Safety risk prediction method and system for power field operations based on artificial intelligence

Abstract

The present invention discloses an artificial intelligence-based safety risk prediction method and system for power field operations, which relates to the field of artificial intelligence technology and is used to improve the problem of secondary damage to substation equipment caused by maintenance personnel and static electricity accumulation when repairing a substation system. The method comprises performing a full-body scan on the maintenance personnel to determine the insulation degree of the maintenance personnel, obtaining various joints of the maintenance personnel's body and determining the humidity of each joint of the maintenance personnel according to image recognition technology, using a hierarchical analysis method to evaluate the personal risk of the maintenance personnel and obtain the evaluation results; performing a power-on experiment on various equipment of the substation system to measure the leakage of various equipment of the substation system, performing moisture classification on each equipment according to the leakage of each equipment, setting alarm rules based on the moisture classification of each equipment of the substation system and the personal risk evaluation results of the maintenance personnel, detecting the distance between each joint of the maintenance personnel and each equipment of the substation system, and performing an instantaneous alarm according to the alarm rules.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and more specifically, to an artificial intelligence-based method and system for predicting safety risks in power field operations. Background Art

[0002] Artificial intelligence technology is a branch of computer technology that aims to automatically run and implement operating instructions. Applying artificial intelligence in the prediction of safety risks in power field operations can improve the safety of maintenance personnel and maintenance efficiency.

[0003] The prior art has the following deficiencies:

[0004] Substation maintenance is usually carried out without power. The long-term operation of the substation system will cause a large amount of static electricity to accumulate. In order to improve maintenance efficiency, some maintenance personnel arbitrarily repair the substation system without taking adequate protective measures. The safety risks of maintenance personnel are greatly increased, and it is easy to damage the equipment in the substation system. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an artificial intelligence-based method and system for predicting safety risks in power field operations, which sets alarm rules by analyzing the electric shock risks of maintenance personnel and the leakage conditions of different equipment caused by moisture, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The safety risk prediction method for power field operations based on artificial intelligence includes the following steps:

[0008] Step S1: Perform a full body scan on the maintenance personnel to determine the full body insulation level of the maintenance personnel, divide the full body of the maintenance personnel into regions, obtain the humidity readings of the joint points of each detection area of ​​the maintenance personnel and mark them separately;

[0009] Step S2: Filter out the joint points marked as wet nodes in each detection area and count their number, comprehensively evaluate the personal risk of the maintenance personnel based on the number of wet nodes in each detection area and the insulation degree of the maintenance personnel, and obtain the evaluation results;

[0010] Step S3: conducting a power-on test on each device of the substation system, setting the experimental conditions according to the humidity readings of each joint point, measuring the leakage of each device of the substation system under the experimental conditions, and classifying the moisture of each device according to the leakage of each device;

[0011] Step S4: setting alarm rules based on the moisture classification of each device in the comprehensive substation system and the personal risk assessment results of the maintenance personnel, detecting the distance between each joint point of the maintenance personnel and each device in the substation system, and giving an instant alarm according to the alarm rules.

[0012] In a preferred embodiment, in step S1, a full body scan is performed on the maintenance personnel before the maintenance personnel enters the substation maintenance room to determine the full body insulation degree of the maintenance personnel, and the main active parts of the maintenance personnel's body are used as detection areas for regional division, which are respectively divided into a head area, a chest area, an arm area, and a leg area;

[0013] Extract clothing and equipment from the part images of each detection area, compare them with the insulating materials in the insulating material database, determine whether the clothing and equipment in the part images of each detection area are insulating, and determine the insulating area of ​​each detection area based on the part images of each detection area after the comparison and matching;

[0014] The insulation material database contains the models and information of all known insulation equipment, which is used to determine whether the maintenance personnel's clothing is insulating.

[0015] In a preferred embodiment, in step S2, when the humidity reading of each joint point in the detection area exceeds a preset humidity threshold, the corresponding joint point is marked as a wet node, otherwise it is marked as a dry node;

[0016] The normalized wet node data and insulation area of ​​each detection area are summed up to obtain the calculated result as the detection area risk coefficient; the risk weight of each detection area is determined using the hierarchical analysis method based on the detection area risk coefficient.

