Electric power industry power transformation unplanned operation monitoring system based on artificial intelligence

Abstract

The invention discloses a power industry power transformation unplanned operation monitoring system based on artificial intelligence, which relates to the technical field of substation monitoring and comprises a monitoring equipment deployment module, a region segmentation module, a first acquisition module, a video image processing module, a security prediction module and a control warning module. The monitoring equipment deployment module is used for installing infrared cameras with night vision and panoramic functions at key positions around the transformer substation, connecting the cameras with the central monitoring platform through a wired or wireless network, and uploading data to the cloud; the region division module is used for performing region division on continuous frames in the captured power transformation equipment video; by integrating various intelligent technologies and modules, the internal and external environments of the transformer substation can be effectively monitored, potential safety hazards can be found in time, and corresponding measures are taken, so that the stable operation of the transformer substation is guaranteed, and a series of intelligent technologies and modules are integrated and applied.

Description

Technical Field

[0001] The present invention relates to the technical field of substation monitoring, and in particular to an artificial intelligence-based monitoring system for unplanned substation operations in the power industry. Background Art

[0002] With the rapid development of the power industry, the safety management and maintenance of substations have become particularly important. Traditional substation safety management mainly relies on manual patrols and regular inspections, which is not only inefficient but also difficult to detect potential safety hazards in a timely manner. In recent years, with the development of the Internet of Things, big data, cloud computing and artificial intelligence technologies, intelligent substation monitoring systems have gradually become a research hotspot. These systems can realize remote monitoring and intelligent analysis of substations, thereby improving operational efficiency and safety.

[0003] The patent publication number is CN118262274A, which provides a method and system for identifying and detecting unplanned operations of substations based on artificial intelligence, belonging to the technical field of substation operation detection. The identification and detection method includes: constructing an action data model and a safety data model for substation planned operations.

[0004] Although the existing technology provides some useful technical ideas, it still faces some of the challenges mentioned above in practical applications. Traditional substation monitoring is difficult to achieve real-time monitoring and rapid response to abnormalities, especially at night or under severe weather conditions. The monitoring efficiency is low. There are many devices in the substation. Once a failure or abnormal behavior occurs, if it is not discovered in time, it may cause serious safety accidents. When staff perform unplanned operations, if there is no appropriate supervision mechanism, illegal operations may occur, increasing the risk of accidents. The existing monitoring system may not be able to effectively process large amounts of video data, and it is difficult to extract valuable information from massive information, affecting the monitoring efficiency and accuracy. The safety assessment and early warning mechanism is relatively simple and cannot effectively respond to complex scenarios. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the deficiencies in the prior art, the present invention provides an artificial intelligence-based monitoring system for unplanned operations of power transformation in the electric power industry, which solves the problems in the above-mentioned background technology.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: an artificial intelligence-based power industry substation unplanned operation monitoring system, including a monitoring equipment deployment module, a regional segmentation module, a first acquisition module, a video image processing module, a security prediction module and a control warning module;

[0009] The monitoring equipment deployment module is used to install infrared cameras with night vision and panoramic functions at key locations around the substation, connect the cameras to the central monitoring platform via a wired or wireless network, and upload the data to the cloud;

[0010] The region segmentation module is used to divide the continuous frames in the captured substation equipment video into several groups of mesh regions, establish plane coordinates, calibrate the mesh region where the entity is located, and determine the gray value status of the pixels of two adjacent frames in real time;

[0011] The first acquisition module is used to set the video acquisition mode based on event triggering, automatically record the video when an abnormality is detected, integrate the data from the camera and sensor equipment, and generate the first data group; monitor the working status of the main power supply and backup power supply of the substation to generate the second data group, and compare the relevant pixel difference data information between adjacent video frames according to the mesh area where the substation is located to generate the third data group;

[0012] The video image processing module is used to identify faces and abnormal behaviors in the video in real time; after denoising, defogging and adjusting the brightness of the collected video images, key feature information is extracted from the video stream;

[0013] The security prediction module is used to extract key features from the first data group, the second data group, and the third data group to obtain the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz, the wearing qualification index Cdhg, the behavior standard index Xwgf, and the equipment surface temperature fluctuation difference Wdcz. Based on the extracted features, a machine learning algorithm is used to comprehensively analyze these features, and the first security assessment index Afzs and the second operation behavior safety index Xwzs are fitted and calculated;

[0014] The control warning module is used to evaluate the first security assessment index Afzs and the second operation behavior safety index Xwzs, generate corresponding assessment report results, and take different early warning measures according to the corresponding assessment report results.

