optional embodiment, based on the second sensing data, the type matching relationship and location spatial relationship between multiple second targets can be determined. For example, the relationship between personnel targets and power equipment, and the relationship between tool targets and personnel targets can be determined. The type matching relationship and location spatial relationship can be matched with preset safety rules. The preset safety rules may include prohibiting personnel targets from approaching certain dangerous equipment, restricting the scope of use of tool targets, etc. Based on the matching results of the safety rules, it can be determined whether there are safety hazards or whether safety conditions are met. If the matching result meets the safety rules, it means that the power operation can be safely performed. Finally, based on the matching result, a second prediction result can be determined, i.e., whether the personnel targets can safely perform the power operation. This may include whether further safety measures need to be taken, whether the position of the power equipment needs to be adjusted, etc. Through the implementation of the above steps, the safety status of personnel targets in power distribution network operations can be effectively predicted and evaluated, ensuring the smooth progress of power operations and protecting personnel safety.

[0054] Optionally, the type matching relationship includes at least the target type matching relationship between the type of personnel target and the type of tool target, and the location spatial relationship includes at least the target location spatial relationship between the location of personnel target and the location of power equipment; determining the type matching relationship and location spatial relationship between multiple second targets based on the second sensing data includes: acquiring the first type of personnel target, the first location of personnel target, the second type of tool target and the second location of power equipment from the second sensing data; determining the target type matching relationship based on the first type and the second type; determining the target location spatial relationship based on the first location and the second location; and determining the type matching relationship and location spatial relationship based on the target type matching relationship and the target location spatial relationship.

[0055] In an optional embodiment, a first type of personnel target, a first location of personnel target, a second type of tool target, and a second location of power equipment can be extracted from the second sensing data. The type matching relationship between personnel target and tool target can be determined based on the correspondence between the first type and the second type. The spatial relationship between the first location of personnel target and the second location of power equipment can be determined by comparing the first location of personnel target and the second location of power equipment. The type matching relationship and the spatial relationship between the target can be determined by comprehensively considering the target type matching relationship and the target spatial relationship. Through the above steps, accurate judgment and warning of the type matching relationship and spatial relationship between personnel target, tool target and power equipment can be achieved in power distribution network operations, thereby improving the safety and efficiency of power distribution network operations.

[0056] Optionally, determining the second prediction result based on the matching result includes: constructing a dangerous behavior pattern feature library, wherein the dangerous behavior pattern feature library includes at least feature data of mismatch between personnel and the electrical tools used, and feature data of personnel not maintaining a safe distance from the electrical equipment being operated; and matching the matching result with the dangerous behavior pattern feature library to determine the second prediction result.

[0057] In one optional embodiment, data related to mismatches between personnel and the electrical tools they use, and failure to maintain a safe distance between personnel and the electrical equipment they operate, can be collected. This data can be obtained through on-site observation, recording, employee training, etc. Feature extraction can be performed on the collected data to transform it into feature data that can be processed by a computer. For example, features of mismatches between personnel and electrical tools may include tool type, personnel identity information, etc.; features of failure to maintain a safe distance between personnel and electrical equipment may include distance, equipment type, etc. The extracted feature data can be organized into a feature library, including feature data of mismatches between personnel and the electrical tools they use, and failure to maintain a safe distance between personnel and the electrical equipment they operate. The hazardous behavior pattern feature library can be used to store reference data of hazardous behavior patterns. In actual operations, data can be collected by monitoring equipment and processed in real time to perform matching analysis on the relationship between personnel and tools / equipment. The matching results can be matched with the hazardous behavior pattern feature library to determine a second prediction result. If the matching results show the existence of a hazardous behavior pattern, the system will issue a warning and remind the operator to correct it in time. Through the above steps, a hazardous behavior pattern feature library can be effectively constructed, and the monitoring and early warning of hazardous behaviors can be realized in power distribution network operations, thereby improving the safety awareness and behavioral norms of operators.

[0058] The technical solution proposed in this application is described below in conjunction with an optional application scenario. This application proposes an intelligent warning method for proactive early warning of safety overruns, which is detailed as follows: First sensing data of the power distribution network operation site is obtained; the location, type and movement state of the first target approaching the power distribution network operation site are determined based on the first sensing data; and the location, type and movement state of the first target are used to determine whether the first target will enter the power distribution network operation site.

[0059] In this embodiment, multiple high-definition cameras, radar sensors, and infrared thermal imagers installed around the work site can be used to acquire first perception data of the power distribution network work site in real time. The high-definition cameras continuously capture video footage of the work site to generate first video data. The radar sensors continuously scan the surrounding space to capture the distance and speed information of the target and generate first radar data. The infrared thermal imager detects differences in thermal radiation to help identify targets at night or in poor visibility conditions, and its data serves as auxiliary perception data.