[0017] In a preferred embodiment, in step S2, the risk weight of each detection area is set using the analytic hierarchy process, and the specific steps are as follows:

[0018] Constructing a risk matrix: Taking the risk coefficient of the detection area as input, constructing a judgment matrix to compare the risk weights of each detection area;

[0019] Calculate risk vector: Calculate the risk vector of each detection area through the risk coefficient of each detection area;

[0020] Calculate the risk weight of each detection area: Calculate the risk weight of each detection area through the risk vector of each detection area.

[0021] In a preferred embodiment, in step S2, after obtaining the risk weight of each detection area, the humidity readings of each wet node in each detection area are calculated and the average value is taken as the humidity of the detection area. The humidity of the detection area and the risk weight of each detection area are weighted and summed to obtain the personal risk assessment coefficient of the maintenance personnel. The maintenance personnel are classified into low-risk maintenance personnel, medium-risk maintenance personnel and high-risk maintenance personnel by comparing the personal risk assessment coefficient of the maintenance personnel with a preset double-layer risk threshold.

[0022] In a preferred embodiment, in step S3, after calculating the humidity of the detection area, the humidity of each detection area is arranged from small to large, each humidity reading is used as a test condition, and multiple humidity environments are set to perform separate power-on tests on each device in the substation system. The specific steps are as follows:

[0023] Connect the leakage detector to the equipment in the substation system, detect the leakage of equipment in the substation system by adjusting the humidity in different areas and record the measurement results, classify each device according to the leakage situation, and determine the moisture level of each device in the substation system.

[0024] In a preferred embodiment, in step S4, after the maintenance personnel enter the substation maintenance room, the distances between each wet node marked on the maintenance personnel and each device in the substation system are measured in real time.

[0025] In a preferred embodiment, in step S4, the alarm rules are set based on the moisture classification of each device in the integrated substation system and the personal risk assessment results of the maintenance personnel. The specific steps are as follows:

[0026] Set the initial distance threshold: Set the initial distance threshold according to the distance between each wet node marked on the maintenance personnel and the substation system;

[0027] Parameter assignment: Assign parameters to maintenance personnel of different categories and moisture levels of each device;

[0028] Setting distance threshold adjustment rules: When calculating the distance thresholds between maintenance personnel of different categories and damp-affected equipment of different levels, the parameters assigned to maintenance personnel of different categories and the parameters assigned to damp-affected equipment of different levels are multiplied as the distance threshold adjustment ratio, and the product of the initial distance threshold and the distance threshold adjustment ratio is used as the adjusted distance threshold for subsequent judgment;

[0029] Set alarm rules: When the distance between the maintenance personnel's wet node and the corresponding equipment in the substation system is lower than the adjusted distance threshold, an alarm will sound, otherwise, no alarm will be issued.

[0030] An artificial intelligence-based power field operation safety risk prediction system, used to implement the above-mentioned artificial intelligence-based power field operation safety risk prediction method, including a data acquisition module, a regional division module, a risk assessment module and an alarm response module;

[0031] The data acquisition module is used to obtain the whole body insulation degree of maintenance personnel, the humidity of joints, and the leakage status of each device in the substation system after the power-on test, and transmit the acquired data to subsequent modules for system growth and operation;

[0032] The area division module is used to divide all parts of the maintenance personnel and obtain multiple detection areas, which facilitates the data collection module to collect data according to the detection areas;

[0033] The risk assessment module is used to receive the whole body insulation degree and humidity of joints of maintenance personnel to conduct risk assessment on maintenance personnel, and also classify the moisture degree of each equipment in the substation system;

[0034] The alarm response module is used to integrate the risk assessment results of maintenance personnel and the moisture level of each device in the substation system to specify alarm rules and detect the behavior of maintenance personnel in real time to determine whether to issue an alarm.

