Electric power engineering supervision and shooting method and system based on artificial intelligence

Abstract

The invention belongs to the technical field of power equipment monitoring, and particularly relates to an artificial intelligence-based power engineering supervision and monitoring method and system, and the method comprises the steps: obtaining an authorized technical improvement and overhaul project list from a power grid resource business platform in response to successful login, building the correlation between a supervision and monitoring device and a target project, and carrying out the supervision and monitoring of the target project; downloading feasibility research report structured data corresponding to the project, an associated power equipment asset code and power industry mandatory regulation text data of a specialty to which the project belongs; according to the invention, through double authentication of face living body detection and passwords in the step S1, unique binding of'human-machine-account 'is ensured, and false use of the identity is completely eradicated; and S3, Beidou differential positioning and anti-metal electronic tag scanning are creatively fused, the positioning precision is improved to a meter level / sub-meter level in a 500kV strong electromagnetic interference environment, whether a supervisor accurately arrives at an equipment operation point or not is directly checked, the recognition rate of virtual post behaviors such as card punching and regional wandering is improved, and supervision is guaranteed to perform in place from the source.

Description

A method and system for monitoring and filming power engineering based on artificial intelligence Technical Field

[0001] This invention belongs to the field of power equipment monitoring technology, specifically relating to a power engineering supervision and monitoring method and system based on artificial intelligence. Background Technology

[0002] As the State Grid Corporation of China continues to raise its requirements for the health level of power grid equipment and the reliability of power supply, the number and investment scale of power grid technical renovation and overhaul projects have been increasing year by year. However, the allocation of supporting project management personnel, on-site safety capacity, and qualified construction forces are relatively insufficient, and management shortcomings are becoming increasingly prominent. Although existing power engineering supervision technology has incorporated positioning and video recording methods, it has obvious defects in reliability, intelligence, and systemicity, making it difficult to meet the needs of lean power grid management and control: First, the positioning accuracy of on-site personnel is low, relying heavily on GPS / base station positioning via mobile apps or ordinary check-in devices, with an accuracy of hundreds of meters. This makes it impossible to verify whether the supervisors have arrived at specific substation intervals or equipment operation points, easily leading to problems such as fake check-ins and fictitious positions. Second, supervision records are fragmented and have low credibility. Unstructured data such as images and logs are easily tampered with and lack strong correlation with projects, processes, and equipment, making it difficult to trace and determine responsibility when quality disputes or safety incidents occur. Third, regulatory response is lagging. Supervision data is not integrated with the company-level project management and asset management systems, making it impossible for managers to monitor dispersed sites in real time. They rely more on post-event spot checks, making timely intervention difficult. Summary of the Invention

[0003] To address the aforementioned shortcomings of existing technologies, this invention provides an artificial intelligence-based method, system, terminal, and storage medium for power engineering supervision and monitoring.

[0004] In a first aspect, the present invention provides an artificial intelligence-based power engineering supervision and monitoring method, comprising: S1, performing dual authentication of user identity and logging into an intelligent management and control platform through a supervision and monitoring device, wherein the dual authentication includes password verification and facial liveness detection verification; S2, in response to successful login, obtaining an authorized list of technical renovation and overhaul projects from the power grid resource business platform, establishing an association between the supervision and monitoring device and the target project, and downloading the structured data of the feasibility study report corresponding to the project, the associated power equipment asset code, and the text data of the mandatory power industry regulations of the project's professional category; S3, obtaining initial geographic coordinates and equipment electronic tag information, performing intelligent location calibration through a multi-source fusion algorithm, and generating meter-level to sub-meter-level calibration positioning data; S4, activating the high-definition image recording function of the supervision and monitoring device, collecting real-time video streams from the construction site and caching them to the local storage unit; S5, triggering the edge AI inference module of the supervision and monitoring device to process step S4. The on-site video stream undergoes polarization imaging and multispectral fusion preprocessing to generate a fused video stream. OCR recognition is performed on the nameplates of power equipment in the fused video stream, and time-series analysis is conducted on the construction activities. The recognition results are then checked for consistency with the structured data from the feasibility study report downloaded in S2 and the text data of mandatory power industry regulations. A sequence of process events carrying confidence markers is output. In S6, the process event sequence output in step S5 is received. When the confidence level of key nodes in the process event sequence exceeds a set threshold, a high-resolution image is automatically captured. The captured image is bound to the calibrated positioning data generated in step S3, the identity information of the supervisor certified in step S1, the current timestamp, the associated power equipment asset code, and the verification results from step S5 to generate an immutable digital trace containing a national cryptographic SM2 digital signature. In S7, the digital trace and full-process supervision metadata are uploaded to the intelligent management and control platform through a national cryptographic SM4 encrypted channel.

[0005] Further improvements to this technical solution include step S1, which includes: S11, the supervisor enters the username on the touch interface of the monitoring device. With password After receiving the input information, the core control unit of the monitoring and surveillance device calls the SHA-256 hash algorithm to encrypt the input password and generate a password hash value. The encryption formula is: ;in, The hash value of the input password; The original password entered by the supervisor is transmitted in plaintext to the core control unit and then immediately hashed to avoid the risk of plaintext storage; S12, the core control unit synchronizes the pre-stored password hash value corresponding to the username from the intelligent management and control platform. Compared with pre-stored facial feature vectors ,Will and A consistency comparison is performed. If the two are completely identical, the password verification passes; if they are different, an audible and visual alarm is triggered, the alarm type is recorded as "password verification failed," and the information is uploaded to the intelligent management platform. The comparison formula is: ;in, This is the password verification result; This indicates that the verification has passed; Verification failed; S13, after password verification is successful, the device's front-facing camera starts face capture, continuously capturing 3 frames of the supervisor's face image. , , The face liveness detection unit performs dynamic feature analysis on the acquired images, and determines whether a person is alive by calculating the grayscale change value between adjacent frames. The determination formula is as follows: ;in, The liveness detection threshold is denoted by ; M and N are the number of rows and columns of pixels in the face image, respectively. Let (i,j) be the grayscale value of the k-th frame image at pixel (i,j); when If the liveness detection is successful, the face feature extraction unit will process the second frame image. Otherwise, it will be considered a fake medium, triggering an alarm and terminating authentication. (S14) Perform feature extraction to generate a 1024-dimensional face feature vector. Calculate the vector and the pre-stored face feature vector. The cosine similarity is calculated using the following formula: ;in, The similarity of feature vectors is defined, with values ​​ranging from [0,1]. The k-th dimension of the input face feature vector; For the pre-stored k-th dimension component of the face feature vector; when At that time, facial verification is successful; S15, after facial verification is successful, the core control unit generates a unique identity authentication identifier. The formula is: ;in, The unique serial number for the monitoring and surveillance device; This is the current login timestamp; Used for binding digital traces with the identity of supervisors in step S6.