[0060] Target detection algorithms (e.g., YOLO) can be used to analyze the initial sensing data to quickly locate the position of the first target approaching the work site, such as pedestrians, vehicles, or animals, and mark the target bounding box. Image classification algorithms (e.g., ResNet) can be applied to further classify the image within the target bounding box to determine the type of the first target, such as a person, vehicle, or animal. By comparing the pixel position changes of the first target in consecutive video frames and combining the camera's frame rate parameters, the movement speed and direction of the first target can be calculated. Distance and speed information from radar data can be integrated to further verify and correct the target's motion state. Based on the position, type, and motion state of the first target, a weighted fusion method is used to calculate a comprehensive risk value. The comprehensive risk value considers the target's distance from the work site, the potential threat of its type, and the changing trend of its motion state. If the comprehensive risk value exceeds a preset risk threshold, the system can determine that the first target may enter the work site and trigger an early warning mechanism.

[0061] In this embodiment, by acquiring and processing multi-source sensing data in real time, the system can quickly and accurately identify targets approaching the work site and predict their movement trajectory, thereby providing early warnings. By combining various technologies such as high-definition video, radar scanning, and infrared thermal imaging, comprehensive monitoring of the work site and its surrounding environment can be achieved. At the same time, by utilizing image processing and machine learning algorithms, intelligent detection and recognition of targets are realized, improving the system's intelligence level.

[0062] In some embodiments, the first sensing data may include first image data, first video data, and first radar data; determining the location, type, and motion state of a first target approaching the power distribution network operation site based on the first sensing data, and determining whether the first target will enter the power distribution network operation site based on the location, type, and motion state of the first target, specifically includes: preprocessing the first sensing data to obtain third sensing data; determining the location of the first target using a target detection algorithm based on the third sensing data; determining the type of the first target using an image classification algorithm based on the third sensing data; determining the motion state of the first target based on the pixel position changes of the first target in two consecutive frames, camera frame rate parameters, and direction of position change based on the third sensing data; determining a comprehensive risk value based on the location, type, and motion state of the first target using a weighted fusion method; if the comprehensive risk value exceeds a preset risk threshold, it is determined that the first target will enter the power distribution network operation site.

[0063] In this embodiment, first perception data, including first image data, first video data, and first radar data, is acquired and stored in a perception data cache. The first perception data is read from the perception data cache and preprocessed using algorithms such as image enhancement and noise removal to obtain third perception data. Based on the third perception data, a pre-trained target detection model is used to extract image features and perform bounding box regression to determine the pixel coordinates of the first target in each frame. Based on the third perception data, a pre-trained image classification model is used to extract image features of the target region and compare them with a preset target type feature library to determine the type of the first target. The pixel coordinates of the first target in two consecutive frames are acquired, and its position change is calculated. Simultaneously, the camera's frame rate parameters are acquired, and the instantaneous velocity of the target is calculated based on the position change and frame rate. By analyzing the horizontal and vertical components of the target's instantaneous velocity and setting a velocity threshold, it is determined whether the target is stationary, moving at a constant speed, or accelerating. The position, type, and motion state information of the first target are fused to generate structured target description data.

[0064] Based on the acquired information such as location, type, and movement status, a support vector machine algorithm is used to predict the risk level of the first target, obtaining a first risk prediction value. Geographical location information of the power distribution network operation site is acquired, and a risk level model for the operation site is established based on this information. The location information of the first target is input into the risk level model of the operation site, and the on-site risk level value of the first target is calculated through the model. According to pre-set risk level weights, the first risk prediction value and the on-site risk level value are weighted and fused to obtain a comprehensive risk value for the first target. It is determined whether the comprehensive risk value of the first target exceeds a preset risk threshold. If it exceeds the risk threshold, it is determined that the first target will enter the power distribution network operation site and an early warning is triggered. If the comprehensive risk value of the first target does not exceed the risk threshold, the changes in the target's location are continuously monitored, the target's location information is updated in real time, and the above risk prediction and judgment process is repeated.

[0065] Furthermore, the preprocessing of the first perception data specifically includes: using a median filtering algorithm to remove noise from the first image data, and then using a scale-invariant feature transform (SIFT) algorithm to extract the scale-invariant feature transform features of the first image data; using a frame difference method to detect moving targets in the first video data; and using a Kalman filter algorithm to remove noise from the first radar data and extract the distance and speed information of moving targets in the first radar data.

[0066] In this embodiment, the first image data can be denoised using a median filtering algorithm to obtain the second image data; for the second image data, a scale-invariant feature transform algorithm is used to extract scale-invariant feature transform features to obtain the first feature data; the first video data is acquired, and the frame difference method is used to detect moving targets to determine whether there are moving targets; the first radar data is denoised using a Kalman filter to obtain the second radar data; the distance and speed information of the moving targets are extracted from the second radar data to obtain the second feature data; based on the first feature data and the second feature data, a support vector machine algorithm is used for target recognition, and the recognition result is output.

[0067] For example, firstly, the acquired first image data can be denoised using a median filtering algorithm. Median filtering is a non-linear filtering method that removes salt-and-pepper noise from the image by replacing the gray value of a pixel with the median of its neighborhood gray values. A 5×5 filtering window can be set, and for each pixel in the image, the 25 pixel values ​​within its 5×5 neighborhood are taken, sorted by size, and the value at the median is used as the filtering result for that point. This effectively smooths the image while preserving edge details. Then, the scale-invariant feature transform (SMT) algorithm is used to extract the scale-invariant features of the image. The SMT algorithm constructs a scale space using a difference-of-Gaussian pyramid, detects extreme points at multiple scales as keypoints, and calculates the gradient direction histogram of the keypoints as feature descriptors. Setting four scales, each with five levels, can extract a total of 1024-dimensional SMT feature vectors. The above parameter settings can also be determined according to actual needs and are not limited here.