[0035] Technical effects and advantages of the artificial intelligence-based power field operation safety risk prediction method and system of the present invention:

[0036] The present invention determines the whole body insulation degree of the maintenance personnel by performing a full body scan on the maintenance personnel, obtains the joints of the maintenance personnel's body according to the image recognition technology and determines the humidity of each joint of the maintenance personnel, evaluates the personal risk of the maintenance personnel and obtains the evaluation result by comprehensively analyzing the whole body insulation degree of the maintenance personnel and the humidity readings of each joint. The present invention firstly performs risk analysis on the maintenance personnel to provide a data basis for the subsequent formulation of alarm rules, performs power-on experiments on each device of the substation system, sets experimental conditions according to the humidity readings of each joint, measures the leakage of each device of the substation system under the experimental conditions, classifies each device according to the leakage of each device, sets alarm rules based on the moisture classification of each device of the substation system and the personal risk evaluation results of the maintenance personnel, detects the distance between each joint of the maintenance personnel and each device of the substation system, and issues an instant alarm according to the alarm rules, thereby improving the safety of the maintenance personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the safety risk prediction method for power field operations based on artificial intelligence of the present invention,

[0038] Figure 2 This is a flow chart of the safety risk prediction system for power field operations based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] The present invention determines the whole body insulation degree of the maintenance personnel by performing a full body scan on the maintenance personnel, obtains the joints of the maintenance personnel's body and determines the humidity of each joint of the maintenance personnel according to the image recognition technology, evaluates the personal risk of the maintenance personnel based on the whole body insulation degree of the maintenance personnel and the humidity readings of each joint and obtains the evaluation result; performs a power-on experiment on each device of the substation system, sets the experimental conditions based on the humidity readings of each joint, measures the leakage of each device of the substation system under the experimental conditions, classifies the moisture of each device according to the leakage of each device, sets the alarm rules based on the moisture classification of each device of the substation system and the personal risk evaluation results of the maintenance personnel, detects the distance between each joint of the maintenance personnel and each device of the substation system, and issues an instant alarm based on the alarm rules, thereby improving the safety of the maintenance personnel.

[0041] Example 1: A method for predicting safety risks of power field operations based on artificial intelligence, such as Figure 1 As shown, the following steps are included:

[0042] Step S1: Perform a full body scan on the maintenance personnel to determine the full body insulation level of the maintenance personnel, divide the full body of the maintenance personnel into regions, obtain the humidity readings of the joint points of each detection area of ​​the maintenance personnel and mark them separately;

[0043] Step S2: Filter out the joint points marked as wet nodes in each detection area and count their number, comprehensively evaluate the personal risk of the maintenance personnel based on the number of wet nodes in each detection area and the insulation degree of the maintenance personnel, and obtain the evaluation results;

[0044] Step S3: conducting a power-on test on each device of the substation system, setting the experimental conditions according to the humidity readings of each joint point, measuring the leakage of each device of the substation system under the experimental conditions, and classifying the moisture of each device according to the leakage of each device;

[0045] Step S4: setting alarm rules based on the moisture classification of each device in the comprehensive substation system and the personal risk assessment results of the maintenance personnel, detecting the distance between each joint point of the maintenance personnel and each device in the substation system, and giving an instant alarm according to the alarm rules.

[0046] The specific implementation is as follows:

[0047] In step S1, a full body scan is performed on the maintenance personnel before they enter the substation maintenance room to determine the insulation level of the maintenance personnel. The main active parts of the maintenance personnel's body are used as detection areas for area division, and the maintenance personnel's body is divided into a head area, a chest area, an arm area, and a leg area, wherein the arm area includes the left and right arms and hands, and the leg area includes the left and right legs and feet.

[0048] The place for full-body scanning of maintenance personnel can be selected at the entrance of the substation maintenance room, so as to ensure that the maintenance personnel pass the full-body scan before entering the substation maintenance room. When performing full-body scans on maintenance personnel, the part images of each detection area of ​​the maintenance personnel can be determined by a three-dimensional scanner, and then the clothing and equipment are extracted from the part images of each detection area, and then proofread with the insulating materials in the insulating material database to determine whether the clothing and equipment in the part images of each detection area are insulated. After proofreading and matching, the insulation area of ​​each detection area is determined according to the part images of each detection area.

[0049] Since maintenance personnel may accidentally touch other substation equipment when repairing designated equipment in the substation system, joint points of each detection area of ​​the maintenance personnel are marked during the full-body scan, such as hand joints, foot joints, etc. The humidity readings of each marked joint point of the maintenance personnel are obtained through the humidity sensor.

