A power marketing field operation inspection efficiency optimization method, device and equipment

Abstract

This application relates to the field of power marketing technology. It discloses a method, apparatus, and equipment for optimizing the efficiency of on-site operations in power marketing. The method includes: acquiring multi-source data of on-site operations from multiple locations; performing independent and cross-validation on the multi-source data of on-site operations at each location to generate operation status verification results; based on the operation status verification results of the multi-source data of on-site operations from multiple locations, selecting target efficiency optimization areas from the multiple locations and determining the type of the target efficiency optimization area; determining target historical optimization data based on the type of the target efficiency optimization area, and determining historical optimization measures and target correlation coefficients based on the target historical optimization data; and generating optimization implementation guidance information based on the on-site operation multi-source data, historical optimization measures, and target correlation coefficients corresponding to the target efficiency optimization area. This application achieves the integration and correlation analysis of cross-modal data and can generate targeted optimization implementation guidance information.

Description

Technical Field

[0001] This application relates to the field of power marketing technology, and more specifically, to a method, apparatus, and equipment for optimizing the efficiency of on-site power marketing operations inspection. Background Technology

[0002] Electricity marketing field operations involve multiple types of data, including video recordings, equipment parameters, and work logs. However, traditional operation inspections often rely on manual verification of single-type data, lacking the ability to integrate and analyze cross-modal data. For example, during meter installation inspections, operators often check paper log entries or review on-site videos for compliance, but struggle to correlate the operational steps in the videos with changes in equipment parameters, leading to information gaps between data.

[0003] Meanwhile, traditional operational inspection efficiency optimization lacks precise regional positioning logic, often adopting a uniform adjustment approach across the entire process, which is difficult to match with the actual problems in different areas. For example, in power marketing sites, there are both operational execution areas with chaotic data integration and terminal equipment areas with slow data transmission. However, traditional optimization usually involves uniformly increasing inspection personnel or upgrading equipment, which cannot specifically solve the problems of chaotic data integration and redundant processes in the operational execution areas, nor can it efficiently improve the problems of slow data transmission and delays in the terminal equipment areas. Instead, it leads to resource misallocation and poor optimization results. Summary of the Invention

[0004] In view of the above situation, this application provides a method, apparatus and equipment for optimizing the efficiency of on-site inspection of power marketing operations, which aims to solve the above problems or at least partially solve the above problems.

[0005] Firstly, this application provides a method for optimizing the efficiency of on-site inspections in electricity marketing, including: Acquire multi-source data on field operations at multiple locations; Based on pre-set verification standards, the multi-source data of the on-site operations at each location are independently and cross-verified to generate operation status verification results. Based on pre-set area type determination rules, and according to the multi-source data of field operations at multiple locations and their corresponding operation status verification results, target efficiency optimization areas are selected from multiple locations and the type of the target efficiency optimization areas is determined. Based on a pre-set optimization database, historical optimization data of the target is determined according to the type of the target efficiency optimization region, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data of the target. Based on a pre-set optimization generation model, optimization implementation guidance information is generated according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0006] Secondly, this application provides a device for optimizing the efficiency of on-site inspections in power marketing, comprising: The acquisition module is used to acquire multi-source data on field operations at multiple locations; The verification module is used to perform independent and cross-verification of the multi-source data of the on-site operation at each location based on the pre-set verification standards, and generate operation status verification results. The region filtering module is used to filter out target efficiency optimization regions from multiple locations and determine the type of the target efficiency optimization regions based on pre-set region type determination rules, according to the multi-source data of the on-site operations at multiple locations and their corresponding operation status verification results. The query module is used to determine the target historical optimization data based on the pre-set optimization database and the type of the target efficiency optimization region, and to determine the historical optimization measures and target correlation coefficient based on the target historical optimization data; The generation module is used to generate optimization implementation guidance information based on a pre-set optimization generation model, according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0007] Thirdly, this application provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the power marketing field operation inspection efficiency optimization method as described in the first aspect.

[0008] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the power marketing field operation inspection efficiency optimization method described in the first aspect.

[0009] The above-described technical solutions adopted in the embodiments of this application can achieve the following beneficial effects: This application achieves cross-modal data integration and correlation analysis by independently verifying and cross-verifying multi-source data from on-site operations. This improves the accuracy of efficiency optimization analysis. Further analysis of the verification results identifies target efficiency optimization areas, generates targeted optimization measures and implementation guidelines, and matches different adjustment methods to problems in different areas, thereby improving resource utilization and optimization effectiveness. Attached Figure Description

[0010] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of an application environment for a method for optimizing the efficiency of on-site operations inspection in power marketing, according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating a method for optimizing the efficiency of on-site inspections in power marketing, according to an embodiment of the present invention. Figure 3 This is a schematic diagram of a device for optimizing the efficiency of on-site operations inspection in power marketing, according to one embodiment of the present invention. Figure 4 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. 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.

[0012] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the term "comprising" and its variations should be interpreted as open-ended terms meaning "including but not limited to."

[0013] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0014] As mentioned earlier, current operational inspections largely rely on manual verification of single-type data, lacking sufficient capabilities for cross-modal data integration and correlation analysis. Efficiency optimization also lacks precise regional positioning logic, making it difficult to match the actual problems in different regions. To address this technical issue, this application provides a method for optimizing the efficiency of on-site operational inspections in power marketing.

[0015] The method for optimizing the efficiency of on-site inspections in power marketing provided in this embodiment of the invention can be applied to, for example... Figure 1In this application environment, the device communicates with the server via a network. The server can acquire multi-source data of field operations from multiple locations through the device; based on pre-set verification standards, it performs independent and cross-verification on the multi-source data of field operations at each location, generating operation status verification results; based on pre-set area type determination rules, it filters out target efficiency optimization areas from multiple locations and determines the type of the target efficiency optimization area based on the multi-source data of field operations at multiple locations and their corresponding operation status verification results; based on a pre-set optimization database, it determines target historical optimization data according to the type of the target efficiency optimization area, and determines historical optimization measures and target correlation coefficients based on the target historical optimization data; based on a pre-set optimization generation model, it generates optimization implementation guidance information based on the multi-source data of field operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficients. This application achieves cross-modal data integration and correlation analysis by independently verifying and cross-verifying multi-source data from on-site operations. This improves the accuracy of efficiency optimization analysis. Further analysis of the verification results identifies target efficiency optimization areas, generates targeted optimization measures and implementation guidelines, and matches different adjustment methods to problems in different areas, thereby improving resource utilization and optimization effectiveness.

[0016] The device side can include, but is not limited to, various personal computers, laptops, smartphones, tablets, portable wearable devices, and various device sensors. The server side can be implemented using a standalone server or a server cluster consisting of multiple servers. The invention will now be described in detail through specific embodiments.

[0017] Please see Figure 2 As shown, Figure 2 A flowchart illustrating the method for optimizing the efficiency of on-site inspections in power marketing, as provided in this embodiment of the invention, includes the following steps: S1: Acquire multi-source data of field operations at multiple locations.

[0018] In one embodiment, the multi-source data for on-site operations includes video operation data, equipment operation data, and operation log data.