[0068] Next, the frame difference method can be used to detect moving targets on the first video data. The frame difference method extracts the moving region by comparing the differences between two adjacent frames. The video frame rate can be set to 25 frames per second. If the absolute value of the grayscale difference between corresponding pixels in two consecutive frames is greater than a threshold of 30, it is determined to be a moving pixel, forming a binary motion mask. For connected component analysis, pixels with an area greater than 100 are considered moving targets. The above parameter settings can also be determined according to actual needs; no limitation is made here.

[0069] Finally, Kalman filtering can be used to denoise the first radar data. Kalman filtering recursively estimates the system state, continuously correcting the deviation between the predicted and observed values ​​to obtain the optimal result. Using the target's distance and velocity as state variables, a uniform linear motion model is established, setting the state transition matrix and observation matrix. The state at the next moment is predicted based on the current state, and the predicted value is then corrected using new observation data, making the estimated value continuously approach the true value, eliminating the influence of random noise, and improving the accuracy and stability of radar data. By integrating the processing results of multi-source data such as images, videos, and radar data, key information such as the target's location, type, and motion state can be accurately obtained, providing a reliable basis for subsequent risk warning and security precautions.

[0070] When it is determined that the first target will enter the power distribution network operation site, the target type can be matched with the preset authorized target feature library to determine whether the first target is an authorized target. If the first target is not an authorized target, an alarm will be issued.

[0071] In this embodiment, when the system determines that a first target is about to enter the work site, an authorization verification step is added to further ensure work safety. This step determines whether the target has been authorized to enter the work site by comparing the type of the first target with a preset authorized target feature library.

[0072] Specifically, after determining that the first target will enter the work site, the system first reviews and confirms the target's type, such as worker, manager, maintenance personnel, or external personnel. The system accesses a pre-set authorized target feature database, which contains all target types permitted to enter the work site and their corresponding feature information, such as identification, work badge number, RFID tag, appearance characteristics, clothing color and style, and time restrictions. Based on the type of the first target, the system searches for the corresponding entry in the feature database.

[0073] The system compares the actual characteristics of the first target with entries in the feature database, including verifying the validity of the identification, checking for consistency in appearance, and confirming whether the target is within the permitted entry time period. If all characteristics match, the first target is determined to be an authorized target; otherwise, it is an unauthorized target. If the first target is determined to be an unauthorized target, the system immediately triggers an alarm mechanism. Alarm methods may include displaying warning messages at the monitoring center, issuing audible and visual alarms, and sending SMS or email notifications to security management personnel. Simultaneously, the system may also activate other security measures, such as automatically locking the work site entrance and activating cameras to track unauthorized targets.

[0074] In this embodiment, by introducing an authorization verification step, the system can effectively prevent unauthorized personnel from entering the power distribution network work site, thereby significantly reducing the risk of safety accidents. The authorized target feature library can be dynamically updated and adjusted according to actual conditions to adapt to different safety requirements and work scenarios.

[0075] In some embodiments, the type of the first target can be matched with a preset authorized target feature library to determine whether the first target is an authorized target. If the first target is not an authorized target, an alarm is issued. Specifically, this includes: when the type of the first target is a non-human object, determining that the first target is not an authorized target and issuing an alarm; when the type of the first target is a human, acquiring the facial features of the first target, matching the facial features of the first target with the preset authorized target feature library, and using a feature matching algorithm to calculate the similarity between the facial features of the first target and the facial features of each authorized target in the authorized target feature library; if the similarity between the facial features of the first target and the facial features of any authorized target in the authorized target feature library is higher than a first preset similarity threshold, then determining that the first target is an authorized target.

[0076] In this embodiment, when the system detects that the type of the first target is a non-human object, such as an animal, it can directly determine it as an unauthorized illegal intrusion target, trigger an audible and visual alarm to issue a warning, and notify on-duty personnel to go to the scene for handling via SMS, telephone, or other means.

[0077] When the first target is detected as a person, the system automatically captures their facial image, extracts a 128-dimensional feature vector, and compares it one-to-one with each sample in the authorized personnel face database using the Euclidean distance metric, calculating a similarity score. If the similarity with any authorized target exceeds a threshold of 8, the person is deemed legitimate and authorized to enter the work site, and the system automatically switches the gate to passage mode. If the similarity is below 8, the person is identified as a stranger, the system keeps the gate closed, and provides a voice prompt, "Please contact on-site staff to complete the entry procedures." Simultaneously, the captured image of the person is displayed on a large screen for manual verification. The entire process can be timed within 1 second to ensure real-time response. The above parameter settings can be adjusted according to actual needs and are not limited here.

[0078] It can acquire second perception data from the power distribution network operation site, and determine the type and location of multiple second targets within the power distribution network operation site based on the second perception data. The types of second targets include personnel targets, power equipment targets, and tool targets.