[0006] Further improvements to this technical solution include step S2, which includes: S21, the monitoring and surveillance device sends a project list request to the power grid resource business platform via the 5G communication module, and the request data packet carries the identity authentication identifier generated in step S1. This is used to verify the project authorization permissions of the supervisors in the middle platform. The request data packet format must meet the "Power Industry Data Transmission Protocol Specification". The formula is: ;in, Request a data packet for the item list; This is the header information, which includes the protocol version number and data length; The unique serial number for the monitoring and surveillance device hardware; Send a timestamp to request; The checksum is calculated using the following formula: ;in, S22: The data packet contains the i-th byte; n is the total number of bytes in the data packet; S22: The power grid resource business platform receives the project list request data packet. Then, first pass the checksum. Verify the integrity of the data packet, then parse the authentication identifier. It also queries the associated authorized project list, filters out technical renovation and major repair projects with a status of "awaiting supervision," and generates a project list: Where m is the number of authorized projects; each project Includes project number Project Name Planned supervision cycle The middle platform will include a project list. The encrypted data is returned to the monitoring device. The encryption uses the national standard SM4 algorithm, and the formula is as follows: ;in, This is the encrypted list of projects; The session key is negotiated and generated between the monitoring and surveillance device and the power grid resource business platform through the ECC key exchange protocol; S23, the monitoring and surveillance device receives the encrypted project list. ,pass Decryption yields the project list The project list is displayed on the touchscreen for the supervisors to choose from; the supervisors then select the target project. Subsequently, the device sends a target project data request to the power grid resource service platform, carrying the target project number in the request. S24, The power grid resource business platform, based on the target project number... The system retrieves the original feasibility study report data and power equipment asset codes corresponding to the project from the PMS system, and retrieves the text data of the mandatory power industry regulations for the project's specialty from the risk control platform. It then performs structured processing on the original feasibility study report data, extracting three core data categories: "equipment model standards," "construction procedure requirements," and "safety distance thresholds," to generate structured feasibility study report data. The structured processing formula is: ;in, This is an XML structuring conversion function that converts unstructured feasibility study report text into tagged XML format. This is the original data for the feasibility study report; This is an associated tag used to bind power equipment asset codes. , The unique asset code for the k-th device; S25, the power grid resource business platform will structure the data from the feasibility study report. Power equipment asset code and mandatory regulations text data in the power industry Packaged as target project dataset The data is transmitted to the monitoring and surveillance device via a 5G encrypted channel; upon receiving the data, the device will process the target project dataset. Stored to a local encrypted partition, the encryption formula is as follows: ;in, The local hardware encryption key for the monitoring and surveillance device.

[0007] Further improvements to this technical solution include step S3, which includes: S31, the electromagnetic interference sensor of the monitoring and surveillance device collects the electromagnetic interference intensity E at the construction site in real time at a frequency of 1Hz, and simultaneously activates the Beidou differential positioning module to collect the initial geographic coordinates at a frequency of 1Hz. S32. For outliers in the initial geographic coordinates caused by strong electromagnetic interference, a Kalman filter algorithm is used for noise reduction to obtain the filtered BeiDou coordinates. S33. Activate the anti-metal electronic tag scanning module to scan the UHF RFID anti-metal tag affixed to the surface of the target power equipment and obtain the preset coordinates of the equipment built into the tag. and equipment asset codes The scanning distance must meet the requirements. t is the thickness of the equipment's metal casing; S34, calculate the average value of the filtered BeiDou coordinates. The formula is: ; then calculate and Positional deviation The formula is: S35. Based on positional deviation The weights of BeiDou positioning and tag positioning are dynamically allocated based on the electromagnetic interference intensity E to generate calibration positioning data. The weighting and standard formula are: when hour: , , ;when hour: , ; ; ; ;in, Weighting for BeiDou positioning; Weighting for anti-metal tags; when hour ,when hour ,make sure The positioning accuracy provides precise coordinates for the digital trace location stamp in step S6.

[0008] Further improvements to this technical solution include step S4, which includes: S41, the environmental sensing module of the monitoring and surveillance device collects real-time data on the light intensity at the construction site. With sound decibel value The sampling frequency is 1Hz. The recording parameters of the high-definition camera are automatically adjusted based on the collected data. The parameter adjustment formula is: When... hour, The shutter speed of the HD camera was adjusted to 1 / 50s, and the LED fill light was turned on with a fill light intensity of [insert value here]. ;when hour, The shutter speed of the HD camera was adjusted to 1 / 200s, and the LED fill light was turned off; when When microphone noise reduction is enabled, the noise reduction intensity is [not specified]. ;in, For camera sensitivity; To increase the LED fill light intensity; S42: To reduce noise intensity; Activate the HD camera to record video and generate a real-time video stream. The H.265 video coding standard is used for real-time video streams. Compression processing is performed, compression ratio Based on the calibration and positioning data generated in step S3 The associated construction scenario is dynamically adjusted using the following formula: Bitrate of the compressed video stream The calculation formula is: ;in, For video resolution; To record the frame rate; S43, divide the compressed video stream into video segments according to the time-slicing strategy. Duration of each segment The segmentation formula is: ;in, , This is the start time of recording. ; Add metadata tags to each ,in The associated device asset code downloaded in step S2; This refers to the average calibration positioning data during the recording of this segment; Record timestamps for the segments; S44, add metadata tags. The cached data is stored in a local encrypted storage unit and encrypted using the SM4 algorithm before storage. The encryption formula is as follows: ;in, Use it as a local storage key; and monitor the remaining local storage capacity. ,when At that time, historical video clips that have been uploaded to the intelligent management and control platform and have been fully verified will be automatically deleted.

[0009] Further improvements to this technical solution include step S5, which includes: S51, the edge AI inference module retrieves cached video stream segments from the local storage unit in step S4. First, process each frame of the image. Perform polarization imaging preprocessing to eliminate reflective interference from the equipment's metal casing. The processing formula is as follows: ;in, This is a single-frame image after polarization processing; The original video frame grayscale value; The polarization angle of this frame image is acquired in real time by the camera's built-in polarization sensor; S52, the multispectral fusion algorithm is invoked to process the polarized visible light image. Near-infrared images Fusion, the fusion formula is: ; ; ;in, The gray value of pixel (i,j) in the fused image; For weighting; Image grayscale contrast; , These represent the maximum and minimum grayscale values ​​of the image, respectively; all fused frames form the fused video stream. S53, For fused video streams OCR recognition is performed on the nameplate area of ​​the power equipment. First, the nameplate area is located using the YOLOv8 object detection algorithm, and then the character sequence of the nameplate is extracted. Calculation and download of structured data from the feasibility study report in step S2 Chinese standard nameplate characters The matching rate is calculated using the following formula: ;in, Here is the character matching function; m is the character length; hour ,otherwise ;when If the nameplates match, the system will determine if they do; otherwise, it will mark the nameplate information as mismatched and record the difference characters. S54. For the merged video stream... The execution time sequence analysis of construction behavior in the process is performed, and the construction procedure sequence is extracted through a temporal convolutional network (TCN). Compared with the measured values ​​of process parameters This is compared with the mandatory regulations and rules library of the power industry analyzed in step S2. Standard process Standard parameters Compare and calculate the process compliance rate: ; ; ;in, This is a process sequence compliance function; it conforms to... Logical time ,otherwise ; For parameter-compliant functions; exist When within the threshold range ,otherwise ; Overall process compliance rate; S55, integration and matching rate Overall compliance rate of processes The result is that confidence is added to each construction event. Generate process event sequence .