[0019] In one embodiment, video operation data refers to video data recorded by staff using video recording equipment during on-site operations. For example, video recording of the entire process of metering device installation / replacement: recording the complete operation of staff removing the old meter, installing the new meter, wiring and debugging, and installing the lead seal; video of on-site inspection of electricity usage: capturing images of the operating status of the distribution cabinet of commercial users, investigation of suspected electricity theft points, and the integrity of the metering box seal; video of charging pile operation and maintenance: recording charging pile fault diagnosis (such as gun head testing, motherboard debugging) and on-site repair operations.

[0020] In one embodiment, the equipment operation data refers to the on-site equipment operation data during staff work. For example, real-time data collected by smart meters includes user voltage, current, power factor, daily electricity consumption, peak and off-peak load, etc.; distribution transformer monitoring data covers transformer oil temperature, load rate, three-phase imbalance, active / reactive power, supporting area load prediction and overload warning; charging pile operation parameters include charging voltage, charging current, charging time, charging amount, equipment online status, fault codes, etc.

[0021] In one embodiment, the work log data consists of paper logs recorded by staff during on-site operations. Examples include: on-site work orders (recording staff information, task number, work time, equipment number, operation details, and on-site safety measure confirmation signatures); customer power survey forms (recording the on-site environment, list of electrical equipment, load calculation results, and power supply scheme recommendations for new / upgraded users); and metering device calibration records (paper versions of meter error calibration reports, transformer ratio test records, and seal number registrations).

[0022] In one embodiment, video operation data can be uploaded to the server via recording devices carried by staff, equipment operation data can be uploaded to the server via sensors and detection instruments on the equipment, and operation log data can be uploaded to the server via manual entry or photography.

[0023] In one embodiment, video operation data, equipment operation data, and operation log data include location coordinates and timestamps.

[0024] In one embodiment, the multi-source data of the on-site operation is a data set of multiple sampling times within a preset verification time window. The verification time window can be flexibly set according to the operation type, inspection cycle or user instructions, such as the complete duration of a single operation, a specific working period or a specified historical backtracking interval.

[0025] The method includes: based on a pre-set verification time window, obtaining video operation data, equipment operation data, and operation log data within the verification time window according to the timestamp; and based on the location coordinates, combining the video operation data, equipment operation data, and operation log data whose differences between location coordinates are within a preset distance range to form multi-source data of on-site operations at a location; and finally, obtaining multi-source data of on-site operations at multiple locations.

[0026] S2: Based on the pre-set verification standards, perform independent verification and cross-verification on the multi-source data of the on-site operation at each location to generate operation status verification results.

[0027] In one embodiment, the operation status verification result is a conclusion obtained by comparing the multi-source data of the on-site operation with the operation specification standard, such as whether the operator's operation conforms to the process, and whether the equipment operating parameters are within the standard range.

[0028] In one embodiment, the verification standards include independent verification standards and cross-verification standards. Independent verification standards include standards for various types of data, video operation standards, equipment parameter standards, and logbook standards. Video operation standards include worker attire specifications, requirements for wearing safety protective equipment, compliant on-site operating procedures, key steps in equipment installation or maintenance, standardized safety briefing scripts, clarity of on-site communication instructions, and prohibited scenarios for unauthorized dialogue. Equipment parameter standards include normal range thresholds for key indicators such as voltage, current, power, and temperature of various types of equipment, and criteria for judging abnormal fluctuations. Logbook standards include text verification rules for work order information completeness, chronological logic of operation records, and standardization of signature and approval processes.

[0029] In one embodiment, step S2 includes: S21: Based on a pre-set preprocessing model, perform spatiotemporal alignment and structuring processing on the video operation data, the equipment operation data, and the operation log data.

[0030] In one embodiment, spatiotemporal alignment is performed based on the timestamps in the video operation data, the equipment operation data, and the operation log data. Spatiotemporal alignment is achieved using timeline synchronization and spatiotemporal alignment technology, ensuring a one-to-one correspondence between video frames, equipment parameter sampling points, and log record entries on the timeline.

[0031] After spatiotemporal alignment, based on the pre-set alignment model and structured model, the multi-source data of the on-site operation corresponding to each location is structured. The content extracted from video operation data, equipment operation data, and operation log data is filled into the preset structured template to obtain the structured data corresponding to the video operation data, equipment operation data, and operation log data respectively.

[0032] After spatiotemporal alignment and structuring, the video operation data, equipment operation data, and operation log data proceed to the next steps. The video operation data, equipment operation data, and operation log data in the subsequent steps refer to the data after spatiotemporal alignment and structuring.

[0033] S22: Based on the first identification model and video operation standard in the verification standard, identify the video attire, video dialogue, video actions, and video flow in the video operation data according to the first identification model, and independently verify the video attire, the video dialogue, and the video actions according to the video operation standard.

[0034] In one embodiment, the first recognition model employs image recognition and speech recognition technologies to identify images in video frames, recognizing clothing, actions, and workflow. Based on the changes in video actions over timestamps, the video workflow is identified, referring to the on-site work steps. Image recognition technologies include, but are not limited to, YOLO, Faster R-CNN, Two-Stream Networks, 3D Convolutional Networks (C3D), and OpenPose.

[0035] In one embodiment, audio in video job data is identified using speech recognition technology to recognize video dialogue. Speech recognition technology includes, but is not limited to, Whisper, DeepSpeech, and natural language processing technology.

[0036] In one embodiment, video operation data is used to verify the compliance of the operation process, the completeness and time sequence of key actions, whether the steps are executed in the standard order, whether the key actions are completed completely, and to identify whether there are any violations during the operation, such as not wearing safety equipment or not disconnecting the power as required, in order to determine whether the operation process is compliant.

[0037] In one embodiment, the video attire is compared with the worker attire specifications and safety equipment wearing requirements in the video operation standard to obtain an attire verification result. For example, if the video attire is identified as employee work clothes, safety gloves, and a safety helmet, the comparison result is "Employee work clothes are in compliance, safety gloves are worn correctly, and safety helmets are worn correctly; verification is qualified." The video actions are compared with the compliant actions in the on-site operation procedures in the video operation standard to obtain an action verification result. For example, if the video actions include raising a hand, holding a brake, and pulling down, the comparison result is "Raising a hand is correct, holding a brake is correct, and pulling down is correct; verification is qualified." The video process is compared with the key steps of equipment installation or maintenance in the video operation standard to obtain a step verification result. For example, if the identified video actions change sequentially with the timestamp as raising a hand, holding a brake, and pulling down, and the video process identifies this step as power off, then based on subsequent video actions, the subsequent video process is identified as removing the old meter, installing the new meter, wiring and debugging, and installing the seal; the comparison result is "Power off, removing the old meter, installing the new meter, wiring and debugging, and installing the seal are correct steps; verification is qualified." The video script is compared with the safety briefing script specifications, clarity of on-site communication instructions, and prohibited dialogue scenarios in the video operation standard to obtain the script verification result. For example, the script text in the video script is identified and verified. The verification result is "The script text is '......', the script is standardized, the instructions are clear, and it does not belong to the prohibited dialogue scenarios, so the verification is qualified".

[0038] In one embodiment, the independent verification result of the video job standard includes pass / fail, as well as specific descriptions of the identified video attire, video script, video actions, and video process, and the verification result of the corresponding verification standard.

[0039] S23: Based on the second identification model and the device parameter standards in the verification standard, identify various device parameters in the device operation data according to the second identification model, and independently verify the various device parameters according to the device parameter standards.