[0015] Preferably, the region segmentation module includes a region identification unit and a pixel depth value unit;

[0016] The area identification unit is used to determine the fixed coordinates of the substation equipment in the continuous frames, and mark them as the first fixed coordinate q1, the second fixed coordinate q2, the third fixed coordinate q3, ..., the nth fixed coordinate qn, respectively, and compare the adjacent continuous frames to detect the pixel changes in the image;

[0017] The pixel depth value unit is used to mark corresponding pixels in two adjacent frames, determine the positions of the two pixels in the plane coordinates, and monitor relevant pixel difference data information, including the first pixel depth value Su-1 (x1, y1) and the second pixel depth value Su (x2, y2).

[0018] Preferably, the first acquisition module includes an external acquisition unit and an internal acquisition unit;

[0019] The external collection unit is used to collect and record the damage data information of the relevant equipment of the substation damaged by the external environment, including the damaged area of ​​the outer surface of the equipment line psmj, the loose pressure value of the equipment interface Ylz, the vibration frequency of the equipment Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd, as the first data group;

[0020] The internal acquisition unit is used to monitor and record relevant power status data information inside the substation in real time through a visual display cloud platform, including the real-time power SsGl of the main power line and the real-time power Bygl of the backup power line as the second data group.

[0021] Preferably, the video image processing module includes a face recognition unit and an image enhancement unit;

[0022] The face recognition unit is used to deploy a face recognition system at the entrance of the substation using face recognition technology to perform identity authentication and restrict the entry of unauthorized persons;

[0023] The image enhancement unit is used to utilize wavelet denoising technology to decompose the data information in several image frames through wavelet transform, remove the noise component in the wavelet domain, and extract the key feature information in the continuous frames through image enhancement technology.

[0024] 6. Preferably, the security prediction module includes a third analysis unit, an analysis subunit, a behavior analysis unit and a comparison subunit; the third analysis unit is used to extract the equipment line outer surface damage area psmj, the equipment interface loose pressure value Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd in the first data group, and after dimensionless processing, calculate the equipment state coefficient Dyzt through the following formula:

[0025]

[0026] Among them, α, b, d, f and e represent the damaged area of ​​the equipment line outer surface psmj, the loose pressure value of the equipment interface Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCfs and the vibration intensity Zdqd, respectively. The weight value is adjusted and set by the user, and C1 represents the first normal value;

[0027] The real-time power of the main power supply line SsGl and the real-time power of the backup power supply line Bygl in the second data group are extracted, and after dimensionless processing, the equipment power operation index Yxzs is calculated by the following formula:

[0028]

[0029] In the formula, the power operation index Yxzs reflects the operating efficiency of the equipment under different power supply conditions.

[0030] Preferably, the behavior analysis unit is used to collect the wearing conditions and behaviors of the workers in the unplanned operation, and calculate and obtain the wearing qualification index Cdhg and the behavior standard index Xwgf;

[0031] The wearing qualification index Cdhg is used to evaluate the wearing of safety protection equipment by workers in unplanned operations. The wearing qualification index Cdhg is calculated by the following formula:

[0032]

[0033] In the formula, Ei represents whether the i-th protective equipment is worn correctly, 0 represents not worn, and 1 represents worn correctly; N represents the total number of protective equipment that needs to be worn;

[0034] The behavior norm index Xwgf is used to evaluate whether the behavior of workers in unplanned operations complies with operating procedures. The behavior norm index Xwgf is calculated using the following formula:

[0035]

[0036] In the formula, Bi represents whether the jth behavioral norm is complied with, 0 represents not complied with, and 1 represents compliance; M represents the total number of behavioral norms that need to be complied with.

[0037] Preferably, the analysis subunit is used to use the third data group to compare the first pixel depth value Su-1 (x1, y1) with the second pixel depth value Su (x2, y2) to obtain the pixel change difference Jdcz, and the pixel change difference Jdcz is obtained by the following formula:

[0038] Jdcz=|Su-1(x1, y1)-Su(x2, y2)|.

[0039] Preferably, the comparison subunit is used to pre-set a difference threshold W, and compare and analyze the difference threshold W with the pixel change difference Jdcz to determine whether an active area appears in the local area of ​​the current substation;

[0040] If the pixel change difference Jdcz is greater than or equal to the difference threshold W, it means that an active area has appeared locally in the current substation. At this time, the cloud platform will receive an intermittent red flashing warning notification;

[0041] If the pixel change difference Jdcz is less than the difference threshold W, it means that there is no active area in the current substation, and no additional measures will be taken.