[0010] Further improvements to this technical solution include step S6, which includes: S61, the key node determination module of the monitoring and surveillance device receives the process event sequence output in step S5. First, predefined key node events including equipment placement, grounding welding, test wiring, and removal of old equipment are selected to form a candidate set of key nodes. ;in, For key nodes The sequence number in S62; setting the confidence threshold for key nodes. Compare one by one Confidence of each event and The determination formula is: ;in, This indicates that the key node image capture has been triggered; Indicates that it will not be triggered; when At that time, record the event type of the key node. Corresponding verification results S63, triggers the HD camera to capture high-resolution images. The resolution was set to 4K, and the timestamp of the photo was recorded. ; Call the calibration positioning data generated in step S3 S64, will , The identity authentication identifier generated in step S1 , Step S2 downloads the power equipment asset code. and Packaged as raw trace data The format is: S65. Using the national cryptographic SM2 algorithm to... Perform digital signatures to generate immutable digital traces. The formula for the signature process is: ; ; ;in, for The SHA-256 hash value; A private key for the monitoring and surveillance device; For SM2 digital signature.

[0011] Secondly, this invention provides an artificial intelligence-based power engineering supervision and monitoring system, comprising: an identity authentication module for performing dual authentication of user identity and logging into the intelligent management and control platform through the supervision and monitoring device, the dual authentication including password verification and facial liveness detection verification; a project association module for obtaining an authorized list of technical renovation and overhaul projects from the power grid resource business platform in response to successful login, establishing an association between the supervision and monitoring device and the target project, and downloading the structured data of the feasibility study report corresponding to the project, the associated power equipment asset code, and the text data of the mandatory power industry regulations of the project's professional field; a location calibration module for obtaining initial geographic coordinates and equipment electronic tag information, performing intelligent location calibration through a multi-source fusion algorithm, and generating calibration positioning data at the meter to sub-meter level; a video acquisition module for activating the high-definition video recording function of the supervision and monitoring device, acquiring the construction site video stream in real time and caching it to the local storage unit; and a video stream fusion module for triggering the edge AI inference module of the supervision and monitoring device to perform video fusion. The video stream acquired by the frequency acquisition module undergoes polarization imaging and multispectral fusion preprocessing to generate a fused video stream. OCR recognition is performed on the nameplates of power equipment within the fused video stream, and time-series analysis is conducted on construction activities. The recognition results are then compared with the downloaded structured data from the feasibility study report and the text data of mandatory power industry regulations. A sequence of process events carrying confidence levels is output. The data binding module receives the process event sequence output by the video stream fusion module. When the confidence level of key nodes in the process event sequence exceeds a set threshold, a high-resolution image is automatically captured. This captured image is bound to the calibrated positioning data generated by the position calibration module, the identity information of the supervisor authenticated by the identity authentication module, the current timestamp, the associated power equipment asset code, and the verification results from the video stream fusion module, generating an immutable digital trace containing a national cryptographic SM2 digital signature. The data upload module uploads the digital trace and full-process supervision metadata to the intelligent management platform via a national cryptographic SM4 encrypted channel.

[0012] Thirdly, the present invention provides a terminal, comprising: a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the terminal performs the aforementioned terminal method.

[0013] Fourthly, the present invention provides a computer storage medium storing instructions that, when executed on a computer, cause the computer to perform the methods described in the above aspects.

[0014] The beneficial effects of this invention are as follows: by using face liveness detection and password dual authentication in step S1, the unique binding of "human-machine-account" is ensured, and identity theft is prevented; step S3 innovatively integrates Beidou differential positioning and anti-metal electronic tag scanning, which improves the positioning accuracy to meter-level / sub-meter-level in a 500kV strong electromagnetic interference environment, directly verifying whether the supervisor has accurately arrived at the equipment operation point, improving the identification rate of false work behaviors such as proxy attendance and wandering around the area, and ensuring that the supervisor performs his duties in a timely manner from the source.

[0015] Steps S4-S6 strongly associate and bind high-definition video streams, AI recognition results, location data, personnel identities, and equipment asset codes using the national cryptographic SM2 / SM4 algorithm to form an immutable digital trace with a spatiotemporal stamp and digital signature. Step S5 uses polarization imaging and multispectral fusion preprocessing to improve the nameplate OCR recognition rate under complex working conditions such as metal reflection and low light, improve the accuracy of process compliance judgment in construction behavior time sequence analysis, and realize automated consistency verification of "on-site behavior - design feasibility study - mandatory procedures" to ensure that the evidence chain is authentic, complete, and traceable.

[0016] Step S7 connects the field devices and the intelligent management and control platform through a national cryptographic encryption channel, and synchronizes the supervision data with the company-level PMS and risk control system in real time; Step S5 outputs a process event sequence that supports hierarchical early warning, turning post-event discovery into in-process intervention, and shortening the average problem response time from 24 hours to 15 minutes; Step S6 strongly binds the digital traces with the equipment asset code, realizes data traceability throughout the entire equipment life cycle, and makes the evaluation of the supervisor's performance quantifiable and objective, significantly improving project control and business penetration. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 is a schematic flowchart of a method according to an embodiment of the present invention.

[0019] Figure 2 is a schematic block diagram of a system according to an embodiment of the present invention.

[0020] Figure 3 is a schematic diagram of the structure of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the specific embodiments. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0023] Figure 1 is a schematic flowchart of an artificial intelligence-based power engineering supervision and monitoring method provided by the present invention. The executing entity in Figure 1 can be an artificial intelligence-based power engineering supervision and monitoring system. Depending on different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted.