[0040] In one embodiment, the second recognition model is an OCR model capable of extracting fields and recognizing the parameter values ​​of each parameter. Based on the parameter values, parameter volatility can also be calculated. Volatility can be the proportion of the difference between the parameter value and a preset benchmark value, or it can be the standard deviation, coefficient of variation, or variance of the parameter. Independent verification can also be performed on the parameter volatility.

[0041] In one embodiment, the device operation data is verified for compliance with numerical range and stability of fluctuations.

[0042] In one embodiment, the standard for each device parameter is the normal allowable range and the allowable fluctuation range for each parameter. The verification result of the device operation data includes the parameter name, parameter value, volatility, and whether it is qualified or unqualified. The verification result of the device operation data may also include the parameter trend, which refers to the trend of the latter parameter value relative to the former parameter value between two adjacent parameter values, such as rising, falling, or stable. It is judged according to the preset parameter value difference range. When the difference between the former and latter parameter values ​​is within the preset parameter value difference range, it is considered stable. When the latter parameter value is greater than the former parameter value and the difference between the former and latter parameter values ​​exceeds the preset parameter value difference range, it is considered rising. When the latter parameter value is less than the former parameter value and the difference between the former and latter parameter values ​​exceeds the preset parameter value difference range, it is considered falling.

[0043] S24: Based on the third identification model and the work log standard in the verification standard, identify the text fields, text actions, and operation processes in the work log data according to the third identification model, and independently verify the text fields and text actions according to the work log standard.

[0044] In one embodiment, the third recognition model is an OCR model capable of extracting fields, including text fields, text actions, and operation processes.

[0045] In one embodiment, the work log verification checks the completeness and standardization of required fields. For example, the verification of on-site work orders requires checking whether the required fields, such as worker information, task number, work time, equipment number, operation content, and safety measure confirmation signature, are complete and standardized. Completeness refers to whether all required fields are filled in; all fields are considered complete, and incomplete fields are considered incomplete. Standardization refers to whether the prescribed format and steps are used for signature approval. For example, the standard format for work time is "year-month-day-hour-minute," and only the year, month, day, and hour in the recognized text fields are considered non-standard. The verification result is "Required fields complete or incomplete, standardized or non-standardized, incomplete items are…, non-standardized items are…, inspection passed or failed."

[0046] In one embodiment, the text actions are compared with the compliant actions in the on-site operation procedures in the work log standard to obtain the action verification result. For example, if the identified text action includes "pulling the switch," the action verification result after comparison is "The switch-pulling action is correct, and the verification is qualified." According to the sequence of the identified text actions changing with the timestamp, the actions are: pulling the switch, removing the old meter, installing the new meter, wiring and debugging, and installing the lead seal; this constitutes the operation procedure.

[0047] S25: Based on the cross-validation criteria in the verification standard, perform consistency verification on the video flow and the operation flow, perform consistency verification on the video action and the various device parameters, and perform consistency verification on the text field and the various device parameters.

[0048] In one embodiment, the cross-validation standard sets a baseline range for verifying the consistency of any two data sets. This baseline range can be a range of matching degrees. Consistency is determined by the matching degree.

[0049] In one embodiment, the method for performing consistency verification on the video flow and the operation flow further includes: identifying the device number in the video job data and the device number in the job log data; calculating the matching degree between the device number in the video job data and the device number in the job log data; calculating the matching degree between the video flow and the operation flow; calculating the matching degree between the timestamp of the video flow and the timestamp of the operation flow; performing a first matching degree calculation based on the preset weights of each dimension and the matching degree calculation results of each dimension; and obtaining a consistency verification result based on the preset matching degree range and the first matching degree calculation result.

[0050] For example, the matching degree between the equipment number in the video operation data and the equipment number in the operation log data can be calculated using a pre-set string similarity algorithm, resulting in a continuous value between 0 and 1. A complete match is 1, and a complete mismatch is 0. String similarity algorithms include, but are not limited to, cosine similarity algorithms or Jaccard similarity coefficients. For instance, if the equipment number in the video operation data is "P-01-A01-23-001" and the equipment number in the operation log data is "P-01-A01-23-002", with segment weights of major category code (0.3) - minor category code (0.2) - department code (0.2) - year code (0.1) - serial number (0.2), then the serial number matching degree is 0, and the equipment number matching degree = 1×0.3 + 1×0.2 + 1×0.2 + 1×0.1 + 0×0.2 = 0.8.

[0051] For example, when calculating the matching degree between the timestamps of the video process and the operation process, a pre-set timestamp matching degree calculation formula can be used. For instance, the deviation between the timestamps of the video process and the operation process can be calculated; the smaller the deviation, the higher the matching degree. The formula can be set as: Timestamp Matching Degree = 1 - |Timestamp of Video Process - Timestamp of Operation Process| / Preset Maximum Allowable Deviation. If the preset maximum allowable deviation is 60 seconds, and the actual deviation is 10 seconds, then the timestamp matching degree = 1 - 10 / 60 ≈ 0.83.

[0052] For example, when calculating the matching degree between the video flow and the operation flow, the matching degree of the operation content can be calculated using a preset action sequence similarity algorithm. This algorithm includes, but is not limited to, the Dynamic Time Warping (DTW) algorithm. The formula can be set as: Operation content matching degree = Number of similar actions / Total number of actions. For instance, if the operation content in the video flow contains 3 actions, and the operation flow includes 2 similar actions, the operation content matching degree = 2 / 3 ≈ 0.67.

[0053] In one embodiment, the preset matching degree range includes a first matching degree range, a second matching degree range, and a third matching degree range. When the first matching degree is within the first matching degree range, the verification result is "First matching degree is ..., high consistency"; when the first matching degree is within the second matching degree range, the verification result is "First matching degree is ..., medium consistency, further verification is needed"; when the first matching degree is within the third matching degree range, the verification result is "First matching degree is ..., low consistency, there may be violations or recording errors, further verification is needed". For example, the first matching degree range is matching degree ≥ 0.8, the second matching degree range is 0.6 ≤ matching degree < 0.8, and the third matching degree range is matching degree < 0.6. If the weight of the device number dimension is 0.3, the weight of the timestamp dimension is 0.3, and the weight of the operation content dimension is 0.4, then the first matching degree = 0.8×0.3 + 0.83×0.3 + 0.67×0.4 = 0.757. The first matching degree is within the range of the second matching degree, and the output verification result is "The first matching degree is 0.757, consistent, but there are some non-standard parts that need further verification".

[0054] In one embodiment, the method for verifying the consistency between the video action and the various device parameters includes: determining a device number based on the video job data; determining a timestamp of the video action based on the video action; determining a video flow based on the video action; determining a first feature timestamp, a target parameter volatility, and a target parameter trend corresponding to various device parameters based on the video flow; searching for the target parameter volatility and target parameter trend among the corresponding various device parameters based on the timestamp of the video action and the device number; determining a second feature timestamp based on the timestamp of the found target parameter volatility and target parameter trend; calculating a second matching degree based on the first feature timestamp and the second feature timestamp; and obtaining a consistency verification result based on a preset matching degree range and the second matching degree calculation result.