[0079] In this embodiment, in order to comprehensively monitor and manage various elements at the power distribution network operation site, the system not only focuses on externally approaching targets, but also acquires and analyzes second perception data inside the operation site in real time. This data is used to determine the type and location of multiple second targets at the site, including personnel targets, power equipment targets, and tool targets.

[0080] Specifically, various sensors are deployed at the work site, such as radio frequency identification readers, location trackers, ultrasonic sensors, cameras, and dedicated power equipment status monitoring sensors. These sensors continuously collect real-time data of various targets on site, including location information, status information, and identity information, forming second-level sensing data.

[0081] The collected second-sensor data can be preprocessed, such as through noise reduction and format conversion, to ensure data accuracy and usability. Image recognition technology, such as convolutional neural networks, is used to process camera data, identifying the appearance and location of personnel and some power equipment targets. Radio frequency identification (RFID) tags are read using RFID readers to quickly determine the target type and location of personnel and tools wearing the tags. Power equipment status monitoring sensors directly provide real-time status information of the power equipment, including online status, voltage and current parameters, etc. Based on the processed data, targets are categorized into three main types: personnel targets, power equipment targets, and tool targets. Using the location information provided by the sensors, such as positioning coordinates and relative position coordinates, a real-time location map of each target is plotted in the system. For power equipment targets, in addition to location information, their operating status and key parameters are recorded for real-time monitoring and fault warning. The processed data is displayed in real-time to safety management personnel and operators through a monitoring interface, clearly showing the location, status, and quantity of each type of target. Interactive functions are provided, allowing management personnel to perform operations such as target search, trajectory playback, and status query through the interface.

[0082] In this embodiment, by monitoring and managing various targets at the work site in real time, the system can help workers quickly locate the necessary tools and electrical equipment, reducing search time and improving work efficiency. It can also promptly detect and report potential safety hazards, such as personnel illegally entering dangerous areas or abnormal conditions of electrical equipment, thereby enabling appropriate measures to be taken to prevent accidents.

[0083] In some embodiments, determining the types and locations of multiple second targets within the power distribution network operation site based on the second sensing data specifically includes: acquiring second video data from the second sensing data; extracting image features from the second video data using an edge detection algorithm (e.g., Canny) to obtain third video data; analyzing the third video data using a target detection algorithm to identify the locations of multiple second targets at the power distribution network operation site; and analyzing the third video data using an image classification algorithm to identify the types of multiple second targets at the power distribution network operation site, wherein the types of second targets include personnel targets, power equipment targets, and tool targets.

[0084] In this embodiment, after acquiring the second video data from the second perception data, the video frame is processed using an edge detection algorithm. This algorithm smooths the image using Gaussian filtering, calculates the image grayscale gradient, performs non-maximum suppression, and then extracts the image features in the video frame using a double threshold algorithm and connected component analysis to obtain the third video data.

[0085] Next, a target detection algorithm is used to analyze the third video data. This algorithm divides the image into a 7×7 grid, and each grid predicts two bounding boxes and a confidence score. The confidence score reflects the probability that the bounding box contains a target. Redundant bounding boxes are removed by a non-maximum suppression algorithm, and finally the position coordinates of personnel targets, power equipment targets and tool targets in the power distribution network operation site are obtained. The above parameter settings can also be determined according to actual needs, and are not limited here.

[0086] Simultaneously, image classification algorithms can be used to analyze the third video data. By employing residual learning and using identity mapping, the performance does not degrade as the number of network layers increases. The network can reach 152 layers. This algorithm can accurately identify the specific types of personnel, power equipment, and tools at the power distribution network operation site. The above parameter settings can also be determined according to actual needs, and are not limited here.

[0087] Based on the type matching relationship and location spatial relationship between personnel targets, power equipment targets, and tool targets, it can be determined whether personnel targets violate safety rules. If personnel targets violate safety rules, an alarm will be issued.

[0088] In this embodiment, to ensure the safety of power distribution network operations, the system checks safety rules based on the type matching and spatial relationship between personnel, power equipment, and tools. If a personnel violation of safety rules is detected, the system will immediately issue an alarm.

[0089] Specifically, a series of safety rules are defined in advance within the system. These rules clarify the permitted and prohibited relationships between personnel, electrical equipment, and tools, as well as their safe distances and area restrictions within the work site. For example, it may stipulate that personnel are prohibited from approaching high-voltage equipment within a certain range, or that certain types of tools can only be used in designated areas.

[0090] The system can acquire real-time information on the type and location of personnel, power equipment, and tools. Using spatial analysis algorithms, such as spatial querying and buffer analysis, it calculates the actual distance and relative position between personnel and power equipment / tool ​​targets. Based on the target type, it queries preset safety rules to determine the permissible relationship and safe distance between that type of target and surrounding targets. The actual spatial relationship is matched against the preset safety rules to determine if personnel have violated any rules. Special attention is paid to checking whether personnel have entered prohibited areas, approached hazardous equipment, used incorrect tools, or used tools in inappropriate areas. Once a violation is detected, the system immediately triggers an alarm mechanism. Alarm methods may include highlighting the violator's location on the monitoring interface, issuing audible and visual alarms, and sending emergency notifications to safety management personnel. Simultaneously, the system can record relevant information about the violation, such as the time, person, and location, for subsequent analysis and processing.