[0024] As shown in Figure 1, the method includes: S1. Performing dual authentication of user identity and logging into the intelligent management and control platform through the monitoring and surveillance device, the dual authentication includes password verification and face liveness detection verification; S2. In response to successful login, obtaining the authorized list of technical renovation and overhaul projects from the power grid resource business platform, establishing the association between the monitoring and surveillance device and the target project, and downloading the structured data of the feasibility study report corresponding to the project, the associated power equipment asset code, and the text data of the mandatory power industry regulations of the project's professional category; S3. Obtaining the initial geographic coordinates and equipment electronic tag information, performing intelligent location calibration through a multi-source fusion algorithm, and generating meter-level to sub-meter-level calibration positioning data; S4. Activating the high-definition image recording function of the monitoring and surveillance device, collecting the construction site video stream in real time and caching it to the local storage unit; S5. Triggering the edge AI inference module of the monitoring and surveillance device to perform polarization analysis on the on-site video stream in step S4. The process involves preprocessing the image with multispectral data to generate a fused video stream. OCR recognition is then performed on the nameplates of the power equipment within the fused video stream, and a time-series analysis of the construction activities is conducted. The recognition results are then compared with the structured data from the feasibility study report downloaded in S2 and the text data of mandatory power industry regulations. A sequence of process events carrying confidence markers is output. In S6, the process event sequence output in step S5 is received. When the confidence level of key nodes in the process event sequence exceeds a set threshold, a high-resolution image is automatically captured. The captured image is then bound to the calibrated positioning data generated in step S3, the identity information of the supervisor certified in step S1, the current timestamp, the associated power equipment asset code, and the verification results from step S5 to generate an immutable digital trace containing a national cryptographic SM2 digital signature. Finally, the digital trace and full-process supervision metadata are uploaded to the intelligent management and control platform via a national cryptographic SM4 encrypted channel.

[0025] To facilitate understanding of the present invention, the following description further illustrates the artificial intelligence-based power engineering supervision and monitoring method provided by the present invention, using the principle of the artificial intelligence-based power engineering supervision and monitoring method of the present invention and the process of supervising and monitoring power engineering based on artificial intelligence in the embodiments.

[0026] First, step S1 includes: S11, the supervisor enters the username on the touch screen interface of the monitoring device. With password After receiving the input information, the core control unit of the monitoring and surveillance device calls the SHA-256 hash algorithm to encrypt the input password and generate a password hash value. The encryption formula is: ;in, The hash value of the input password is 256 bits long and is used for subsequent comparison with the pre-stored password hash value; The original password entered by the supervisor is transmitted in plaintext to the core control unit and then immediately hashed to avoid the risk of plaintext storage; S12, the core control unit synchronizes the pre-stored password hash value corresponding to the username from the intelligent management and control platform. Compared with pre-stored facial feature vectors ,Will and A consistency comparison is performed. If the two are completely identical, the password verification passes; if they are different, an audible and visual alarm is triggered, the alarm type is recorded as "password verification failed," and the information is uploaded to the intelligent management platform. The comparison formula is: ;in, This is the password verification result; This indicates that the verification has passed; Verification failed; S13, after password verification is successful, the device's front-facing camera starts face capture, continuously capturing 3 frames of the supervisor's face image. , , The face liveness detection unit performs dynamic feature analysis on the acquired images, and determines whether a person is alive by calculating the grayscale change value between adjacent frames. The determination formula is as follows: ;in, The liveness detection threshold is denoted by ; M and N are the number of rows and columns of pixels in the face image, respectively. Let (i,j) be the grayscale value of the k-th frame image at pixel (i,j); when If the liveness detection is successful, the face feature extraction unit will process the second frame image. Otherwise, it will be considered a fake medium, triggering an alarm and terminating authentication. (S14) Perform feature extraction to generate a 1024-dimensional face feature vector. Calculate the vector and the pre-stored face feature vector. The cosine similarity is calculated using the following formula: ;in, The similarity of feature vectors is defined, with values ​​ranging from [0,1]. The k-th dimension of the input face feature vector; For the pre-stored k-th dimension component of the face feature vector; when At that time, facial verification is successful; S15, after facial verification is successful, the core control unit generates a unique identity authentication identifier. The formula is: ;in, The unique serial number for the monitoring and surveillance device; This is the current login timestamp; This is used to bind digital traces to the identity of supervisors in step S6, ensuring a unique association between "person-trace-project", and is also uploaded to the intelligent management and control platform to record login logs.

[0027] Secondly, step S2 includes: S21, the supervision and monitoring device sends a project list request to the power grid resource business platform through the 5G communication module, and the request data packet carries the identity authentication identifier generated in step S1. This is used to verify the project authorization permissions of the supervisors in the middle platform. The request data packet format must meet the "Power Industry Data Transmission Protocol Specification". The formula is: ;in, Request a data packet for the item list; This is the header information, which includes the protocol version number and data length; The unique serial number for the monitoring and surveillance device hardware; Send a timestamp to request; The checksum is calculated using the following formula: ;in, S22: The data packet contains the i-th byte; n is the total number of bytes in the data packet, used to verify the integrity of the data packet transmission; S22: The power grid resource business platform receives the project list request data packet. Then, first pass the checksum. Verify the integrity of the data packet, then parse the authentication identifier. It also queries the associated authorized project list, filters out technical renovation and major repair projects with a status of "awaiting supervision," and generates a project list: Where m is the number of authorized projects, ranging from 3 to 10; each project Includes project number Project Name Planned supervision cycle The middle platform will include a project list. The encrypted data is returned to the monitoring device. The encryption uses the national standard SM4 algorithm, and the formula is as follows: ;in, This is the encrypted list of projects; The session key, 128 bits long, is negotiated between the monitoring and surveillance device and the power grid resource business platform via the ECC key exchange protocol to ensure the secure transmission of the project list; S23, the monitoring and surveillance device receives the encrypted project list. ,pass Decryption yields the project list The project list is displayed on the touchscreen for the supervisors to choose from; the supervisors then select the target project. Subsequently, the device sends a target project data request to the power grid resource service platform, carrying the target project number in the request. S24, The power grid resource business platform, based on the target project number... The system retrieves the original feasibility study report data and power equipment asset codes corresponding to the project from the PMS (Power Equipment Management System), and retrieves the text data of the mandatory power industry regulations for the project's specialty from the risk control platform. It then performs structured processing on the original feasibility study report data, extracting three core data categories: "equipment model standards," "construction procedure requirements," and "safety distance thresholds," to generate structured feasibility study report data. The structured processing formula is: ;in, This is an XML structuring conversion function that converts unstructured feasibility study report text into tagged XML format. This is the original data for the feasibility study report; This is an associated tag used to bind power equipment asset codes. , The k-th device has a unique asset code that conforms to the State Grid equipment coding rules; S25, the power grid resource business platform will use the structured data from the feasibility study report. Power equipment asset code and mandatory regulations text data in the power industry Packaged as target project dataset The data is transmitted to the monitoring and surveillance device via a 5G encrypted channel; upon receiving the data, the device will process the target project dataset. Stored to a local encrypted partition, the encryption formula is as follows: ;in, The local hardware encryption key for the monitoring and surveillance device (generated by the TPM 2.0 chip) ensures the security of local data storage and provides standard data support for the consistency verification in step S5.