[0055] For example, the video actions are identified as raising a hand, gripping the brake, and pulling down, and the timestamps for raising the hand, gripping the brake, and pulling down are obtained. Based on the actions of raising the hand, gripping the brake, and pulling down, the video process is determined to be a power outage, and the first feature timestamp is the pull-down timestamp. Then, the current in the parameters of device A shows a decreasing trend in the target parameter at the pull-down timestamp, and the volatility of the target parameter is greater than the preset benchmark volatility. In the current data after the raising hand timestamp in the parameters of device A, the current data whose parameter volatility and trend match the target parameter volatility and trend are identified. After finding the data, the timestamp of the current data that matches the target parameter volatility and trend is determined as the second timestamp. The second matching degree is calculated based on the first feature timestamp and the second feature timestamp. For example, if a similarity algorithm is pre-set, and the calculation formula is: Second Matching Degree = (1 - |First Feature Timestamp - Second Feature Timestamp| / Preset Time Threshold), and the first feature timestamp is "March 1, 2026, 5:17:15 AM", the second feature timestamp is "March 1, 2026, 5:17:16 AM", and the preset time threshold is 10 seconds, the second matching degree is 0.9. Since the second matching degree falls within the range of the first matching degree, the output result is "Second matching degree is 0.9, high consistency".

[0056] In one embodiment, the consistency verification of the text field and the various device parameters includes: extracting the device parameter type, device parameter value, timestamp corresponding to the parameter value, and device number from the text field; searching for matching parameter values ​​and timestamps corresponding to the matching parameter values ​​among the various device parameters based on the device number, the device parameter type, and the timestamp corresponding to the parameter value; calculating a third matching degree based on the device parameter value, the matching parameter value, the timestamp corresponding to the parameter value, and the timestamp corresponding to the matching parameter value; and obtaining a consistency verification result based on a preset matching degree range and the third matching degree calculation result.

[0057] For example, the device parameter type, device parameter value, timestamp corresponding to the parameter value, and device number extracted from the text field are as follows: current, 5A, "March 1, 2026, 5:17:15", and "P-01-A01-23-001". Based on "P-01-A01-23-001", current, and "March 1, 2026, 5:17:15", the matching parameter value and the timestamp corresponding to the matching parameter value are searched in various device parameters. For example, 5.1A and "March 1, 2026, 5:17:16". The preset current threshold is 0.5A, the preset time threshold is 10 seconds, the dimension weight of current is 0.4, and the dimension weight of timestamp is 0.6. The third matching degree = (1-|5-5.1| / 0.5)×0.4+(1-1 / 10)×0.6=0.86. The third matching degree falls within the range of the first matching degree, and the output test result is "The third matching degree is 0.86, and the consistency is high".

[0058] In one embodiment, for nonconformities, the inspection results may also include the deviation value, risk level, and mismatch description corresponding to the nonconformity. The deviation value is the deviation rate between the identified data and the data in the verification standard and the preset baseline deviation. Examples include the deviation rate of equipment operating parameters from the standard range, the similarity deviation rate between video actions and the standard process, and the deviation rate between ledger records and actual time points. The risk level is automatically determined based on the magnitude of the deviation value, the scope of influence of the deviation item, and the importance of the operational step, using preset risk mapping rules. The larger the deviation, the wider the scope of influence, and the more critical the operational step, the higher the risk level. The mismatch description refers to the structured text description formed by the server automatically extracting information such as the feature dimension, location, numerical difference, deviation value, and risk level of the mismatch based on the comparison results. A nonconformity refers to each item that needs to be inspected; each item represents a feature dimension, including video attire, video dialogue, video actions, each type of parameter in various equipment parameters, text fields, text actions, operating procedures, and three items for consistency verification.

[0059] S26: Generate the job status verification result based on the independent verification results and the cross-verification results.

[0060] In one embodiment, the structured text of the independent verification results of each item and the cross-validation results of each item constitutes the job status verification result.

[0061] This application generates verification results for each data object, enabling comprehensive and thorough verification of all aspects of on-site electricity marketing operations, thus providing a basis for subsequent regional division.

[0062] S3: Based on the pre-set area type determination rules, according to the multi-source data of the on-site operations at multiple locations and their corresponding operation status verification results, select the target efficiency optimization area from multiple locations and determine the type of the target efficiency optimization area.

[0063] In one embodiment, the region type determination rule includes a correlation strength calculation rule, a correlation strength evaluation threshold determination rule, a first optimized region division rule, a priority calculation rule, a priority threshold determination rule, a second optimized region division rule, and a region classification rule.

[0064] In one embodiment, step S3, based on pre-set area type determination rules, filters out target efficiency optimization areas from multiple locations and determines the type of the target efficiency optimization areas according to the multi-source data of on-site operations at multiple locations and their corresponding operation status verification results, including: S31: Based on the association strength calculation rules, calculate the data association strength of each location according to the operation status verification results of each type of data at each location.

[0065] In one embodiment, data association strength refers to the degree of association between different types of data in cross-modal multi-source data, including the degree of association between video operation data and equipment operation data, the degree of association between video operation data and operation log data, and the degree of association between equipment operation data and operation log data, which can be calculated by a first matching degree, a second matching degree, and a third matching degree.

[0066] In one embodiment, the association strength calculation rule refers to calculating the data association strength based on the first matching degree, the second matching degree, the third matching degree, and their corresponding preset weights. For example, if the weight of the first matching degree is 0.3, the weight of the second matching degree is 0.3, and the weight of the third matching degree is 0.4, then the data association strength = 0.757 × 0.3 + 0.9 × 0.3 + 0.86 × 0.4 = 0.841. Based on the above association strength calculation rule, the data association strength of each location is calculated according to the first matching degree, the second matching degree, and the third matching degree in the operation status verification result of each location.

[0067] In one embodiment, a correlation strength distribution map is generated based on the location coordinates of each location.

[0068] In one embodiment, based on pre-set first, second, and third association strength ranges, the data association strength of each location is compared with these ranges. Locations within the first association strength range are considered high-association regions, those within the second are medium-association regions, and those within the third are low-association regions. For example, the first association strength range is defined as data association strength ≥ 0.7, the second as 0.5 ≤ matching degree < 0.7, and the third as < 0.5. If the data association strength of a location is 0.841, then that location is considered a high-association region.

[0069] S32: Based on the association strength assessment threshold determination rule, determine the association strength assessment threshold according to the data association strength of multiple locations.

[0070] In one embodiment, the association strength assessment threshold determination rule refers to statistically analyzing the data association strength at all locations within the verification time window to obtain a data association strength statistical curve, and then extracting an association strength assessment threshold from the data association strength statistical curve based on a preset confidence level. The association strength assessment threshold is a critical value that distinguishes between strong and weak associations.

[0071] In one embodiment, the data association strength corresponding to a preset confidence level of the data association strength statistical curve is identified according to the association strength assessment threshold determination rule, which is the association strength assessment threshold. For example, if there are 20 data association strengths on the association strength statistical curve, sorting them from smallest to largest gives an ordered sequence: [0.12, 0.23, 0.27, 0.31, 0.35, 0.42, 0.48, 0.53, 0.57, 0.59, 0.61, 0.63, 0.68, 0.72, 0.75, 0.77, 0.81, 0.85, 0.89, 0.94]. The data association strength corresponding to an 80% confidence level is identified as 0.77, i.e., the association strength assessment threshold is 0.77.

[0072] The correlation strength assessment threshold in this application is no longer a fixed manual parameter, but an adaptive parameter dynamically generated by the server based on the actual situation of the work site data. This ensures the compatibility between the correlation strength assessment threshold and the work status verification results, and also eliminates the problems of subjectivity, lag and insufficient adaptability caused by purely manual setting.