[0091] In this embodiment, through real-time monitoring and alarm mechanisms, the system can promptly detect and prevent violations by personnel, effectively preventing security incidents. Automated security rule checks and alarms reduce the need for human judgment and intervention, lowering the security risks caused by human negligence.

[0092] In some embodiments, determining whether a personnel target violates safety rules based on the type matching relationship and spatial relationship between personnel targets, power equipment targets, and tool targets includes: constructing a 3D model for each power equipment target at the power distribution network operation site, and setting a safety distance boundary based on the 3D model; using a target detection algorithm to determine whether the distance between the personnel target and the safety distance boundary is less than a preset distance threshold; if so, using an image classification algorithm to determine whether the identity type of the personnel target matches the power equipment target, and whether the personnel target carries a tool target of the corresponding type; if the identity type of the personnel target does not match the power equipment target, or if the personnel target does not carry a tool target of the corresponding type, then it is determined that the personnel target violates safety rules.

[0093] In this embodiment, 3D point cloud data of each power equipment target can be acquired based on the actual conditions of the power distribution network operation site. The acquired 3D point cloud data is preprocessed to remove noise points and outliers, resulting in a clear 3D point cloud model. A triangulation algorithm is used to convert the preprocessed 3D point cloud model into a 3D mesh model. Based on the type and specifications of the power equipment, corresponding 3D parametric models are obtained from a pre-established 3D model library. A 3D matching algorithm is used to match the 3D mesh model with the 3D parametric model to obtain an accurate 3D model of each power equipment target. According to the safety regulations for power operations and the 3D model of each power equipment target, corresponding safety distance boundaries are calculated and set. The 3D model and safety distance boundary information of each power equipment target are then visualized and updated in real time within the visualization system at the power distribution network operation site.

[0094] A pre-trained target detection model can be used to process monitoring images, identify personnel targets in the images, and obtain their location coordinates. Based on pre-defined safety distance boundary coordinates, the Euclidean distance between each personnel target and the boundary is calculated, and it is determined whether it is less than a preset distance threshold. If a personnel target's distance to the boundary is less than the threshold, the image region of that personnel target is captured, and an image classification model is used for identity type identification. Based on the model's classification results, it is determined whether the personnel target's identity type matches the required worker identity for the power equipment, such as electrician, maintenance worker, etc. Simultaneously, the target detection model is used to identify tools in the captured personnel image, determining whether they are carrying tools related to the power equipment, such as insulating gloves, safety helmets, etc. Combining the results of the identity type and tool carrying judgments, if the identity does not meet the requirements or the designated tools are not being carried, it is determined that the personnel have violated safety rules, an alarm message is generated, and relevant personnel are notified to handle the situation.

[0095] In any of the above embodiments, the method further includes: creating a dangerous behavior pattern feature library; acquiring behavioral data of all human targets within the power distribution network operation site and within a preset distance outside the power distribution network operation site; matching the behavioral data with the feature set of each dangerous behavior pattern in the dangerous behavior pattern feature library; if the similarity between the behavioral data and any dangerous behavior pattern in the dangerous behavior pattern feature library is higher than a second preset similarity threshold, then determining that the human target has potential dangerous behavior; determining the degree of danger score of the potential dangerous behavior of the human target based on the type of dangerous behavior pattern matched by the behavioral data and the corresponding similarity; and determining corresponding warning measures based on the degree of danger score.

[0096] In this embodiment, key features of various dangerous behaviors are extracted by analyzing historical hazardous event data, and a hazardous behavior pattern feature library is constructed, with each hazardous behavior pattern represented by a set of features. Video monitoring equipment is deployed at the power distribution network operation site to collect video image data within and outside the operation site at a preset distance in real time. Target detection algorithms are used to analyze the acquired video images, identify all human targets in the images, and extract the key point coordinate data of each human target. Based on the key point coordinate data, the posture, movement, and other behavioral features of each human target are calculated to form behavioral data. The behavioral data of each human target is matched with the feature sets in the hazardous behavior pattern feature library for similarity, and the similarity between the two is calculated using algorithms such as cosine similarity. A second preset similarity threshold is set. If the similarity between the behavioral data and any hazardous behavior pattern exceeds the threshold, it is determined that the human target currently has a corresponding potential hazardous behavior. If a potential hazardous behavior is detected, an early warning mechanism is triggered, warning the human target through voice, flashing lights, etc., and simultaneously reporting information such as screenshots of the hazardous behavior to the monitoring platform to notify safety management personnel to handle the situation promptly.