[0028] Next, step S3 includes: S31, the electromagnetic interference sensor of the monitoring and surveillance device collects the electromagnetic interference intensity E of the construction site in real time at a frequency of 1Hz, and simultaneously starts the Beidou differential positioning module to collect the initial geographic coordinates at a frequency of 1Hz. S32. For outliers in the initial geographic coordinates caused by strong electromagnetic interference, a Kalman filter algorithm is used for noise reduction to obtain the filtered BeiDou coordinates. The filtering formula is: ; ; ;in, The coordinates after the kth filtering (i.e. F is the state transition matrix (values ​​are an identity matrix). ); Kalman gain; The initial geographic coordinates for the kth time (i.e.) H is the observation matrix (a matrix with values ​​of identity). Let I be the state covariance matrix at the k-th iteration; I is the identity matrix. To observe the noise covariance matrix (which dynamically adjusts with the electromagnetic interference intensity E; the larger E is, the better). The larger the value, the more complex the formula becomes. , (Base value under no interference); S33, start the anti-metal electronic tag scanning module, scan the UHF RFID anti-metal tag affixed to the surface of the target power equipment (such as transformer, switch cabinet), and obtain the preset coordinates of the equipment built into the tag. and equipment asset codes (Downloaded in step S2) (Consistent), scanning distance must meet t is the thickness of the equipment's metal casing (in mm), to avoid metal obstruction causing scanning failure; S34, calculate the average value of the filtered BeiDou coordinates. The formula is: ; then calculate and Positional deviation The formula is: S35. Based on positional deviation The weights of BeiDou positioning and tag positioning are dynamically allocated based on the electromagnetic interference intensity E to generate calibration positioning data. The weighting and standard formula are: when hour: , , ;when hour: , ; ; ; ;in, Weighting for BeiDou positioning; Weighting for anti-metal tags; when hour ,when hour ,make sure The positioning accuracy provides precise coordinates for the digital trace location stamp in step S6.

[0029] In addition, step S4 includes: S41, the environmental sensing module of the monitoring and surveillance device collects real-time data on the light intensity at the construction site. With sound decibel value The sampling frequency is 1Hz. The recording parameters of the high-definition camera are automatically adjusted based on the collected data. The parameter adjustment formula is: When... hour, The shutter speed of the HD camera was adjusted to 1 / 50s, and the LED fill light was turned on with a fill light intensity of [insert value here]. ;when hour, The shutter speed of the HD camera was adjusted to 1 / 200s, and the LED fill light was turned off; when When microphone noise reduction is enabled, the noise reduction intensity is [not specified]. ;in, For camera sensitivity; To increase the LED fill light intensity; The noise reduction intensity is set (value range 0-0.8); the parameters are dynamically adjusted to ensure the video stream clarity is ≥1080P, providing high-quality raw data for the image preprocessing in step S5; S42: Start the high-definition camera to perform image recording and generate a real-time video stream. The H.265 video coding standard is used for real-time video streams. Compression processing is performed, compression ratio Based on the calibration and positioning data generated in step S3 The associated construction scenario is dynamically adjusted using the following formula: Bitrate of the compressed video stream The calculation formula is: ;in, For video resolution (1080P corresponds to 2,073,600 pixels, 4K corresponds to 8,294,400 pixels). Set the recording frame rate (fixed at 25fps); The unit is Mbps; ensure normal scenarios. Key Scenarios To balance storage usage and image quality; S43, the compressed video stream is divided into video segments according to a time-slicing strategy. Duration of each segment The segmentation formula is: ;in, , This is the start time of recording. ; Add metadata tags to each ,in The associated device asset code downloaded in step S2; The average calibration positioning data during the recording of this segment (taken from step S3, and averaged after sampling at 1Hz). Record timestamps for the segments; S44, add metadata tags. The cached data is stored in a local encrypted storage unit and encrypted using the SM4 algorithm before storage. The encryption formula is as follows: ;in, The local storage key (generated by the TPM 2.0 chip of the monitoring device and cannot be exported) is used; the remaining local storage capacity is also monitored. ,when At that time, historical video clips that have been uploaded to the intelligent management and control platform and have been fully verified are automatically deleted to ensure sufficient cache space and provide continuous video clip input for edge AI inference in step S5.

[0030] Then, step S5 includes: S51, the edge AI inference module retrieves the cached video stream segments from the local storage unit in step S4. First, process each frame of the image. Perform polarization imaging preprocessing to eliminate reflective interference from the equipment's metal casing. The processing formula is as follows: ;in, This is a single-frame image after polarization processing; The original video frame grayscale value; The polarization angle of this frame image is acquired in real time by the camera's built-in polarization sensor, with a value ranging from 0°. 90°); S52, Call the multispectral fusion algorithm to convert the polarization-processed visible light image. Near-infrared images Fusion, the fusion formula is: ; ; ;in, The gray value of pixel (i,j) in the fused image; The fusion weight (value range 0.3-0.7); Image grayscale contrast; , These represent the maximum and minimum grayscale values ​​of the image, respectively; all fused frames form the fused video stream. S53, For fused video streams OCR recognition is performed on the nameplate area of ​​the power equipment. First, the nameplate area is located using the YOLOv8 object detection algorithm, and then the character sequence of the nameplate is extracted. Calculation and download of structured data from the feasibility study report in step S2 Chinese standard nameplate characters The matching rate is calculated using the following formula: ;in, Here is the character matching function; m is the character length; hour ,otherwise ;when If the nameplates match, the system will determine if they do; otherwise, it will mark the nameplate information as mismatched and record the difference characters. S54. For the merged video stream... The execution time sequence analysis of construction behavior in the process is performed, and the construction procedure sequence is extracted through a temporal convolutional network (TCN). Compared with the measured values ​​of process parameters This is compared with the mandatory regulations and rules library of the power industry analyzed in step S2. Standard process Standard parameters Compare and calculate the process compliance rate: ; ; ;in, This is a process sequence compliance function; it conforms to... Logical time ,otherwise ; For parameter-compliant functions; exist When within the threshold range ,otherwise ; Overall process compliance rate; S55, integration and matching rate Overall compliance rate of processes The result is that confidence is added to each construction event. Generate process event sequence This provides a basis for determining the key nodes in step S6.