[0073] The correlation strength assessment threshold in this application is based on the statistical analysis of the data correlation strength at all locations within each verification time window. When the overall data correlation strength is low, the correlation strength assessment threshold is also low, thus tightening the correlation strength assessment threshold. When the overall data correlation strength is high, the correlation strength assessment threshold is also high, thus relaxing the correlation strength assessment threshold. This allows for dynamic adjustment of the threshold's leniency or strictness.

[0074] S33: Based on the first optimization region division rule, select the positions where the data association strength is greater than the association strength evaluation threshold as the first optimization region.

[0075] In one embodiment, multiple locations are initially screened using a correlation strength assessment threshold. Regions exhibiting issues with cross-modal data fusion and collaboration are identified as candidate regions requiring efficiency optimization, i.e., the first optimization region. For example, when the correlation strength assessment threshold is 0.77, the locations corresponding to data correlation strengths of 0.81, 0.85, 0.89, and 0.94 are all located within the first optimization region.

[0076] S34: Based on the priority calculation rules, calculate the priority coefficient of each location according to the operation status verification results of the multi-source data of the on-site operation at each location in the first optimization area, and sort the multiple locations by priority.

[0077] In one embodiment, the priority calculation rule quantifies the risk level as a value between 0 and 1 based on the risk level of the operation status verification result at each location and a pre-set level quantification rule. For example, the quantified value of the first risk level is 0, the quantified value of the second risk level is 0.25, the quantified value of the third risk level is 0.5, the quantified value of the fourth risk level is 0.75, and the quantified value of the fifth risk level is [value missing].

[0078] In one embodiment, the priority calculation rule may consider factors such as the importance of the work area and the frequency of historical problems, in addition to the risk level. The importance of the work area and the frequency of historical problems are calculated using historical statistical data and stored in a database. During application, in response to a call command, the system calls the appropriate method based on the location coordinates of each location within the first optimization area. Both the importance of the work area and the frequency of historical problems are quantified values ​​between 0 and 1.

[0079] In one embodiment, the priority calculation rule calculates the priority coefficient for each location based on the risk level, the importance of the work area, the quantified value of the frequency of historical problems, and the corresponding preset priority weight, and then sorts them according to the priority coefficient. For example, the priority coefficient for a location with a data association strength of 0.81 is 0.9, the priority coefficient for a location with a data association strength of 0.85 is 0.8, the priority coefficient for a location with a data association strength of 0.89 is 0.7, and the priority coefficient for a location with a data association strength of 0.94 is 0.6. The sorting result is: location with a data association strength of 0.94, location with a data association strength of 0.89, location with a data association strength of 0.85, and location with a data association strength of 0.81.

[0080] S35: Based on the priority threshold determination rule, determine the priority threshold according to the priority coefficient of each position in the first optimization region.

[0081] In one embodiment, the priority threshold determination rule uses a preset quantile value to determine the priority threshold based on the priority coefficients of multiple positions. For example, if the preset quantile value is 15%, the priority threshold for the priority ordered sequence [0.6, 0.7, 0.8, 0.9] is 0.645.

[0082] S36: Based on the second optimization region division rule, select positions with priority coefficients greater than the priority threshold as the second optimization region.

[0083] In one embodiment, a priority threshold is introduced to further analyze the urgency of optimization of multiple locations within the first optimization region, thereby addressing the problem of insufficient resources when multiple locations within the first optimization region are optimized simultaneously, and achieving a reasonable allocation of optimization resources.

[0084] For example, if the priority threshold is 0.645, then positions with priority coefficients greater than the priority threshold are optimized first. That is, positions with priority coefficients of 0.9, 0.8, and 0.7 are all located in the second optimization region.

[0085] S37: Based on the region classification rules, extract the first type of features and the second type of features from the multi-source data of the field operation corresponding to each location in the second optimization region, and select the location where the first type of features or the second type of features are extracted as the target efficiency optimization region. The target efficiency optimization region type where the first type of features are extracted is the first type of efficiency optimization region, and the target efficiency optimization region type where the second type of features are extracted is the second type of efficiency optimization region.

[0086] In one embodiment, the first type of features includes data fusion integrity features, job resource scheduling features, and process execution continuity features; the second type of features includes delay repair timeliness features and data transmission stability features.

[0087] In one embodiment, the first type of efficiency optimization area is an area where data is closely correlated but fusion is problematic. The second type of efficiency optimization area is an area where data correlation meets the standards but transmission is problematic. For example, in a certain location, video operation data and equipment operation data have a high degree of correlation, but there is information duplication during data integration; this type of area belongs to the first type of efficiency optimization area. In another example, in a certain location, the data correlation strength meets the standards, but the time for equipment operation data to be transmitted from the field to the backend exceeds the standard limit; this type of area belongs to the second type of efficiency optimization area.

[0088] In one embodiment, first-type features and second-type features are extracted at each location in the second optimization region. Based on the extracted feature type, the corresponding location is divided into a first-type efficiency optimization region or a second-type efficiency optimization region. That is, the location where the first-type feature is extracted is the first-type efficiency optimization region, and the location where the second-type feature is extracted is the second-type efficiency optimization region. This achieves the simultaneous completion of feature extraction and region division, providing a direct basis for subsequent targeted optimization.

[0089] In one embodiment, first-class features and second-class features are extracted based on the multi-source data of field operations after preprocessing in step S21.

[0090] In one embodiment, the process of extracting first-type and second-type features from multi-source data of on-site operations corresponding to a location is described: A. For the data fusion integrity feature, the fusion integrity score is calculated by analyzing the information overlap, missing rate and matching degree of different modal data.

[0091] Specifically, firstly, by analyzing the interaction logic and integration path of historical video operation data, historical equipment operation data, and historical operation log data for each location, a first feature parsing model containing interaction logic rules and integration path templates is trained and pre-set in the system. Then, upon receiving new video operation data, new equipment operation data, and new operation log data, the system automatically parses these data based on the pre-set first feature parsing model, thereby extracting the corresponding data fusion integrity features.

[0092] Specifically, the interaction logic rules and integration path templates in the first feature parsing model are not manually drafted on an ad-hoc basis, but are configured during the system initialization phase, becoming a fixed automated processing flow. When new multi-source data from on-site operations arrives, the preset first feature parsing model is directly invoked for comparison and judgment, without manual intervention. The interaction logic rules define the transmission, matching, and correlation relationships between video operation data, equipment operation data, and operation log data during the operation process. The integration path template presets the specific process for data aggregation and verification.

[0093] For example, in the meter installation area, the interaction logic rules in the preset first feature parsing model are invoked to automatically match the wiring operation steps recorded by the video operation data, the meter voltage changes recorded by the equipment operation data, and the operation time nodes recorded by the operation log data. The interaction logic among the three is that the operation steps trigger voltage changes and record the time nodes synchronously. Then, the data integrity is verified by integrating the path template. The integrated path template is that the field terminal first summarizes the wiring operation steps and the meter voltage changes, then compares them with the operation time nodes and uploads them to the backend. Finally, based on the predefined key information items and corresponding weights of data fusion integrity, the quantitative result of data fusion integrity features is automatically generated, thereby extracting the data fusion integrity features.

[0094] Specifically, the method includes: pre-defining key information items and corresponding weights for data fusion integrity in the first feature parsing model; then verifying whether each key information item is missing, incorrect, or mismatched based on the interaction logic rules and integration path template in the first feature parsing model; and calculating the information integrity rate based on the verification results. The information integrity rate is calculated as: (sum of weights of verified key information items) / (sum of weights of all key information items) × 100%. The information integrity rate is the data fusion integrity feature. The data fusion integrity feature reflects the completeness of information coverage during the integration process.