[0097] Real-time behavioral data of human targets is acquired and compared with patterns in a pre-defined dangerous behavior pattern library to calculate the similarity between each dangerous behavior pattern and the current behavioral data. Based on the similarity vector obtained from the matching, a support vector machine algorithm is used to score the degree of danger of potential dangerous behaviors, resulting in a danger score. The danger score is compared with a pre-defined threshold; if it exceeds the threshold, a potential danger is identified, and corresponding warning measures are determined based on the danger score. The prior probabilities of various dangerous behavior patterns are obtained, and combined with the similarity vector, a Bayesian network is used to calculate the posterior probabilities of various dangerous behaviors to determine the most likely type of dangerous behavior. For different types of dangerous behaviors, corresponding warning measures are pre-defined, and appropriate measures are retrieved from the warning measure library based on the type of dangerous behavior. If multiple potential dangers exist simultaneously, the danger scores of each danger are weighted and summed to determine the overall danger level, and the warning measure with the highest danger level is selected. The behavioral changes of human targets are continuously tracked and monitored, the danger level assessment is updated in real time, and the warning measures are dynamically adjusted until the danger is eliminated.

[0098] For example, to create a dangerous behavior pattern feature library, 1000 typical dangerous behavior video samples were first collected, including climbing utility poles and touching live equipment. The OpenPose algorithm was used to extract the spatiotemporal trajectory features of key human points in each video, and the trajectory features were encoded using a temporal pyramid pooling method, resulting in a 128-dimensional feature vector. Then, the K-means clustering algorithm was used to cluster the feature vectors, grouping similar dangerous behavior patterns into one class. The center point of each class represents the feature of that class of dangerous behavior patterns. The system acquires real-time monitoring videos of the power distribution network operation site and its surrounding 50-meter range, applies a target detection algorithm to detect all human targets, and uses the OpenPose algorithm to extract the key point trajectory features of each human target. The extracted trajectory features are compared with each feature in the dangerous behavior pattern feature library using cosine similarity calculation. If the similarity exceeds a threshold of 85, the human target is determined to have a potential dangerous behavior. Furthermore, a danger level score is calculated based on the type and similarity of the matched dangerous behavior pattern, ranging from 0 to 100 points, with higher scores indicating higher danger levels. When the danger level score exceeds 60 points, the system automatically sends a voice warning to the on-site workers and marks the dangerous target on the monitoring interface; when the score exceeds 80 points, the system automatically calls the on-site supervisor and sends a warning text message, while simultaneously triggering on-site warning lights and alarms. Through these methods, dangerous behaviors within and around the power distribution network work site can be identified promptly and accurately, and corresponding warning measures can be taken, effectively improving the safety of the work site. The above parameter settings can also be determined according to actual needs and are not limited here.

[0099] Furthermore, corresponding warning measures are determined based on the danger level score. Specifically, this includes: collecting warning parameter combinations under different warning intensity levels, determining the optimal warning parameter combination for each warning intensity level; determining the current warning intensity level based on the danger level score, and determining the corresponding optimal warning parameter combination based on the current warning intensity level; inputting the optimal warning parameter combination corresponding to the current warning intensity level into the warning device to warn the human target. In this embodiment, based on preset warning intensity levels, the warning parameter range corresponding to each level is obtained, forming an initial set of warning parameter combinations. For each warning intensity level, all possible warning parameter combinations are generated from the corresponding initial set of warning parameter combinations using an exhaustive method. Historical warning data is acquired, and the historical data is grouped according to the warning intensity level and warning parameter combinations to obtain warning effect evaluation data under different warning intensity levels and warning parameter combinations. For each warning intensity level, a multi-objective optimization algorithm is used, with the warning effect evaluation index as the optimization objective, to search and determine the optimal warning parameter combination from all warning parameter combinations under that level, as the optimal warning parameter combination for that warning intensity level. If the evaluation index of the optimal warning parameter combination at a certain warning intensity level is lower than a preset threshold, the initial warning parameter range for that level is expanded, and the process is repeated until the evaluation index of the optimal warning parameter combination at that level meets the requirements. The optimal warning parameter combination corresponding to each warning intensity level is stored in a preset database, forming a mapping relationship between warning intensity levels and optimal warning parameter combinations. During actual warning execution, the corresponding optimal warning parameter combination is retrieved from the mapping relationship data based on the warning intensity level, and this combination of parameters is applied to implement the warning, achieving the best warning effect at that warning intensity level.

[0100] Based on a hazard level score, a classification algorithm is used to determine the warning intensity level, establishing a mapping relationship between the score and the level. Multiple preset combinations of warning parameters are obtained, and a decision tree algorithm selects the optimal combination based on the warning intensity level. The selected warning parameter combination is input into the warning device, triggering the warning action through a control algorithm. A human detection algorithm is used to obtain the location coordinates of the target human body, which are then input into the warning device. Based on feedback from the warning device, an evaluation algorithm assesses the warning effect; if the expected effect is not achieved, the parameters are adjusted. Environmental factor parameters are obtained, and a fusion algorithm is used to optimize the combination of environmental factors and warning parameters, improving warning accuracy. A correlation model between warning effect and target human reaction is established; by analyzing changes in the target human's behavior, the warning method is dynamically adjusted.