[0031] Additionally, step S6 includes: S61, the key node determination module of the monitoring and surveillance device receives the process event sequence output in step S5. First, predefined key node events including equipment placement, grounding welding, test wiring, and removal of old equipment are selected to form a candidate set of key nodes. ;in, For key nodes The sequence number in S62; setting the confidence threshold for key nodes. Compare one by one Confidence of each event and The determination formula is: ;in, This indicates that the key node image capture has been triggered; Indicates that it will not be triggered; when At that time, record the event type of the key node. Corresponding verification results S63, triggers the HD camera to capture high-resolution images. The resolution was set to 4K, and the timestamp of the photo was recorded. ; Call the calibration positioning data generated in step S3 Ensure precise correlation with the construction location; S64, will , The identity authentication identifier generated in step S1 , Step S2 downloads the power equipment asset code. and Packaged as raw trace data The format is: S65. Using the national cryptographic SM2 algorithm to... Perform digital signatures to generate immutable digital traces. The formula for the signature process is: ; ; ;in, for The SHA-256 hash value; A private key for the monitoring and surveillance device; For SM2 digital signature; This data is used to verify the authenticity of the data after it is uploaded in step S7 by the intelligent management and control platform.

[0032] In some embodiments, the AI-based power engineering supervision and monitoring system 200 may include multiple functional modules composed of computer program segments. The computer programs of each program segment in the AI-based power engineering supervision and monitoring system 200 may be stored in the memory of a computer device and executed by at least one processor to perform (see Figure 1 for details) the AI-based power engineering supervision and monitoring functions.

[0033] In this embodiment, the AI-based power engineering supervision and monitoring system 200 can be divided into multiple functional modules according to its functions, as shown in Figure 2. These functional modules may include: an identity authentication module 210, a project association module 220, a location calibration module 230, a video acquisition module 240, a video stream fusion module 250, a data binding module 260, and a data upload module 270. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, stored in memory. In this embodiment, the functions of each module will be detailed in subsequent embodiments.

[0034] Specifically, the identity authentication module 210 is used to perform dual authentication of user identity and log in to the intelligent management and control platform through the supervision and monitoring device. The dual authentication includes password verification and facial liveness detection verification. The project association module 220 is used to obtain the authorized list of technical renovation and overhaul projects from the power grid resource business platform in response to successful login, establish the association between the supervision and monitoring device and the target project, and download the structured data of the feasibility study report corresponding to the project, the associated power equipment asset code, and the text data of the power industry mandatory regulations of the project's professional field. The location calibration module 230 is used to obtain the initial geographic coordinates and equipment electronic tag information, perform intelligent location calibration through a multi-source fusion algorithm, and generate calibration positioning data at the meter to sub-meter level. The video acquisition module 240 is used to activate the high-definition video recording function of the supervision and monitoring device, acquire the construction site video stream in real time, and cache it to the local storage unit. The video stream fusion module 250 is used to trigger the edge AI inference module of the supervision and monitoring device to perform on-site video acquisition by the video acquisition module. The video stream undergoes polarization imaging and multispectral fusion preprocessing to generate a fused video stream. OCR recognition is performed on the nameplates of power equipment within the fused video stream, and time-series analysis is conducted on construction activities. The recognition results are then compared with the downloaded structured data from the feasibility study report and the text data of mandatory power industry regulations. A sequence of process events carrying confidence levels is output. The data binding module 260 receives the process event sequence output by the video stream fusion module. When the confidence level of key nodes in the process event sequence exceeds a set threshold, it automatically captures high-resolution images. These captured images are bound to calibrated positioning data generated by the position calibration module, the identity information of the supervisor authenticated by the identity authentication module, the current timestamp, the associated power equipment asset code, and the verification results from the video stream fusion module, generating an immutable digital trace containing a national cryptographic SM2 digital signature. The data upload module 270 uploads the digital trace and full-process supervision metadata to the intelligent management platform via a national cryptographic SM4 encrypted channel.

[0035] Figure 3 is a schematic diagram of the structure of a terminal 300 provided in an embodiment of the present invention. The terminal 300 can be used to execute the power engineering supervision and monitoring method based on artificial intelligence provided in the embodiment of the present invention.

[0036] The terminal 300 may include a processor 310, a memory 320, and a communication module 330. These components communicate via one or more buses. Those skilled in the art will understand that the server structure shown in the figure does not constitute a limitation of the present invention. It may be a bus topology or a star topology, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0037] The memory 320 can be used to store the execution instructions of the processor 310. The memory 320 can be implemented by any type of volatile or non-volatile memory terminal or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. When the execution instructions in the memory 320 are executed by the processor 310, the terminal 300 is able to perform some or all of the steps in the above method embodiments.

[0038] The processor 310 serves as the control center of the storage terminal, connecting various parts of the electronic terminal via various interfaces and lines. It executes software programs and / or modules stored in the memory 320, and calls data stored in the memory to perform various functions of the electronic terminal and / or process data. The processor can be composed of integrated circuits (ICs), such as a single packaged IC or multiple packaged ICs with the same or different functions connected together. For example, the processor 310 may consist only of a central processing unit (CPU). In this embodiment of the invention, the CPU may have a single processing core or include multiple processing cores.

[0039] The communication module 330 is used to establish a communication channel, enabling the storage terminal to communicate with other terminals. It receives user data sent by other terminals or sends user data to other terminals.

[0040] The present invention also provides a computer storage medium, wherein the computer storage medium may store a program, which, when executed, may include some or all of the steps provided in the embodiments of the present invention. The storage medium may be a magnetic disk, an optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0041] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium such as a USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or other media capable of storing program code. It includes several instructions to cause a computer terminal (which may be a personal computer, server, or a second terminal, network terminal, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0042] The same or similar parts between the various embodiments in this specification can be referred to mutually. In particular, the terminal embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the description in the method embodiments.

[0043] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or modules may be electrical, mechanical, or other forms.

[0044] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0045] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0046] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. A power engineering supervision and monitoring method based on artificial intelligence, characterized in that, include: S1. Perform dual authentication of user identity and log in to the intelligent management platform through the monitoring and surveillance device. The dual authentication includes password verification and facial liveness detection verification. S2. Upon successful login, obtain the authorized list of technical renovation and overhaul projects from the power grid resource business platform, establish the association between the monitoring and surveillance device and the target project, and download the structured data of the feasibility study report corresponding to the project, the associated power equipment asset code, and the text data of the mandatory power industry regulations of the project's specialty. S3. Obtain initial geographic coordinates and equipment electronic tag information, perform intelligent location calibration through a multi-source fusion algorithm, and generate meter-level to sub-meter-level calibration positioning data. S4. Activate the high-definition video recording function of the monitoring and surveillance device, collect the construction site video stream in real time, and cache it to the local storage unit. S5. Trigger the edge AI inference module of the monitoring and surveillance device to process the steps. S4 performs polarization imaging and multispectral fusion preprocessing on the on-site video stream to generate a fused video stream. It then performs OCR recognition on the nameplates of the power equipment in the fused video stream, performs time-series analysis on the construction activities, and verifies the consistency of the recognition results with the structured data of the feasibility study report downloaded in S2 and the text data of mandatory power industry regulations. It outputs a sequence of process events carrying confidence level markers. S6 receives the process event sequence output in step S5. When the confidence level of key nodes in the process event sequence exceeds a set threshold, it automatically captures high-resolution images. The captured images are then bound to the calibrated positioning data generated in step S3, the identity information of the supervisor certified in step S1, the current timestamp, the associated power equipment asset code, and the verification results from step S5 to generate an unalterable digital trace containing a national cryptographic SM2 digital signature. S7. Upload digital traces and full-process supervision metadata to the intelligent management and control platform through the national cryptographic SM4 encryption channel.