[0042] Preferably, the security prediction module further includes a third correlation unit, which performs dimensionless processing on the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz, and correlates the wearing qualification index Cdhg and the behavior specification index Xwgf to obtain the first security assessment index Afzs and the second operation behavior safety index Xwzs:

[0043] The first security assessment index Afzs is obtained by the following formula:

[0044]

[0045] Wherein, Wdcz represents the equipment surface temperature fluctuation difference, which is measured by the temperature sensor; G1, G2, G3 and G4 are the weight values ​​of the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz, respectively; C2 represents the second correction normal value;

[0046] The second operation behavior safety index Xwzs is obtained by the following formula:

[0047]

[0048] Wherein, G5 and G6 are the weight values ​​of the wearing qualification index Cdhg and the behavior standard index Xwgf respectively, and C3 is the third modified normal value.

[0049] Preferably, the control warning module includes a first evaluation unit and a second evaluation unit;

[0050] The first evaluation unit is used to compare and evaluate the preset evaluation threshold E with the first security evaluation index Afzs to obtain the first level report result, including:

[0051] Among them, the evaluation threshold E includes the first evaluation threshold E1 and the second evaluation threshold E2, and the first evaluation threshold E1> the second evaluation threshold E2. The specific contents of the first-level report results are as follows:

[0052] If the first security assessment index Afzs≤the second assessment threshold E2, the cloud platform will continuously receive security warning notifications, indicating that the current substation is in a safe state. At this time, normal monitoring measures will be maintained and equipment and system maintenance will be performed regularly;

[0053] If the second assessment threshold E2 is less than the first security assessment index Afzs ≤ the first assessment threshold E1, it means that the cloud platform will receive an interval non-safe flashing warning notification, indicating that the current substation is in a non-safe state, and the monitoring frequency will be strengthened at this time;

[0054] If the first security assessment index Afzs> the first assessment threshold E1, it means that the cloud platform will receive a continuous abnormal risk flashing notification, indicating that the current substation is in an unsafe state, and an emergency plan is collected for maintenance;

[0055] The second evaluation unit is used to preset a third evaluation threshold value X, compare the second operation behavior safety index Xwzs with the third evaluation threshold value X, and obtain a second evaluation result, including:

[0056] When the second operation behavior safety index Xwzs ≤ the third assessment threshold X, it indicates that the personnel behavior risk is abnormal, and a continuous personnel risk flashing notification is sent on the cloud platform;

[0057] When the second operation behavior safety index Xwzs> the third assessment threshold X, it means that the personnel behavior risk is normal and continuous monitoring is required.

[0058] (III) Beneficial effects

[0059] The present invention provides an artificial intelligence-based monitoring system for unplanned substation operations in the power industry. It has the following beneficial effects:

[0060] (1) By integrating a variety of intelligent technologies and modules, it can effectively monitor the internal and external environment of the substation, promptly discover potential safety hazards, and take corresponding measures to ensure the stable operation of the substation. Through the integrated application of a series of intelligent technologies and modules, efficient monitoring and safety management of substation operations are achieved: by deploying infrared cameras with night vision and panoramic functions, combined with regional segmentation technology, comprehensive coverage and accurate monitoring of key areas of the substation can be achieved. This helps to promptly discover and locate any abnormal situation, ensuring immediate response and processing.

[0061] (2) The system adopts an event-triggered video acquisition mode, which can intelligently identify abnormal behaviors and faces, effectively filter irrelevant information, and reduce false alarm rates. At the same time, by performing noise reduction and defogging on video images, the accuracy and reliability of image recognition are improved, ensuring the effective extraction of key feature information, and calculating the wearing qualification index Cdhg and the behavior standard index Xwgf.

[0062] (3) By comprehensively analyzing data from multiple dimensions such as equipment operation status and staff behavior, the system can accurately calculate the first security assessment index Afzs and the second operational behavior safety index Xwzs. The behavioral analysis unit collects data on the wearing of safety protection equipment and operational behavior of the staff to calculate the wearing qualification index Cdhg and the behavioral standard index Xwgf; the wearing qualification index Cdhg is based on whether the staff wears the prescribed protective equipment correctly, and the behavioral standard index Xwgf is based on whether the staff complies with the operating procedures. Both indexes use binary marks 0 or 1 to indicate whether they meet the requirements, and the final index value is calculated through a simple mathematical formula. The higher the value, the better the safety protection and behavioral standards.

[0063] (4) With the help of multi-level evaluation thresholds and intelligent control warning modules, the system can take corresponding early warning and response measures according to different levels of safety status. This not only improves the intelligence level of the system, but also greatly reduces the possibility of accidents, ensuring the safety of personnel and the normal operation of the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The present invention is a block diagram and flow chart of the power industry substation unplanned operation monitoring system based on artificial intelligence. DETAILED DESCRIPTION

[0065] 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.