[0101] For example, to determine the optimal combination of warning parameters for different warning intensity levels, the system first defines five warning intensity levels, each corresponding to a different range of danger level scores. For instance, levels 1 (0-20 points), 2 (21-40 points), 3 (41-60 points), 4 (61-80 points), and 5 (81-100 points) can be set. Then, for each warning intensity level, the system designs warning parameters including warning volume, warning frequency, warning duration, warning light color, and flashing frequency, and uses orthogonal experimental design to generate multiple different combinations of warning parameters. Next, the system invites 50 participants to perform simulated tasks under different combinations of warning parameters, recording their reaction time and subjective feelings. Through analysis of variance and a multi-objective optimization algorithm, the system determines the optimal combination of warning parameters for each warning intensity level, resulting in the shortest average reaction time and the highest subjective feelings score for the participants at that level. Once the system determines the current warning intensity level based on the danger level score, it can retrieve the corresponding warning parameter combination from the preset optimal warning parameter combination library and input it into the warning device, including sound and warning lights, to provide targeted warnings to human targets exhibiting dangerous behavior, reminding them to take timely safety precautions to avoid accidents. The above parameter settings can also be determined according to actual needs, and are not limited here.

[0102] According to another aspect of the present invention, a safety warning device for power distribution network operations is also provided. This device can execute the safety warning method for power distribution network operations described in the above embodiments. The specific implementation method and preferred application scenarios are the same as those described in the above embodiments, and will not be repeated here.

[0103] Figure 2 This is a schematic diagram of a safety warning device for power distribution network operations according to an embodiment of this application, such as... Figure 2 As shown, the device includes the following: an acquisition module 202, a first prediction module 204, a second prediction module 206, and an execution module 208.

[0104] The system includes: an acquisition module for acquiring first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site; a first prediction module for predicting the movement behavior of a first target within the first preset area based on the first sensing data, and obtaining a first prediction result, wherein the first prediction result indicates whether the first target will enter the power distribution network operation site; a second prediction module for predicting the power operation behavior of a personnel target among multiple second targets within the second preset area based on the second sensing data, and obtaining a second prediction result, wherein the second prediction result indicates whether the personnel target can safely perform power operations, wherein the multiple second targets include at least a personnel target, power equipment within the second preset area, and a tool target used by the personnel target, wherein the tool target indicates the power tool used by the personnel target to perform power operations; and an execution module for executing a safety warning operation based on the first and second prediction results.

[0105] The first prediction module is used to determine the position, type, and motion state of the first target based on the first perception data; and to determine the first prediction result based on the position, type, and motion state.

[0106] The first prediction module is used to determine a first risk prediction value based on location, type and motion status; determine the on-site risk level value of the power distribution network operation site based on the type information of the power operation; fuse the first risk prediction value and the on-site risk level value to obtain a target risk value; and determine the first prediction result based on the comparison result between the target risk value and the preset risk threshold.

[0107] The second prediction module is used to determine the type matching relationship and location spatial relationship between multiple second targets based on the second perception data; match the type matching relationship and location spatial relationship with preset security rules to obtain the matching result; and determine the second prediction result based on the matching result.

[0108] The type matching relationship includes at least the target type matching relationship between the type of personnel target and the type of tool target, and the location spatial relationship includes at least the target location spatial relationship between the location of personnel target and the location of power equipment; the second prediction module is used to acquire the first type of personnel target, the first location of personnel target, the second type of tool target and the second location of power equipment in the second perception data; determine the target type matching relationship based on the first type and the second type; determine the target location spatial relationship based on the first location and the second location; and determine the type matching relationship and the location spatial relationship based on the target type matching relationship and the target location spatial relationship.

[0109] The second prediction module is used to construct a dangerous behavior pattern feature library, which includes at least feature data on mismatch between personnel and the electrical tools they use, and feature data on personnel not maintaining a safe distance from the electrical equipment they operate; the second prediction result is determined by matching the matching results with the dangerous behavior pattern feature library.

[0110] Embodiments of this application also provide an electronic device, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of the present invention during runtime.

[0111] The aforementioned memory can refer to devices inside a computer used to store data and programs, including RAM, hard disks, etc. RAM can be used to temporarily store running programs and data, while hard disks can be used to store programs and data long-term. Memory enables the computer to read and write data and execute programs. The aforementioned processor is responsible for executing instructions in computer programs and performing data processing. It can also be responsible for controlling and executing various operations, including arithmetic operations, logical operations, and data transmission.

[0112] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0113] The aforementioned computer storage media can refer to the media used in computer memory to store certain discontinuous physical quantities. Computer storage media mainly include semiconductors, magnetic cores, magnetic drums, magnetic tapes, laser discs, etc. Computer-readable storage media include stored programs, which can be a set of instructions that a computer can recognize and execute, running on an electronic computer to meet certain information needs.

[0114] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0115] The aforementioned computer program products can refer to software programs that have been written, tested, and released, and can run on computers or other devices. Computer program products can include application programs, operating systems, utility software, etc., used to achieve specific functions or solve specific problems.

[0116] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of the present invention.

[0117] The aforementioned non-volatile computer-readable storage medium can refer to a medium for storing data. Non-volatile computer-readable storage media can retain data without loss when power is off and can be used to store long-term data, such as operating systems, applications, and user files. Non-volatile storage media can include hard disk drives, solid-state drives, optical disks, and flash memory storage devices, etc.

[0118] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of the present invention.