2. The power engineering supervision and monitoring method based on artificial intelligence according to claim 1, characterized in that, Step S1 includes: S11, the supervisor enters the username on the touch screen interface of the monitoring device. With password After receiving the input information, the core control unit of the monitoring and surveillance device calls the SHA-256 hash algorithm to encrypt the input password and generate a password hash value. The encryption formula is: ;in, The hash value of the input password; The original password entered by the supervisor is transmitted in plaintext to the core control unit and then immediately hashed to avoid the risk of plaintext storage; S12, the core control unit synchronizes the pre-stored password hash value corresponding to the username from the intelligent management and control platform. Compared with pre-stored facial feature vectors ,Will and A consistency comparison is performed. If the two are completely identical, the password verification passes; if they are different, an audible and visual alarm is triggered, the alarm type is recorded as "password verification failed," and the information is uploaded to the intelligent management platform. The comparison formula is: ;in, This is the password verification result; This indicates that the verification has passed; Verification failed; S13, after password verification is successful, the device's front-facing camera starts face capture, continuously capturing 3 frames of the supervisor's face image. 、 、 The face liveness detection unit performs dynamic feature analysis on the acquired images, and determines whether a person is alive by calculating the grayscale change value between adjacent frames. The determination formula is as follows: ;in, The liveness detection threshold is denoted by ; M and N are the number of rows and columns of pixels in the face image, respectively. Let (i,j) be the grayscale value of the k-th frame image at pixel (i,j); when If the liveness detection is successful, the face feature extraction unit will process the second frame image. Otherwise, it will be considered a fake medium, triggering an alarm and terminating authentication. (S14) Perform feature extraction to generate a 1024-dimensional face feature vector. Calculate the vector and the pre-stored face feature vector. The cosine similarity is calculated using the following formula: ;in, The similarity of feature vectors is defined, with values ​​ranging from [0,1]. The input is the k-th dimension component of the face feature vector; For the pre-stored k-th dimension component of the face feature vector; when At that time, facial verification is successful; S15, after facial verification is successful, the core control unit generates a unique identity authentication identifier. The formula is: ;in, The unique serial number for the monitoring and surveillance device; This is the current login timestamp; Used for binding digital traces with the identity of supervisors in step S6.

3. The power engineering supervision and monitoring method based on artificial intelligence according to claim 2, characterized in that, Step S2 includes: S21, the monitoring and surveillance device sends a project list request to the power grid resource business platform through the 5G communication module, and the request data packet carries the identity authentication identifier generated in step S1. This is used to verify the project authorization permissions of the supervisors in the middle platform. The request data packet format must meet the "Power Industry Data Transmission Protocol Specification". The formula is: ;in, Request a data packet for the item list; This is the header information, which includes the protocol version number and data length; The unique serial number for the monitoring and surveillance device hardware; Send a timestamp to request; The checksum is calculated using the following formula: ;in, S22: The data packet contains the i-th byte; n is the total number of bytes in the data packet; S22: The power grid resource business platform receives the project list request data packet. Then, first pass the checksum. Verify the integrity of the data packet, then parse the authentication identifier. It also queries the associated authorized project list, filters out technical renovation and major repair projects with a status of "awaiting supervision," and generates a project list: Where m is the number of authorized projects; each project Includes project number Project Name Planned supervision cycle The middle platform will include a project list. The encrypted data is returned to the monitoring device. The encryption uses the national standard SM4 algorithm, and the formula is as follows: ;in, This is the encrypted list of projects; The session key is negotiated and generated between the monitoring and surveillance device and the power grid resource business platform through the ECC key exchange protocol; S23, the monitoring and surveillance device receives the encrypted project list. ,pass Decryption yields the project list The project list is displayed on the touchscreen for the supervisors to choose from; the supervisors then select the target project. Subsequently, the device sends a target project data request to the power grid resource service platform, carrying the target project number in the request. S24, The power grid resource business platform, based on the target project number... The system retrieves the original feasibility study report data and power equipment asset codes corresponding to the project from the PMS system, and retrieves the text data of the mandatory power industry regulations for the project's specialty from the risk control platform. It then performs structured processing on the original feasibility study report data, extracting three core data categories: "equipment model standards," "construction procedure requirements," and "safety distance thresholds," to generate structured feasibility study report data. The structured processing formula is: ;in, This is an XML structuring conversion function that converts unstructured feasibility study report text into tagged XML format. This is the original data for the feasibility study report; This is an associated tag used to bind power equipment asset codes. ， The unique asset code for the k-th device; S25, the power grid resource business platform will structure the data from the feasibility study report. Power equipment asset code and mandatory regulations text data in the power industry Packaged as target project dataset The data is transmitted to the monitoring and surveillance device via a 5G encrypted channel; upon receiving the data, the device will process the target project dataset. Stored to a local encrypted partition, the encryption formula is as follows: ;in, The local hardware encryption key for the monitoring and surveillance device.

4. The artificial intelligence-based power engineering supervision and monitoring method according to claim 3, characterized in that, Step S3 includes: S31, the electromagnetic interference sensor of the monitoring and surveillance device collects the electromagnetic interference intensity E of the construction site in real time at a frequency of 1Hz, and simultaneously starts the Beidou differential positioning module to collect the initial geographic coordinates at a frequency of 1Hz. S32. For outliers in the initial geographic coordinates caused by strong electromagnetic interference, a Kalman filter algorithm is used for noise reduction to obtain the filtered BeiDou coordinates. S33. Activate the anti-metal electronic tag scanning module to scan the UHF RFID anti-metal tag affixed to the surface of the target power equipment and obtain the preset coordinates of the equipment built into the tag. and equipment asset codes The scanning distance must meet the requirements. t is the thickness of the equipment's metal casing; S34, calculate the average value of the filtered BeiDou coordinates. The formula is: ; then calculate and Positional deviation The formula is: S35. Based on positional deviation The weights of BeiDou positioning and tag positioning are dynamically allocated based on the electromagnetic interference intensity E to generate calibration positioning data. The weighting and standard formula are: when hour: ， ， ;when hour: ， ； ； ； ;in, Weighting for BeiDou positioning; Weighting for anti-metal tags; when hour ,when hour ,make sure The positioning accuracy provides precise coordinates for the digital trace location stamp in step S6.