[0050] By marking the joint points of the body parts in each detection area, the accuracy of the risk assessment of the maintenance personnel's status is improved, and it also facilitates subsequent behavioral detection.

[0051] It should be noted that a 3D scanner is a device that can capture the shape and features of an object's surface, and generate a 3D model or image through scanning for analysis. A humidity sensor is a device used to detect ambient humidity. When the maintenance personnel's body parts are wet, the moisture accumulates and exceeds the normal humidity concentration in the air, causing abnormal detection. The insulation material database contains the models and information of all known insulation equipment, which can be used to determine whether the maintenance personnel's clothing is insulating.

[0052] In step S2, the humidity readings of the humidity sensor on the wet parts of the maintenance personnel's body are greater than the values ​​displayed on the non-wet parts. When the humidity readings of each joint point in the detection area exceed the preset humidity threshold, the corresponding joint point is marked as a wet node, otherwise it is marked as a dry node. The wetter the joint point, the greater the harm of touching the accumulated static electricity.

[0053] The wet nodes in each detection area of ​​the maintenance personnel are counted, and the number of wet nodes in each detection area is determined. The risk assessment of various parts of the maintenance personnel's body is carried out using the hierarchical analysis method based on the insulation area of ​​each detection area and the number of wet nodes in each detection area. The number of wet nodes in each detection area and the insulation area of ​​each detection area are normalized respectively, and the calculation result obtained by subtracting the insulation area of ​​each detection area from the number of wet nodes is used as the risk coefficient of the detection area.

[0054] When normalizing the number of wet nodes in each detection area and the insulation area of ​​each detection area, the Max-Min normalization method can be used for normalization, that is, the maximum and minimum values ​​of the number of wet nodes or the insulation area in each detection area are selected, and normalized using the Max-Min normalization formula, which is: ,in To detect the wet node data or insulation area data of the area, It is the normalized result of the wet node data of the detection area or the corresponding data of the insulation area. is the minimum number of wet nodes or insulation area in each detection area, It is the maximum value of the number of wet nodes or insulation area in each detection area.

[0055] Sort the risk coefficients of the detection areas from small to large, use the analytic hierarchy process to set the weights, and determine the risk weights of each detection area. The specific steps are as follows:

[0056] Constructing a risk matrix: The risk weights of each detection area are compared by constructing a judgment matrix, using a 7-point scale from 1 to 7, where 1 means the risk is the same, 3 means the risk is slightly higher, 5 means the risk is moderately higher, and 7 means the risk is highly higher, as shown in Table 1:

[0057] Table 1

[0058]

[0059] Calculate the risk vector: In Table 1, the risk of each detection area in each column corresponds to the regional risk size of each row and is scored according to a 7-point scale. For example, in the second row and second column, the risk of detection area A is 1 / 3 of that of detection area B, that is, the risk of detection area B is higher than that of detection area A. Calculate the risk vector of each detection area. The risk vector is the weighted average of each column of the risk matrix. Before calculating the risk vector of the detection area, normalize each column of the matrix so that the sum of the normalized results of each column is 1. Then use the weighted average as the element value of the risk vector of the corresponding detection area. The matrix after normalizing the above risk matrix is ​​shown in Table 2:

[0060] Table 2

[0061]

[0062] Calculate the risk weight for each detection area:

[0063] The risk weight of detection area A = (0.0625 + 0.0358 + 0.0441 + 0.0852) / 4 = 0.0569;

[0064] The risk weight of detection area B = (0.1875 + 0.1070 + 0.0735 + 0.1193) / 4 = 0.121825;

[0065] The risk weight of detection area C = (0.3125 + 0.3220 + 0.2206 + 0.1987) / 4 = 0.26345;

[0066] The risk weight of detection area D = (0.4375+0.5352+0.6618+0.5968) / 4=0.557825.