[0119] The aforementioned computer program can refer to a set of instructions used to tell the computer to perform specific tasks or operations. Computer programs can be written by programmers using specific programming languages ​​and can include algorithms, data structures, logic, and control flow. Computer programs can be used for a variety of purposes, including application software, operating systems, etc.

[0120] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0121] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0122] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0123] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0124] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0125] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A safety warning method for power distribution network operations, characterized in that, include: Acquire first sensing data of a first preset area outside the power distribution network operation site, and second sensing data of a second preset area inside the power distribution network operation site; Based on the first sensing data, the motion behavior of the first target within the first preset area is predicted to obtain a first prediction result, wherein the first prediction result is used to indicate whether the first target will enter the power distribution network operation site; Based on the second perception data, the power operation behavior of personnel targets among multiple second targets in the second preset area is predicted to obtain a second prediction result. The second prediction result is used to indicate whether the personnel targets can safely perform power operations. The multiple second targets include at least the personnel targets, power equipment in the second preset area, and tool targets used by the personnel targets. The tool targets are used to indicate the power tools used by the personnel targets to perform the power operations. Perform a safety warning operation based on the first prediction result and the second prediction result; The method of predicting the power operation behavior of personnel targets among multiple second targets in the second preset area based on the second sensing data to obtain a second prediction result includes: determining the type matching relationship and location spatial relationship among the multiple second targets based on the second sensing data; matching the type matching relationship and the location spatial relationship with preset safety rules to obtain a matching result; and determining the second prediction result based on the matching result. The type matching relationship includes at least the target type matching relationship between the type of the personnel target and the type of the tool target, and the location spatial relationship includes at least the target location spatial relationship between the location of the personnel target and the location of the power equipment. Determining the type matching relationship and location spatial relationship among the plurality of second targets based on the second sensing data includes: acquiring the first type of the personnel target, the first location of the personnel target, the second type of the tool target, and the second location of the power equipment from the second sensing data; determining the target type matching relationship based on the first type and the second type; determining the target location spatial relationship based on the first location and the second location; and determining the type matching relationship and the location spatial relationship based on the target type matching relationship and the target location spatial relationship.

2. The safety warning method for power distribution network operations according to claim 1, characterized in that, Based on the first sensing data, the motion behavior of the first target within the first preset area is predicted to obtain a first prediction result, including: The location, type, and motion state of the first target are determined based on the first sensing data; The first prediction result is determined based on the location, the type, and the motion state.

3. The safety warning method for power distribution network operations according to claim 2, characterized in that, Determining the first prediction result based on the location, the type, and the motion state includes: Based on the location, the type, and the motion state, a first risk prediction value is determined; Based on the type information of the power operation, the on-site risk level value of the power distribution network operation site is determined; The first risk prediction value and the on-site risk level value are fused to obtain the target risk value; The first prediction result is determined based on the comparison between the target risk value and the preset risk threshold.

4. The safety warning method for power distribution network operations according to claim 1, characterized in that, Determining the second prediction result based on the matching result includes: Construct a feature library of dangerous behavior patterns, wherein the feature library of dangerous behavior patterns includes at least feature data of mismatch between personnel and the electrical tools they use, and feature data of personnel not maintaining a safe distance from the electrical equipment they operate; The second prediction result is determined by matching the matching result with the dangerous behavior pattern feature library.

5. A safety warning device for power distribution network operations, characterized in that, include: The acquisition module is used to acquire first sensing data of a first preset area outside the power distribution network operation site and second sensing data of a second preset area inside the power distribution network operation site. The first prediction module is used to predict the motion behavior of the first target in the first preset area based on the first sensing data, and obtain a first prediction result, wherein the first prediction result is used to indicate whether the first target will enter the power distribution network operation site. The second prediction module is used to predict the power operation behavior of personnel targets among multiple second targets in the second preset area based on the second perception data, and obtain a second prediction result. The second prediction result is used to indicate whether the personnel targets can safely perform power operations. The multiple second targets include at least the personnel targets, power equipment in the second preset area, and tool targets used by the personnel targets. The tool targets are used to indicate the power tools used by the personnel targets to perform the power operations. The execution module is used to perform a security warning operation based on the first prediction result and the second prediction result; The second prediction module is further configured to determine the type matching relationship and location spatial relationship among the plurality of second targets based on the second perception data; match the type matching relationship and the location spatial relationship with preset security rules to obtain a matching result; and determine the second prediction result based on the matching result. The type matching relationship includes at least the target type matching relationship between the type of the personnel target and the type of the tool target, and the location spatial relationship includes at least the target location spatial relationship between the location of the personnel target and the location of the power equipment; the second prediction module is further configured to acquire the first type of the personnel target, the first location of the personnel target, the second type of the tool target, and the second location of the power equipment in the second sensing data; determine the target type matching relationship based on the first type and the second type; determine the target location spatial relationship based on the first location and the second location; and determine the type matching relationship and the location spatial relationship based on the target type matching relationship and the target location spatial relationship.

6. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program executes the safety warning method for power distribution network operations as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device where the storage medium is located to perform the safety warning method for power distribution network operations as described in any one of claims 1 to 4.

8. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the safety warning method for power distribution network operations according to any one of claims 1 to 4.

Images