5. The power engineering supervision and monitoring method based on artificial intelligence according to claim 4, characterized in that, Step S4 includes: S41, the environmental sensing module of the monitoring and surveillance device collects real-time data on the light intensity at the construction site. With sound decibel value The sampling frequency is 1Hz. The recording parameters of the high-definition camera are automatically adjusted based on the collected data. The parameter adjustment formula is: When... hour, The shutter speed of the HD camera was adjusted to 1 / 50s, and the LED fill light was turned on with a fill light intensity of [insert value here]. ;when hour, The shutter speed of the HD camera was adjusted to 1 / 200s, and the LED fill light was turned off; when When microphone noise reduction is enabled, the noise reduction intensity is [not specified]. ;in, For camera sensitivity; To increase the LED fill light intensity; S42: To reduce noise intensity; Activate the HD camera to record video and generate a real-time video stream. The H.265 video coding standard is used for real-time video streams. Compression processing is performed, compression ratio Based on the calibration and positioning data generated in step S3 The associated construction scenario is dynamically adjusted using the following formula: Bitrate of the compressed video stream The calculation formula is: ;in, For video resolution; To record the frame rate; S43, divide the compressed video stream into video segments according to the time-slicing strategy. Duration of each segment The segmentation formula is: ;in, ， This is the start time of recording. ； Add metadata tags to each ,in The associated device asset code downloaded in step S2; This refers to the average calibration positioning data during the recording of this segment; Record timestamps for the segments; S44, add metadata tags. The cached data is stored in a local encrypted storage unit and encrypted using the SM4 algorithm before storage. The encryption formula is as follows: ;in, Use it as a local storage key; and monitor the remaining local storage capacity. ,when At that time, historical video clips that have been uploaded to the intelligent management and control platform and have been fully verified will be automatically deleted.

6. The power engineering supervision and monitoring method based on artificial intelligence according to claim 5, characterized in that, Step S5 includes: S51, the edge AI inference module retrieves the cached video stream segments from the local storage unit in step S4. First, process each frame of the image. Perform polarization imaging preprocessing to eliminate reflective interference from the equipment's metal casing. The processing formula is as follows: ;in, This is a single-frame image after polarization processing; The original video frame grayscale value; The polarization angle of this frame image is acquired in real time by the camera's built-in polarization sensor; S52, the multispectral fusion algorithm is invoked to process the polarized visible light image. Near-infrared images Fusion, the fusion formula is: ； ； ;in, The gray value of pixel (i,j) in the fused image; For weighting; Image grayscale contrast; 、 These represent the maximum and minimum grayscale values ​​of the image, respectively; all fused frames form the fused video stream. S53, For fused video streams OCR recognition is performed on the nameplate area of ​​the power equipment. First, the nameplate area is located using the YOLOv8 object detection algorithm, and then the character sequence of the nameplate is extracted. Calculation and download of structured data from the feasibility study report in step S2 Chinese standard nameplate characters The matching rate is calculated using the following formula: ;in, Here is the character matching function; m is the character length; hour ,otherwise ;when If the nameplates match, the system will determine if they do; otherwise, it will mark the nameplate information as mismatched and record the difference characters. S54. For the merged video stream... The execution time sequence analysis of construction behavior in the process is performed, and the construction procedure sequence is extracted through a temporal convolutional network (TCN). Compared with the measured values ​​of process parameters This is compared with the mandatory regulations and rules library of the power industry analyzed in step S2. Standard process Standard parameters Compare and calculate the process compliance rate: ； ； ;in, This is a process sequence compliance function; it conforms to... Logical time ,otherwise ； For parameter-compliant functions; exist When within the threshold range ,otherwise ； Overall process compliance rate; S55, integration and matching rate Overall compliance rate of processes The result is that confidence is added to each construction event. Generate process event sequence 。 7. The power engineering supervision and monitoring method based on artificial intelligence according to claim 6, characterized in that, Step S6 includes: S61, the key node determination module of the monitoring and surveillance device receives the process event sequence output in step S5. First, predefined key node events including equipment placement, grounding welding, test wiring, and removal of old equipment are selected to form a candidate set of key nodes. ;in, For key nodes The sequence number in S62; setting the confidence threshold for key nodes. Compare one by one Confidence of each event and The determination formula is: ;in, This indicates that the key node image capture has been triggered; Indicates that it will not be triggered; when At that time, record the event type of the key node. Corresponding verification results S63, triggers the HD camera to capture high-resolution images. The resolution was set to 4K, and the timestamp of the photo was recorded. ; Call the calibration positioning data generated in step S3 S64, will 、 The identity authentication identifier generated in step S1 、 Step S2 downloads the power equipment asset code. and Packaged as raw trace data The format is: S65. Using the national cryptographic SM2 algorithm to... Perform digital signatures to generate immutable digital traces. The formula for the signature process is: ； ； ;in, for The SHA-256 hash value; A private key for the monitoring and surveillance device; For SM2 digital signature.

8. A power engineering supervision and monitoring system based on artificial intelligence, characterized in that, include: The identity authentication module is used to perform dual authentication of user identity and log in to the intelligent management and control platform through the supervision and monitoring device. The dual authentication includes password verification and face liveness detection verification. The project association module is used to obtain the authorized list of technical renovation and overhaul projects from the power grid resource business platform in response to successful login, establish the association between the supervision and monitoring device and the target project, and download the structured data of the feasibility study report, the associated power equipment asset code, and the text data of the power industry mandatory regulations of the project's professional category. The location calibration module is used to acquire initial geographic coordinates and device electronic tag information, and performs intelligent location calibration through a multi-source fusion algorithm to generate calibration positioning data at the meter to sub-meter level. The video acquisition module is used to activate the high-definition video recording function of the supervision and monitoring device, and to acquire the video stream of the construction site in real time and cache it to the local storage unit. The video stream fusion module is used to trigger the edge AI inference module of the supervision and monitoring device to perform polarization imaging and multispectral fusion preprocessing on the on-site video stream acquired by the video acquisition module, generate a fused video stream, and perform OCR recognition on the nameplates of power equipment in the fused video stream, perform time sequence analysis on the construction behavior, and perform consistency verification between the recognition results and the downloaded structured data of the feasibility study report and the text data of the mandatory regulations of the power industry, and output a sequence of process events with confidence level markers; The data binding module is used to receive the process event sequence output by the video stream fusion module, and automatically capture high-resolution images when the confidence of key nodes in the process event sequence exceeds a set threshold. The captured images are bound with the calibrated positioning data generated by the position calibration module, the identity information of the supervisor authenticated by the identity authentication module, the current timestamp, the associated power equipment asset code, and the verification results of the video stream fusion module to generate an immutable digital trace containing the national cryptographic SM2 digital signature. The data upload module is used to upload digital traces and full-process supervision metadata to the intelligent management and control platform through the national cryptographic SM4 encryption channel.

9. A terminal, characterized in that, include: processor; A memory for storing execution instructions of a processor; wherein the processor is configured to perform the method of any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images