[0067] Among them, the A, B, C, and D detection areas correspond to the four detection areas of the head area, chest area, arm area, and leg area in this example. Each detection area is sorted from large to small according to the risk coefficient of each detection area, and then corresponds to the A, B, C, and D detection areas respectively in the order of risk coefficient from large to small. Since the risk coefficient of the detection area is obtained by the normalized difference between the insulation area and the number of wet nodes of the detection area, the larger the risk coefficient of the detection area, the larger the insulation area of ​​the detection area or the smaller the number of wet nodes, the smaller the risk weight of the corresponding area.

[0068] After obtaining the risk weight of each detection area, the humidity readings of each wet node in each detection area are calculated and the average is taken as the humidity of the detection area. The humidity of the detection area and the risk weight of each detection area are weighted and summed to obtain the personal risk assessment coefficient of the maintenance personnel. The maintenance personnel are classified into low-risk maintenance personnel, medium-risk maintenance personnel and high-risk maintenance personnel by comparing the personal risk assessment coefficient of the maintenance personnel with the preset double-layer risk threshold.

[0069] By conducting risk assessment on each part of the maintenance personnel, the overall risk assessment of the maintenance personnel is improved, the accuracy of the assessment results is improved, the maintenance personnel are classified, and a data basis is provided for the subsequent alarm setting.

[0070] It should be noted that the analytic hierarchy process is a method for multi-criteria decision analysis. It can determine the influence weights of multiple related areas by integrating data from multiple related areas. It can be used to evaluate and calculate the risk weight of each detection area in this example. The preset double-layer risk threshold is two risk thresholds. When the personal risk assessment coefficient of the maintenance personnel is lower than the two thresholds, the maintenance personnel is classified as low-risk maintenance personnel. When the personal risk assessment coefficient of the maintenance personnel is between the two risk thresholds, the maintenance personnel is classified as medium-risk maintenance personnel. When the personal risk assessment coefficient of the maintenance personnel exceeds the two risk thresholds, the maintenance personnel is classified as high-risk maintenance personnel. The double-layer risk threshold is set by professionals in this field and will not be described here.

[0071] In step S3, the models of each device in the substation system are determined, and the samples of the same model are subjected to power-on experiments. After the humidity of the detection area is calculated, the humidity of each detection area is arranged from small to large, and each humidity reading is used as a test condition. Multiple humidity environments are set to conduct separate power-on tests on each device in the substation system. The specific steps are as follows:

[0072] Connect the leakage detector to the equipment in the substation system, detect the leakage of the equipment in the substation system by adjusting the humidity in different areas and record the measurement results. Arrange the calculated humidity in the detection area from small to large as the humidity condition for measuring the leakage of the equipment, classify each device according to the leakage situation, and determine the moisture level of each device in the substation system. For example, the humidity readings of the four detection areas calculated in this example are 20%, 30%, 40% and 50%, then measure the substation system with regional humidity environments of 20%, 30%, 40% and 50% respectively. According to the leakage situation of the equipment, the equipment detected with leakage in the regional humidity environment of 20% is marked as the first-level moisture-affected equipment, the equipment detected with leakage in the regional humidity environment of 30% is marked as the second-level moisture-affected equipment, the equipment detected with leakage in the regional humidity environment of 40% is marked as the third-level moisture-affected equipment, and the equipment detected with leakage in the regional humidity environment of 50% is marked as the fourth-level moisture-affected equipment. Among them, the moisture levels of the equipment are first-level moisture-affected equipment > second-level moisture-affected equipment > third-level moisture-affected equipment > fourth-level moisture-affected equipment. It should be explained that the higher the moisture level of the equipment, the easier it is for the equipment to discharge when it touches water.

[0073] It should be noted that a leakage detector is an instrument used to measure leakage current in electrical equipment or circuits. Leakage current refers to the situation where current accidentally flows into the equipment casing or grounding system under normal operating conditions, which can easily cause harm to people.

[0074] In step S4, when the maintenance personnel enter the substation maintenance room, the distance between each wet node marked on the maintenance personnel and each device in the substation system is measured in real time, and the alarm rules are set based on the moisture classification of each device in the substation system and the personal risk assessment results of the maintenance personnel. The specific steps are as follows:

[0075] Set the initial distance threshold: Set the initial distance threshold according to the distance between each wet node marked on the maintenance personnel and the substation system. You can set it yourself. For example, the initial distance threshold is set when the distance between each wet node marked on the maintenance personnel and the substation system is 10 cm.