[0066] Example 1

[0067] See also Figure 1 , the present invention provides an artificial intelligence-based monitoring system for unplanned operations of power transformation in the power industry, including a monitoring equipment deployment module, a region segmentation module, a first acquisition module, a video image processing module, a security prediction module and a control warning module;

[0068] The monitoring equipment deployment module is used to install infrared cameras with night vision and panoramic functions at key locations around the substation, connect the cameras to the central monitoring platform via a wired or wireless network, and upload the data to the cloud;

[0069] The region segmentation module is used to divide the continuous frames in the captured substation equipment video into several groups of mesh regions, establish plane coordinates, calibrate the mesh region where the entity is located, and determine the gray value status of the pixels of two adjacent frames in real time;

[0070] The first acquisition module sets a video acquisition mode based on event triggering, automatically records video when an abnormality is detected, integrates data from cameras and sensor devices, and generates a first data group; monitors the working status of the main power supply and backup power supply of the substation to generate a second data group, and compares the relevant pixel difference data information between adjacent video frames according to the mesh area where the substation is located to generate a third data group;

[0071] The video image processing module is used to identify faces and abnormal behaviors in the video in real time; after denoising, defogging and adjusting the brightness of the collected video images, key feature information is extracted from the video stream;

[0072] The security prediction module is used to extract key features from the first data group, the second data group, and the third data group to obtain the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz, the wearing qualification index Cdhg, the behavior standard index Xwgf, and the equipment surface temperature fluctuation difference Wdcz. Based on the extracted features, a machine learning algorithm is used to comprehensively analyze these features, and the first security assessment index Afzs and the second operation behavior safety index Xwzs are fitted and calculated;

[0073] The control warning module is used to evaluate the first security assessment index Afzs and the second operation behavior safety index Xwzs, generate corresponding assessment report results, and take different early warning measures according to the corresponding assessment report results.

[0074] In this embodiment, infrared cameras with night vision and panoramic functions are deployed at key locations of the substation to capture videos in real time and upload them to the cloud. The system uses regional segmentation technology to divide the video screen into multiple grid areas so as to accurately monitor changes in each area. When an abnormality is detected, the first acquisition module automatically starts video recording and integrates camera and sensor data to form a multi-dimensional data group. The video image processing module is responsible for processing and analyzing video content and identifying abnormal behavior. The security prediction module uses machine learning algorithms to conduct a comprehensive analysis of the collected data, the first security assessment index Afzs and the second operational behavior safety index Xwzs. Finally, the control warning module generates corresponding evaluation report results based on these indexes, and takes corresponding early warning measures to ensure the safe operation of the substation.

[0075] Example 2

[0076] For details, please refer to Figure 1,The region segmentation module includes a region identification unit and a pixel depth value unit;

[0077] The area identification unit is used to determine the fixed coordinates of the substation equipment in the continuous frames, and mark them as the first fixed coordinate q1, the second fixed coordinate q2, the third fixed coordinate q3, ..., the nth fixed coordinate qn, respectively, and compare the adjacent continuous frames to detect the pixel changes in the image;

[0078] The pixel depth value unit is used to mark corresponding pixels in two adjacent frames, determine the positions of the two pixels in the plane coordinates, and monitor relevant pixel difference data information, including the first pixel depth value Su-1 (x1, y1) and the second pixel depth value Su (x2, y2).

[0079] In this embodiment, the region segmentation module determines the fixed coordinates in the continuous frames through the region identification unit, and marks them as q1 to qn, and compares the adjacent frames to detect pixel changes. The pixel depth value unit monitors the position difference of the corresponding pixels in the adjacent frames, records the first pixel depth value Su-1 (x1, y1) and the second pixel depth value Su (x2, y2), so as to identify potential anomalies.

[0080] Example 3

[0081] For details, please refer to Figure 1 , the first acquisition module includes an external acquisition unit and an internal acquisition unit;

[0082] The external collection unit is used to collect and record the damage data information of the relevant equipment of the substation damaged by the external environment, including the damaged area of ​​the outer surface of the equipment line psmj, the loose pressure value of the equipment interface Ylz, the vibration frequency of the equipment Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd, as the first data group;

[0083] The internal acquisition unit is used to monitor and record relevant power status data information inside the substation in real time through a visual display cloud platform, including the real-time power SsGl of the main power line and the real-time power Bygl of the backup power line as the second data group.

[0084] In this embodiment, the first acquisition module collects data through an external acquisition unit and an internal acquisition unit. The external acquisition unit records the damage to the equipment by the outside world, such as the line damage area and the loose pressure of the interface, to form a first data group. The internal acquisition unit monitors the real-time power of the main power supply and the backup power supply in real time through the cloud platform, forming a second data group, so as to fully grasp the operating status of the substation.