[0095] For example, if the integrated data can completely cover all key information regarding operation steps, voltage changes, and time points, the feature performs well; if some data on voltage changes is missing, the feature performs poorly. Assuming the preset weights for operation steps, voltage changes, and time points are 0.4, 0.3, and 0.3 respectively; if the integrated data can completely cover all key information regarding operation steps, voltage changes, and time points, the information completeness rate = (0.4 + 0.3 + 0.3) / (0.4 + 0.3 + 0.3) × 100% = 100%, and the data fusion completeness feature is 100%. If the information completeness rate reaches the preset information completeness rate threshold (e.g., ≥95%), the feature is considered to perform well; if some data on voltage changes is missing, the corresponding key information item fails verification, and the information completeness rate does not reach the preset information completeness rate threshold, the feature is considered to perform poorly.

[0096] B. For the characteristics of work resource scheduling, it is based on the usage records of on-site equipment, personnel, tools and materials, and statistically analyzes the quantitative indicators of resource types, quantities and arrival times.

[0097] Specifically, the resource scheduling characteristics of the operation are obtained by analyzing the allocation methods of various resources during the operation at each location. Resources include personnel, tools, on-site equipment, and materials. Allocation methods refer to the rules governing the distribution and flow of these resources at different stages of the operation.

[0098] Specifically, firstly, a second feature analysis model is trained based on historical video operation data, historical equipment operation data, and historical operation ledger data. The second feature analysis model sets resource allocation and flow rule templates. Through the second feature analysis model, various resources and their allocation methods identified in the video operation data, equipment operation data, and operation ledger data during the operation at a certain location are automatically compared and judged to obtain the operation resource scheduling features.

[0099] Specifically, the resource allocation and flow rule templates set in the second feature parsing model are not manually drafted on an ad-hoc basis, but are configured during the system initialization phase, becoming fixed automated processing logic. When new field data arrives, the preset second feature parsing model is directly invoked for matching and verification, without manual intervention. Resource allocation defines the types, quantities, and configuration standards of resources required for different work stages. Flow rules preset the scheduling sequence, response time limits, and flow paths of resources between each stage.

[0100] For example, in the meter installation area, the resource allocation rules are as follows: Electricians, multimeters, and insulating tape are needed at the beginning of the work; meter replacement is needed during the middle of the work. The flow rules are: resources are delivered to the site 10 minutes before the start of the work, and electricians are arranged to be on site at the same time; upon receiving a request to replace a meter, a spare meter is dispatched to the site within 5 minutes.

[0101] Specifically, the method includes: identifying the resource types, quantities, and arrival times in video operation data, equipment operation data, and operation log data; verifying the identified resource types, quantities, and arrival times based on predefined resource allocation and flow rule templates in the second feature analysis model to obtain the resource allocation rationality rate and scheduling timeliness rate; and obtaining the resource scheduling rate, i.e., the operation resource scheduling feature, based on the resource allocation rationality rate and scheduling timeliness rate. The operation resource scheduling feature is a feature that characterizes the rationality of resource allocation and the timeliness of scheduling. Resource allocation rationality rate = (sum of weights of verified resource types) / (sum of weights of all resource types) × 100%, scheduling timeliness rate = (sum of weights of verified scheduling items) / (sum of weights of all scheduling items) × 100%, resource scheduling rate = weight 1 × resource allocation rationality rate + weight 2 × scheduling timeliness rate, where weight 1 and weight 2 are preset. When the deviation rate between the actual quantity and the theoretical quantity is within the preset quantity deviation rate range, the verification is passed; when it exceeds the preset quantity deviation rate range, the verification fails. The verification is considered successful when the deviation rate between the actual scheduling time and the theoretical scheduling time is within the preset time deviation rate range; otherwise, the verification is considered unsuccessful.

[0102] For example, in the meter installation area, the initial stage of the work requires electricians, multimeters, and insulating tape, while the middle stage requires meter replacement. The flow rule is to deliver resources to the site 10 minutes before the start of the work, and simultaneously arrange for an electrician to be on site; upon receiving the meter replacement request, a spare meter should be dispatched to the site within 5 minutes. The resource types include electricians, multimeters, insulating tape, and spare meters, with corresponding weights of 0.3, 0.2, 0.2, and 0.3, respectively. The actual quantities of these resources are 2, 3, 10, and 2, respectively, and the theoretical quantities are 2, 3, 15, and 2, respectively. The dispatching items include delivering resources to the site 10 minutes before the start of the work, having an electrician on site 10 minutes before the start of the work, and dispatching a spare meter to the site within 5 minutes of receiving the meter replacement request, with corresponding weights of 0.3, 0.4, and 0.3, respectively. The actual dispatching times for these items are 8 minutes, 8 minutes, and 10 minutes, respectively, and the theoretical dispatching times are 10 minutes, 10 minutes, and 5 minutes, respectively. If all resource type items pass the verification, but the verification of "dispatching a backup meter to the site within 5 minutes of receiving a meter replacement request" in the dispatch item fails, the resource allocation rationality rate = (0.3+0.2+0.2+0.3) / (0.3+0.2+0.2+0.3)×100%=100%, the dispatch timeliness rate = (0.3+0.4) / (0.3+0.4+0.3)×100%=70%, and the resource dispatch rate = 0.6×100%+0.4×70%=88%. If the resource dispatch rate reaches the preset resource dispatch rate threshold (e.g., ≥90%), the performance of this feature is considered good; if the resource dispatch rate does not reach the preset resource dispatch rate threshold, the performance of this feature is considered weak. It can be seen that if the resource quantity matches the workload and the dispatch time fits the work rhythm, it indicates that the resource allocation is reasonable and the dispatch timeliness meets the standard, which will make the performance of this feature better.

[0103] C. Regarding the continuity characteristics of process execution, the server automatically identifies process bottlenecks, repetitive operations, and timing misalignments by comparing operation videos, ledger records, and standard process sequences, and generates corresponding continuity quantification values.

[0104] Specifically, the continuity characteristics of the process execution are extracted by analyzing the connection characteristics of each work step at each location. The work steps include meter unpacking, wiring, parameter adjustment, and record keeping. The connection characteristics refer to the transition methods and time intervals between the steps.

[0105] Specifically, firstly, a third feature analysis model is trained based on historical video operation data, historical equipment operation data, and historical operation log data. The third feature analysis model sets up a template for the connection rules of operation links. The third feature analysis model automatically compares and judges the operation links identified in the video operation data, equipment operation data, and operation log data at a certain location, as well as the connection features of each operation link, and extracts the process execution continuity features.

[0106] Specifically, the transition methods and time interval standards between each work step in the third feature parsing model are not manually configured on an ad-hoc basis, but are configured during the system initialization phase, becoming fixed automated processing logic. When new field data arrives, the preset third feature parsing model is directly invoked for matching and verification, without manual intervention. The transition methods and time interval standards between each work step include predefined triggering conditions, timing relationships, and time intervals between each work step.