[0076] Parameter assignment: set low-risk maintenance personnel to 1, medium-risk maintenance personnel to 2, high-risk maintenance personnel to 3, set level 4 damp equipment to 1 / 4, level 3 damp equipment to 1 / 3, level 2 damp equipment to 1 / 2, and level 1 damp equipment to 1;

[0077] Setting distance threshold adjustment rules: When calculating the distance thresholds between maintenance personnel of different categories and damp-affected equipment of different levels, the parameters assigned to maintenance personnel of different categories and the parameters assigned to damp-affected equipment of different levels are multiplied as the distance threshold adjustment ratio, and the product of the initial distance threshold and the distance threshold adjustment ratio is used as the adjusted distance threshold for subsequent judgment;

[0078] Set alarm rules: When the distance between the maintenance personnel's wet node and the corresponding equipment in the substation system is lower than the adjusted distance threshold, an alarm will sound, otherwise, no alarm will be issued.

[0079] It should be noted that according to the above alarm rules, the higher the risk of the maintenance personnel and the higher the equipment level, the larger the adjusted distance threshold will be, thereby reminding the maintenance personnel in advance, providing them with sufficient reaction time, ensuring their safety and reducing the secondary damage to the maintenance equipment.

[0080] Power field operation safety risk prediction system based on artificial intelligence, such as Figure 2 As shown, the method for realizing the above-mentioned artificial intelligence-based power field operation safety risk prediction method includes a data acquisition module, a regional division module, a risk assessment module and an alarm response module;

[0081] The data acquisition module is used to obtain the whole body insulation degree of maintenance personnel, the humidity of joints, and the leakage status of each device in the substation system after the power-on test, and transmit the acquired data to subsequent modules for system growth and operation;

[0082] The area division module is used to divide all parts of the maintenance personnel and obtain multiple detection areas, which facilitates the data collection module to collect data according to the detection areas;

[0083] The risk assessment module is used to receive the whole body insulation degree and humidity of joints of maintenance personnel to conduct risk assessment on maintenance personnel, and also classify the moisture degree of each equipment in the substation system;

[0084] The alarm response module is used to integrate the risk assessment results of maintenance personnel and the moisture level of each device in the substation system to specify alarm rules and detect the behavior of maintenance personnel in real time to determine whether to issue an alarm.

[0085] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0086] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0087] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0088] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0089] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A safety risk prediction method for power field operations based on artificial intelligence, characterized in that: The following steps are involved: Step S1: Perform a full body scan on the maintenance personnel to determine the full body insulation level of the maintenance personnel, divide the full body of the maintenance personnel into regions, obtain the humidity readings of the joint points of each detection area of ​​the maintenance personnel and mark them separately; Step S2: Filter out the joint points marked as wet nodes in each detection area and count their number, comprehensively evaluate the personal risk of the maintenance personnel based on the number of wet nodes in each detection area and the insulation degree of the maintenance personnel, and obtain the evaluation results; Step S3: conducting a power-on test on each device of the substation system, setting the experimental conditions according to the humidity readings of each joint point, measuring the leakage of each device of the substation system under the experimental conditions, and classifying the moisture of each device according to the leakage of each device; Step S4: setting alarm rules based on the moisture classification of each device in the comprehensive substation system and the personal risk assessment results of the maintenance personnel, detecting the distance between each joint point of the maintenance personnel and each device in the substation system, and giving an instant alarm according to the alarm rules.

2. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 1 is characterized in that: In step S1, a full body scan is performed on the maintenance personnel before they enter the substation maintenance room to determine the full body insulation degree of the maintenance personnel, and the main active parts of the maintenance personnel's body are divided into the head area, chest area, arm area and leg area as the detection area; Extract clothing and equipment from the part images of each detection area, compare them with the insulating materials in the insulating material database, determine whether the clothing and equipment in the part images of each detection area are insulating, and determine the insulating area of ​​each detection area based on the part images of each detection area after the comparison and matching; The insulation material database contains the models and information of all known insulation equipment, which is used to determine whether the maintenance personnel's clothing is insulating.

3. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 2 is characterized in that: In step S2, when the humidity reading of each joint point in the detection area exceeds the preset humidity threshold, the corresponding joint point is marked as a wet node, otherwise it is marked as a dry node; The wet node data and insulation area of ​​each detection area are normalized and then summed to obtain the calculated result as the risk coefficient of the detection area; The risk weight of each detection area is determined using the hierarchical analysis method based on the risk coefficient of the detection area.

4. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 3 is characterized in that: In step S2, the risk weight of each detection area is set using the analytic hierarchy process. The specific steps are as follows: Constructing a risk matrix: Taking the risk coefficient of the detection area as input, constructing a judgment matrix to compare the risk weights of each detection area; Calculate risk vector: Calculate the risk vector of each detection area through the risk coefficient of each detection area; Calculate the risk weight of each detection area: Calculate the risk weight of each detection area through the risk vector of each detection area.

5. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 4 is characterized in that: In step S2, after obtaining the risk weight of each detection area, the humidity readings of each wet node in each detection area are calculated and the average is taken as the humidity of the detection area. The humidity of the detection area and the risk weight of each detection area are weighted and summed to obtain the personal risk assessment coefficient of the maintenance personnel. The maintenance personnel are classified into low-risk maintenance personnel, medium-risk maintenance personnel and high-risk maintenance personnel by comparing the personal risk assessment coefficient of the maintenance personnel with the preset double-layer risk threshold.

6. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 1 is characterized in that: In step S3, after calculating the humidity of the detection area, the humidity of each detection area is arranged from small to large, and each humidity reading is used as a test condition. In setting multiple humidity environments, each device in the substation system is individually powered on for testing. The specific steps are as follows: Connect the leakage current detector to the equipment in the substation system, detect the leakage of equipment in the substation system by adjusting the humidity in different areas and record the measurement results. Classify each device according to the leakage situation and determine the moisture level of each device in the substation system.

7. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 3 is characterized by: In step S4, when the maintenance personnel enter the substation maintenance room, the distances between each wet node marked on the maintenance personnel and each device in the substation system are measured in real time.

8. The method for predicting safety risks of power field operations based on artificial intelligence according to claim 7 is characterized in that: In step S4, the specific steps of setting alarm rules based on the moisture classification of each device in the integrated substation system and the personal risk assessment results of maintenance personnel are as follows: Set the initial distance threshold: Set the initial distance threshold according to the distance between each wet node marked on the maintenance personnel and the substation system; Parameter assignment: Assign parameters to maintenance personnel of different categories and moisture levels of each device; Setting distance threshold adjustment rules: When calculating the distance thresholds between maintenance personnel of different categories and damp-affected equipment of different levels, the parameters assigned to maintenance personnel of different categories and the parameters assigned to damp-affected equipment of different levels are multiplied as the distance threshold adjustment ratio, and the product of the initial distance threshold and the distance threshold adjustment ratio is used as the adjusted distance threshold for subsequent judgment; Set alarm rules: When the distance between the maintenance personnel's wet node and the corresponding equipment in the substation system is lower than the adjusted distance threshold, an alarm will sound, otherwise, no alarm will be issued.

9. An artificial intelligence-based power field operation safety risk prediction system, used to implement the artificial intelligence-based power field operation safety risk prediction method according to any one of claims 1 to 8, characterized in that: It includes data acquisition module, area division module, risk assessment module and alarm response module; The data acquisition module is used to obtain the whole body insulation degree of maintenance personnel, the humidity of joint points, and the leakage status of each device in the substation system after the power-on test, and transmit the acquired data to the subsequent modules for the normal operation of the system; The area division module is used to divide all parts of the maintenance personnel and obtain multiple detection areas, which facilitates the data collection module to collect data according to the detection areas; The risk assessment module is used to receive the whole body insulation degree and humidity of joints of maintenance personnel to conduct risk assessment on maintenance personnel, and also classify the moisture degree of each equipment in the substation system; The alarm response module is used to integrate the risk assessment results of maintenance personnel and the moisture level of each device in the substation system to specify alarm rules and detect the behavior of maintenance personnel in real time to determine whether to issue an alarm.

Images