[0085] Example 4

[0086] For details, please refer to Figure 1,The video image processing module includes a face recognition unit and an image enhancement unit;

[0087] The face recognition unit is used to deploy a face recognition system at the entrance of the substation using face recognition technology to perform identity authentication and restrict the entry of unauthorized persons;

[0088] The image enhancement unit is used to utilize wavelet denoising technology to decompose the data information in several image frames through wavelet transform, remove the noise component in the wavelet domain, and extract the key feature information in the continuous frames through image enhancement technology.

[0089] In this embodiment, the video image processing module implements functions through a face recognition unit and an image enhancement unit. The face recognition unit deploys a face recognition system at the entrance of the substation to perform identity authentication and prevent unauthorized personnel from entering. The image enhancement unit uses wavelet denoising technology to decompose image frames and remove noise through wavelet transform, thereby enhancing image quality and extracting key feature information.

[0090] Example 5

[0091] For details, please refer to Figure 1 The security prediction module includes a third analysis unit, an analysis subunit, a behavior analysis unit and a comparison subunit; the third analysis unit is used to extract the equipment line outer surface damage area psmj, the equipment interface loose pressure value Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd in the first data group, and after dimensionless processing, the equipment state coefficient Dyzt is calculated by the following formula:

[0092]

[0093] Among them, α, b, d, f and e represent the damaged area of ​​the equipment line outer surface psmj, the loose pressure value of the equipment interface Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCfs and the vibration intensity Zdqd, respectively. The weight value is adjusted and set by the user, and C1 represents the first normal value;

[0094] The real-time power of the main power supply line SsGl and the real-time power of the backup power supply line Bygl in the second data group are extracted, and after dimensionless processing, the equipment power operation index Yxzs is calculated by the following formula:

[0095]

[0096] In the formula, the power operation index Yxzs reflects the operating efficiency of the equipment under different power supply conditions.

[0097] In this embodiment, the third analysis unit of the security prediction module extracts data on the damaged area of ​​the outer surface of the equipment line, the loose pressure value of the equipment interface, the equipment vibration frequency, the electromagnetic radiation intensity and the vibration intensity from the first data group, and after dimensionless processing, calculates the equipment state coefficient Dyzt; at the same time, the real-time power data of the main power line and the backup power line are extracted from the second data group, and after dimensionless processing, the equipment power operation index Yxzs is calculated to reflect the operating efficiency of the equipment under different power conditions.

[0098] Example 6

[0099] For details, please refer to Figure 1 ,The behavior analysis unit is used to collect the wearing of safety protection equipment and behavior of the workers in unplanned operations, and calculate to obtain the wearing qualification index Cdhg and the behavior standard index Xwgf;

[0100] The wearing qualification index Cdhg is used to evaluate the wearing of safety protection equipment by workers in unplanned operations. The wearing qualification index Cdhg is calculated by the following formula:

[0101]

[0102] In the formula, Ei represents whether the i-th protective equipment is worn correctly, 0 represents not worn, and 1 represents worn correctly; N represents the total number of protective equipment that needs to be worn;

[0103] The behavior norm index Xwgf is used to evaluate whether the behavior of workers in unplanned operations complies with operating procedures. The behavior norm index Xwgf is calculated using the following formula:

[0104]

[0105] In the formula, Bi represents whether the jth behavioral norm is complied with, 0 represents not complied with, and 1 represents compliance; M represents the total number of behavioral norms that need to be complied with.

[0106] In this embodiment, the behavior analysis unit collects the wearing conditions of the safety protection equipment and the work behavior data of the staff, and calculates the wearing qualification index Cdhg and the behavior standard index Xwgf; the wearing qualification index Cdhg is evaluated based on whether the staff wears the prescribed protective equipment correctly, and the behavior standard index Xwgf is evaluated based on whether the staff complies with the operating procedures. Both indexes use binary marks 0 or 1 to indicate whether they meet the requirements, and the final index value is calculated by a simple mathematical formula. The higher the value, the better the safety protection and behavior standard.

[0107] Example 7

[0108] For details, please refer to Figure 1 The analysis subunit is used to use the third data group to compare the first pixel depth value Su-1 (x1, y1) with the second pixel depth value Su (x2, y2) to obtain the pixel change difference Jdcz, and the pixel change difference Jdcz is obtained by the following formula:

[0109] Jdcz=|Su-1(x1, y1)-Su(x2, y2)|.

[0110] In this embodiment, the analysis subunit compares the first pixel depth value Su-1(x1, y1) and the second pixel depth value Su(x2, y2) in the third data group to calculate the pixel change difference Jdcz. This difference reflects the degree of pixel change between adjacent video frames and is used to detect activities or changes in the video surveillance screen.