[0107] For example, in the installation of electricity meters, the work process includes wiring, parameter debugging, and record keeping. The timing relationship between each work process is as follows: after wiring, parameter debugging is performed, and after parameter debugging, record keeping is performed. The triggering conditions and time intervals for parameter debugging are described as follows: after the wiring process is completed, the parameter debugging process should be triggered by a signal indicating that the equipment parameters are normal, and the interval between the two processes should be controlled within 2 minutes; after the parameter debugging process is completed, the record keeping process should be triggered by a signal indicating the debugging result, and the interval should be controlled within 1 minute.

[0108] Specifically, the method includes: identifying work steps, triggering conditions, temporal relationships, and time intervals in video operation data, equipment operation data, and work log data; verifying the identified work steps, triggering conditions, temporal relationships, and time intervals based on the predefined transition methods and time interval standard templates between each work step in the third feature analysis model to obtain the process execution continuity rate, i.e., the process execution continuity feature. Process execution continuity rate = (sum of weights of verified work step items) / (sum of weights of all work step items) × 100%. When a work step is triggered according to a preset triggering condition and the interval time meets the preset time interval threshold requirement, this work step passes verification; when a work step is not triggered according to the preset triggering condition or the interval time does not meet the preset time interval threshold requirement, this work step fails verification.

[0109] For example, during the installation of electricity meters, if the weights of wiring, parameter adjustment, and ledger recording are 0.3, 0.4, and 0.3 respectively, and the wiring, parameter adjustment, and ledger recording are all verified after the inspection is completed, then the process execution continuity rate = (0.3 + 0.4 + 0.3) / (0.3 + 0.4 + 0.3) × 100% = 100%.

[0110] The continuity of process execution is a characteristic that reflects the smoothness of transitions between steps and the continuity of execution. If the triggering signals between steps are clear and the time intervals meet the standards, it indicates a smooth transition. If each step can proceed continuously without stagnation, it indicates continuous execution. These factors will make the performance of this characteristic more prominent.

[0111] In one embodiment, the operational steps—wiring, parameter debugging, and record keeping—can be verified during the independent verification process of the video operation data. Here, the verification results are directly retrieved from step S2. Passing the verification corresponds to qualification, and failing the verification corresponds to non-qualification.

[0112] D. Delay repair timeliness feature: Taking the moment when data transmission delay occurs as the starting node, the time taken from the starting node to the termination node when data transmission returns to normal is counted. The actual time taken is compared with the preset standard repair time. The shorter the time taken, the stronger the timeliness of delay repair. This feature can intuitively reflect the emergency response efficiency after transmission failure.

[0113] Specifically, the method includes: parsing the transmission data of video operation data and equipment operation data according to a pre-set fourth feature parsing model, and obtaining the delay repair timeliness feature based on the parsing results. The fourth feature parsing model includes the judgment criteria for transmission delay, scene classification rules, the mapping relationship between preset scenes and time thresholds, and the delay repair timeliness calculation rules.

[0114] Specifically, the fourth feature analysis model pre-sets multiple scenarios, such as network congestion during the morning rush hour: network congestion caused by multiple devices uploading data at the same time during the morning rush hour, packet loss in areas with weak signal coverage, failure to synchronize meter reading data to the backend in a timely manner, and stagnation in the ledger recording process.

[0115] Specifically, the method includes: determining the transmission delay based on the transmission delay alarm signal extracted from the transmission data of video operation data and equipment operation data. The timestamp of receiving the transmission delay alarm signal is the starting node.

[0116] Specifically, the method includes: after determining the transmission delay, extracting key indicators such as the number of concurrent devices, network bandwidth utilization, and data transmission success rate for the current time period from the transmission data of video operation data and device operation data; classifying the scenario based on pre-set scenario classification thresholds for each key indicator; and determining the scenario type. For example, if the number of concurrent devices is 100 (total number of devices = 120), the network bandwidth utilization is 95%, and the data transmission success rate is 85%, and the thresholds for the number of concurrent devices, network bandwidth utilization, and data transmission success rate during the morning peak network congestion scenario are 50, 80%, and 80%, then when all key indicator scenarios are greater than the corresponding key indicator thresholds for the morning peak network congestion scenario, the scenario type is determined to be the morning peak network congestion scenario.

[0117] Specifically, the method includes: extracting a second timestamp of the repair operation signal and a third timestamp of the termination node where data transmission resumes normal, from the first timestamp of the starting node, based on the transmission data of video operation data and equipment operation data; calculating a first duration between the second timestamp and the first timestamp; calculating a second duration between the third timestamp and the second timestamp; determining a first duration threshold range and a second duration threshold range corresponding to the corresponding scenario type based on a preset mapping relationship between scenarios and time thresholds; calculating the delay repair efficiency, i.e., the delay repair timeliness characteristic, based on the first duration, the first duration threshold range, the second duration, and the second duration threshold range. If the first duration meets the first duration threshold range, the first time item verification passes; otherwise, the first time item verification fails. If the second duration meets the second duration threshold range, the second time item verification passes; otherwise, the second time item verification fails. Delay repair efficiency = (sum of weights of verified time items) / (sum of weights of all time items) × 100%.

[0118] For example, firstly, a delay alarm signal is captured, and the signal trigger time is recorded as the first timestamp. Subsequent repair operation signals and transmission status data are continuously monitored to determine the second and third timestamps. Then, the first duration is calculated as 2 minutes and the second duration as 10 minutes. If the first duration threshold is 0-5 minutes and the second duration threshold is 5-15 minutes, then both the first and second time items pass the verification. If the weight of the first time item is 0.6 and the weight of the second time item is 0.4, then the delay repair efficiency = (0.6 + 0.4) / (0.6 + 0.4) × 100% = 100%. If the delay repair efficiency reaches the preset delay repair efficiency threshold (e.g., ≥90%), then the feature is considered to perform well, with a fast repair response speed. If the delay repair efficiency does not reach the preset delay repair efficiency threshold, then the feature is considered to perform poorly, with delayed repair measures and excessive time consumption.

[0119] First, the latency-generating scenarios and impact range of the second type of efficiency optimization region during cross-modal data transmission are tracked to obtain latency repair timeliness characteristics. Cross-modal data transmission includes the transmission of on-site video and equipment parameter data to the backend system. Latency-generating scenarios refer to the specific operational steps and triggering conditions that cause data transmission delays, while the impact range refers to the operational processes or areas affected by the delay. Based on pre-set and system-implanted transmission latency rule templates, the server automatically tracks and judges the latency-generating scenarios and impact range of the second type of efficiency optimization region during cross-modal data transmission, thereby obtaining latency repair timeliness characteristics.

[0120] E. Data transmission stability characteristics refer to the comprehensive evaluation of data transmission interruption time intervals, interruption durations, and transmission rate fluctuation levels within a specified monitoring period. The lower the frequency of interruptions, the shorter the duration of a single interruption, and the lower the transmission rate fluctuation level, the better the overall stability of data transmission.

[0121] Specifically, the method includes: parsing the transmission data of video operation data and equipment operation data according to a pre-set fifth feature parsing model, and obtaining data transmission stability characteristics based on the parsing results. The fifth feature parsing model includes an interruption time interval threshold, an interruption duration threshold, a transmission rate fluctuation level evaluation rule, and a data transmission stability calculation rule.

[0122] For example, taking meter reading data transmission as an example, based on a continuous transmission duration of 1 hour, the preset interruption interval threshold is 15 minutes, and the interruption duration threshold is 3 seconds. If the interruption interval is greater than or equal to 15 minutes, the interruption interval item passes the verification; otherwise, the interruption interval item fails the verification. If the interruption duration is less than or equal to 3 seconds, the interruption duration item passes the verification; otherwise, the interruption duration item fails the verification.