[0111] Example 8

[0112] For details, please refer to Figure 1 , the comparison subunit is used to pre-set the difference threshold W, and compare and analyze the difference threshold W with the pixel change difference Jdcz to determine whether an active area appears in the local area of ​​the current substation;

[0113] If the pixel change difference Jdcz is greater than or equal to the difference threshold W, it means that an active area has appeared locally in the current substation. At this time, the cloud platform will receive an intermittent red flashing warning notification;

[0114] If the pixel change difference Jdcz is less than the difference threshold W, it means that there is no active area in the current substation, and no additional measures will be taken.

[0115] In this embodiment, the working principle of the scheme is to monitor the activities in the local area of ​​the substation through a comparison subunit. Specifically, the comparison subunit will pre-set a difference threshold W and compare the calculated pixel change difference Jdcz with it. If Jdcz is greater than or equal to W, it is considered that there is activity in the local area, and the system will send an interval red flashing warning notification to the cloud platform; conversely, if Jdcz is less than W, it is considered that no activity has occurred, and the system will not trigger any warning measures. In this way, real-time monitoring can be achieved and timely response to abnormal activities can be made.

[0116] Example 9

[0117] For details, please refer to Figure 1The security prediction module also includes a third correlation unit, which processes the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz dimensionlessly, and correlates the wearing qualification index Cdhg and the behavior specification index Xwgf to obtain the first security assessment index Afzs and the second operation behavior safety index Xwzs:

[0118] The first security assessment index Afzs is obtained by the following formula:

[0119]

[0120] Wherein, Wdcz represents the equipment surface temperature fluctuation difference, which is measured by the temperature sensor; G1, G2, G3 and G4 are the weight values ​​of the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz, respectively; C2 represents the second correction normal value;

[0121] The second operation behavior safety index Xwzs is obtained by the following formula:

[0122]

[0123] Wherein, G5 and G6 are the weight values ​​of the wearing qualification index Cdhg and the behavior standard index Xwgf respectively, and C3 is the third modified normal value.

[0124] In this embodiment, the third associated unit in the security prediction module is responsible for comprehensive analysis of the equipment operation status and staff behavior data. It first performs dimensionless processing on the equipment power operation index Yxzs, the equipment status coefficient Dyzt, the pixel change difference Jdcz, and the equipment surface temperature fluctuation difference Wdcz, and then combines the weight values ​​of these indicators to calculate the first security assessment index Afzs. At the same time, the unit will also calculate the second operation behavior safety index Xwzs based on the wearing qualification index Cdhg and the behavior standard index Xwgf and their respective weight values. These two indexes comprehensively reflect the safety status of the substation and the safety of the operation behavior.

[0125] Example 10

[0126] For details, please refer to Figure 1 , the control warning module includes a first evaluation unit and a second evaluation unit;

[0127] The first evaluation unit is used to compare and evaluate the preset evaluation threshold E with the first security evaluation index Afzs to obtain the first level report result, including:

[0128] Among them, the evaluation threshold E includes the first evaluation threshold E1 and the second evaluation threshold E2, and the first evaluation threshold E1> the second evaluation threshold E2. The specific contents of the first-level report results are as follows:

[0129] If the first security assessment index Afzs≤the second assessment threshold E2, the cloud platform will continuously receive security warning notifications, indicating that the current substation is in a safe state. At this time, normal monitoring measures will be maintained and equipment and system maintenance will be performed regularly;

[0130] If the second assessment threshold E2 is less than the first security assessment index Afzs ≤ the first assessment threshold E1, it means that the cloud platform will receive an interval non-safe flashing warning notification, indicating that the current substation is in a non-safe state, and the monitoring frequency will be strengthened at this time;

[0131] If the first security assessment index Afzs> the first assessment threshold E1, it means that the cloud platform will receive a continuous abnormal risk flashing notification, indicating that the current substation is in an unsafe state, and an emergency plan is collected for maintenance;

[0132] The second evaluation unit is used to preset a third evaluation threshold value X, compare the second operation behavior safety index Xwzs with the third evaluation threshold value X, and obtain a second evaluation result, including:

[0133] When the second operation behavior safety index Xwzs ≤ the third assessment threshold X, it indicates that the personnel behavior risk is abnormal, and a continuous personnel risk flashing notification is sent on the cloud platform;

[0134] When the second operation behavior safety index Xwzs> the third assessment threshold X, it means that the personnel behavior risk is normal and continuous monitoring is required.

[0135] Please refer to Figure 1 In this embodiment, the security assessment index Afzs and the operation behavior safety index Xwzs are assessed by the first assessment unit and the second assessment unit to determine the safety status of the substation. The first assessment unit performs a graded assessment on the first security assessment index Afzs according to the preset first assessment threshold E1 and second assessment threshold E2, and adopts different levels of monitoring and maintenance measures accordingly.

[0136] When Afzs is lower than the second threshold E2, the system considers the substation to be in a safe state and maintains normal monitoring; when Afzs is between E2 and E1, monitoring will be strengthened and maintenance frequency will be increased;

[0137] When Afzs exceeds E1, emergency measures must be taken immediately. The second assessment unit determines the behavioral risk of the staff member based on the comparison between the second operation behavior safety index Xwzs and the third assessment threshold X, and sends a risk notification or continues monitoring accordingly. This hierarchical assessment and response mechanism helps to promptly discover and deal with potential safety issues.

[0138] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The power industry substation unplanned operation monitoring system based on artificial intelligence is characterized by: It includes a monitoring equipment deployment module, a region segmentation module, a first acquisition module, a video image processing module, a security prediction module and a control warning module; The monitoring equipment deployment module is used to install infrared cameras with night vision and panoramic functions at key locations around the substation, connect the cameras to the central monitoring platform via a wired or wireless network, and upload data to the cloud; The region segmentation module is used to divide the continuous frames in the captured substation video into several groups of mesh regions, establish plane coordinates, calibrate the mesh region where the entity is located, and determine the gray value state of pixels of two adjacent frames in real time; The first acquisition module is used to set a video acquisition mode based on event triggering, automatically record video when an abnormality is detected, integrate data from cameras and sensor devices, and generate a first data group; monitor the working status of the main power supply and backup power supply of the substation to generate a second data group, and compare the relevant pixel difference data information between adjacent video frames according to the mesh area where the substation is located to generate a third data group; The video image processing module is used to identify faces and abnormal behaviors in the video in real time; after performing noise reduction, defogging and brightness adjustment on the collected video images, key feature information is extracted from the video stream; The security prediction module is used to extract key features from the first data group, the second data group and the third data group to obtain the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz, the wearing qualification index Cdhg, the behavior standard index Xwgf and the equipment surface temperature fluctuation difference Wdcz. Based on the extracted features, a machine learning algorithm is used to perform a comprehensive analysis on these features, and a first security assessment index Afzs and a second operation behavior safety index Xwzs are fitted and calculated; The control warning module is used to evaluate the first security evaluation index Afzs and the second operation behavior safety index Xwzs, generate corresponding evaluation report results, and take different early warning measures according to the corresponding evaluation report results.

2. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 1 is characterized by: The region segmentation module includes a region identification unit and a pixel depth value unit; The area identification unit is used to determine the fixed coordinates of the substation equipment in the continuous frames, and mark them as the first fixed coordinate q1, the second fixed coordinate q2, the third fixed coordinate q3, ..., the nth fixed coordinate qn, respectively, and compare the adjacent continuous frames to detect the pixel changes in the image; The pixel depth value unit is used to mark corresponding pixels in two adjacent frames, determine the positions of the two pixels in the plane coordinates, and monitor relevant pixel difference data information, including a first pixel depth value Su-1 (x1, y1) and a second pixel depth value Su (x2, y2).

3. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 1 is characterized by: The first acquisition module includes an external acquisition unit and an internal acquisition unit; The external collection unit is used to collect and record the damage data information of the relevant equipment of the substation damaged by the external environment, including the damaged area of ​​the outer surface of the equipment line psmj, the loose pressure value of the equipment interface Ylz, the vibration frequency of the equipment Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd, as the first data group; The internal collection unit is used to monitor and record relevant power status data information inside the substation in real time through a visually displayed cloud platform, including the real-time power SsGl of the main power line and the real-time power Bygl of the backup power line as the second data group.

4. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 1 is characterized by: The video image processing module includes a face recognition unit and an image enhancement unit; The face recognition unit is used to deploy a face recognition system at the entrance of the substation using face recognition technology to perform identity verification and restrict entry of unauthorized persons; The image enhancement unit is used to utilize wavelet denoising technology to decompose data information in a plurality of image frames through wavelet transform, remove noise components in the wavelet domain, and extract key feature information in continuous frames through image enhancement technology.

5. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 4 is characterized by: The security prediction module includes a third analysis unit, an analysis subunit, a behavior analysis unit and a comparison subunit; the third analysis unit is used to extract the equipment line outer surface damage area psmj, the equipment interface loose pressure value Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCFs and the vibration intensity Zdqd in the first data group, and after dimensionless processing, calculate the equipment state coefficient Dyzt through the following formula: Among them, α, b, d, f and e represent the damaged area of ​​the equipment line outer surface psmj, the loose pressure value of the equipment interface Ylz, the equipment vibration frequency Gjpl, the electromagnetic radiation intensity DCfs and the vibration intensity Zdqd, respectively. The weight value is adjusted and set by the user, and C1 represents the first normal value; The real-time power of the main power supply line SsGl and the real-time power of the backup power supply line Bygl in the second data group are extracted, and after dimensionless processing, the equipment power operation index Yxzs is calculated by the following formula: In the formula, the power operation index Yxzs reflects the operating efficiency of the equipment under different power supply conditions.

6. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 5 is characterized by: The behavior analysis unit is used to collect the wearing conditions and behaviors of the workers in the unplanned operation, and calculate and obtain the wearing qualification index Cdhg and the behavior standard index Xwgf; The wearing qualification index Cdhg is used to evaluate the wearing of safety protection equipment by workers in unplanned operations. The wearing qualification index Cdhg is calculated by the following formula: In the formula, Ei represents whether the i-th protective equipment is worn correctly, 0 represents not worn, and 1 represents worn correctly; N represents the total number of protective equipment that needs to be worn; The behavior norm index Xwgf is used to evaluate whether the behavior of workers in unplanned operations complies with operating procedures. The behavior norm index Xwgf is calculated using the following formula: In the formula, Bi represents whether the jth behavioral norm is complied with, 0 represents not complied with, and 1 represents compliance; M represents the total number of behavioral norms that need to be complied with.

7. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 2 is characterized by: The analysis subunit is used to use the third data group to compare the first pixel depth value Su-1 (x1, y1) with the second pixel depth value Su (x2, y2) to obtain a pixel change difference Jdcz, wherein the pixel change difference Jdcz is obtained by the following formula: Jdcz=|Su-1(x1, y1)-Su(x2, y2)|.

8. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 7 is characterized by: The comparison subunit is used to preset a difference threshold W, and compare and analyze the difference threshold W with the pixel change difference Jdcz to determine whether an active area appears in the local area of ​​the current substation; If the pixel change difference Jdcz is greater than or equal to the difference threshold W, it indicates that an active area has appeared locally in the current substation, and the cloud platform will receive an intermittent red flashing warning notification; If the pixel change difference Jdcz is less than the difference threshold W, it means that there is no active area in the local area of ​​the current substation, and no additional measures will be taken at this time.

9. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 8 is characterized by: The security prediction module also includes a third correlation unit, which processes the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz dimensionlessly, and correlates the wearing qualification index Cdhg and the behavior specification index Xwgf to obtain the first security assessment index Afzs and the second operation behavior safety index Xwzs: The first security assessment index Afzs is obtained by the following formula: Wherein, Wdcz represents the equipment surface temperature fluctuation difference, which is measured by the temperature sensor; G1, G2, G3 and G4 are the weight values ​​of the equipment power operation index Yxzs, the equipment state coefficient Dyzt, the pixel change difference Jdcz and the equipment surface temperature fluctuation difference Wdcz, respectively; C2 represents the second correction normal value; The second operation behavior safety index Xwzs is obtained by the following formula: Wherein, G5 and G6 are the weight values ​​of the wearing qualification index Cdhg and the behavior standard index Xwgf respectively, and C3 is the third modified normal value.

10. The artificial intelligence-based power industry substation unplanned operation monitoring system according to claim 9 is characterized by: The control warning module includes a first evaluation unit and a second evaluation unit; The first evaluation unit is used to compare and evaluate the preset evaluation threshold E with the first security evaluation index Afzs to obtain the first level report result. include: The evaluation threshold E includes a first evaluation threshold E1 and a second evaluation threshold E2, and the first evaluation threshold E1> the second evaluation threshold E2. The specific contents of the first-level report results are as follows: If the first security assessment index Afzs≤the second assessment threshold E2, the cloud platform will continuously receive security warning notifications, indicating that the current substation is in a safe state. At this time, normal monitoring measures will be maintained and equipment and system maintenance will be performed regularly; If the second evaluation threshold E2 is less than the first security evaluation index Afzs ≤ the first evaluation threshold E1, it means that the cloud platform will receive an interval non-safe flashing warning notification, indicating that the current substation is in a non-safe state, and the monitoring frequency will be strengthened at this time; If the first security assessment index Afzs> the first assessment threshold E1, it means that the cloud platform will receive a continuous abnormal risk flashing notification, indicating that the current substation is in an unsafe state, and an emergency plan is collected for maintenance; The second evaluation unit is used to preset a third evaluation threshold value X, compare the second operation behavior safety index Xwzs with the third evaluation threshold value X, and obtain a second evaluation result, including: When the second operation behavior safety index Xwzs ≤ the third assessment threshold X, it indicates that the personnel behavior risk is abnormal, and a continuous personnel risk flashing notification is sent on the cloud platform; When the second operation behavior safety index Xwzs> the third assessment threshold X, it means that the personnel behavior risk is normal and continuous monitoring is required.

Images