[0123] For example, the transmission rate fluctuation level evaluation rule takes meter reading data transmission as an example. Based on a continuous transmission duration of 1 hour and a data transmission rate benchmark of 10 Mbps, it presets multiple transmission rate fluctuation threshold ranges: the first range is 0 to 1 Mbps, the second range is 1 Mbps to 2 Mbps, the third range is 2 Mbps to 4 Mbps, and the fourth range exceeds 4 Mbps. When the difference between the real-time transmission rate and the benchmark value is within the first threshold range, the fluctuation level is Level 1; when it is within the second threshold range, the level is Level 2; when it is within the third threshold range, the level is Level 3; and when it is within the fourth threshold range, the level is Level 4. For instance, if the real-time transmission rate is 7 Mbps and the difference between it and the benchmark value is 3 Mbps, the fluctuation level is Level 3. Furthermore, a quantification value is set for each volatility level, such as 1 for the first level, 0.75 for the second level, 0.5 for the third level, and 0.25 for the fourth level. These quantification values ​​are for illustrative purposes only and do not represent actual usage; users can set them according to their needs.

[0124] Specifically, the data transmission stability rate = Interruption time interval item weight × (sum of the number of verified interruption time interval items) / (sum of the number of all interruption time interval items) × 100% + Interruption duration item weight × (sum of the number of verified interruption duration items) / (sum of the number of all interruption duration items) × 100% + Fluctuation level item weight × Fluctuation level quantification value × 100%. For example, based on a continuous transmission duration of 1 hour, if there are 4 interruptions, 1 successful interruption time interval item verification, and 2 successful interruption duration item verifications, with a fluctuation level of Level 1, then the data transmission stability rate = 0.3 × 1 / 4 × 100% + 0.3 × 2 / 4 × 100% + 0.4 × 1 × 100% = 62.5%.

[0125] Data transmission stability characteristics characterize the fluctuations and interruptions during data transmission. This characteristic is more prominent when the signal is stable and transmission is continuous; conversely, it weakens when signal fluctuations are large and transmission interruptions are frequent. The fifth feature analysis model can also generate multi-level transmission stability characteristic grades based on pre-set data transmission stability rate ranges, including four levels: Excellent, Good, Medium, and Poor. When the data transmission stability rate is within the first data transmission stability rate range, with stable signal and continuous transmission, the characteristic is rated Excellent. When the data transmission stability rate is within the second data transmission stability rate range, with small signal fluctuations and occasional short-term interruptions, it is rated Good. When the data transmission stability rate is within the third data transmission stability rate range, with large signal fluctuations and frequent interruptions, it is rated Medium. When the data transmission stability rate is within the fourth data transmission stability rate range, with drastic signal fluctuations and frequent interruptions, it is rated Poor. The entire process is automatically completed by the server based on preset benchmark values ​​and multi-level thresholds, requiring no manual intervention.

[0126] S4: Based on the pre-set optimization database, determine the target historical optimization data according to the type of the target efficiency optimization region, and determine the historical optimization measures and target correlation coefficients based on the target historical optimization data.

[0127] In one embodiment, when the target efficiency optimization region is a first type of efficiency optimization region; step S4, based on a pre-set optimization database, determines the target historical optimization data according to the type of the target efficiency optimization region, and determines historical optimization measures and target correlation coefficients based on the target historical optimization data, including: S411: Based on the optimization database, determine the target historical optimization data according to the first type of feature corresponding to the first type of efficiency optimization region. The target historical optimization data includes historical first type of feature, historical first optimization measure, and historical first optimization magnitude.

[0128] Specifically, the first type of feature corresponding to the first type of efficiency optimization region is the data fusion integrity feature, job resource scheduling feature, and process execution coherence feature extracted in step S3 based on the multi-source data of the on-site operation at this location in step S1. Each location within the first type of efficiency optimization region has its corresponding first type of feature.

[0129] Specifically, based on the first type of features of the first type of efficiency optimization region, similar data is searched from the optimization database according to the pre-set search rules, and similar historical data is determined as target historical optimization data.

[0130] Specifically, the optimization database stores full-process data of all past power marketing field operations, including multi-source data of cross-modal field operations for each batch of operations, operation status verification results, resource scheduling records, process execution logs, first type of features, second type of features, first optimization measures, first optimization effect, and first optimization improvement.

[0131] Specifically, the search rules include pre-setting search threshold ranges for each feature. Features in the first category of the optimization database that fall within these search threshold ranges are considered target first-category features, and the optimization data corresponding to these target first-category features are considered target historical optimization data. The search threshold range for each feature fluctuates by a preset percentage based on the current feature value. For example, if the first-category features at a certain location in the first-category efficiency optimization region include data fusion integrity (0.8), job resource scheduling (0.85), and process execution coherence (0.9), with a preset fluctuation percentage of 10% of the search threshold range, then the search threshold range for the data fusion integrity feature is [0.72, 0.88], the search threshold range for the job resource scheduling feature is [0.77, 0.94], and the search threshold range for the process execution coherence feature is [0.81, 0.99]. Historical optimization data that simultaneously satisfies the following criteria in the optimization database is considered target historical optimization data.

[0132] This range-based fuzzy matching method ensures the similarity of matching scenarios while avoiding missing valuable historical optimization cases due to minor differences in feature values. These target historical optimization data are relevant data from historical operations where there were deviations in data fusion completeness, resource scheduling, and process coherence, and where efficiency improvements were achieved through optimization measures—in other words, the portion of the first type of feature where there is room for efficiency optimization.

[0133] For example, based on the first type of features corresponding to the location of electricity meter installation work in the first type of efficiency optimization area, and based on the search threshold range of each feature, target historical optimization data was selected. The target historical optimization data shows that: the data fusion integrity feature of a certain batch of operations shows that there is a 15% information overlap between equipment parameters and operation records; the operation resource scheduling feature shows that the average waiting time for electricians to use tools is 8 minutes; and the process execution continuity feature shows that the longest interval between links is 10 minutes.

[0134] In one embodiment, the target historical optimization data establishes mapping relationships for location, multi-source on-site operation data, operation status verification results, resource scheduling records, process execution logs, historical first-type features, historical first-type optimization measures, historical first-type optimization effects, and historical first-type optimization magnitude. The target historical optimization data can be determined based on the historical first-type features, and is a complete dataset including the historical first-type features, historical first-type optimization measures, and historical first-type optimization magnitude. Once the target historical optimization data is determined, the historical first-type optimization measures and historical first-type optimization magnitude can be determined.

[0135] In one embodiment, the historical first optimization measure refers to targeted measures that have already been implemented and can be used to generate corresponding optimization implementation guidance information. These include process simplification measures, resource allocation measures, and connection optimization measures. For example, process simplification measures include merging repetitive data collection steps and adjusting the interaction logic of cross-modal data to reduce information duplication and process redundancy; resource allocation measures include adjusting the scheduling order of tools and personnel based on historical data and preparing resources in advance at key work nodes to avoid waiting time; and connection optimization measures include adding automatic verification nodes between steps to ensure that data is synchronized promptly after the previous step is completed, reducing connection delays.

[0136] In one embodiment, the historical first optimization magnitude is a quantification of the optimization effect achieved after the implementation of the historical first optimization measure. For example, in the historical data of meter installation operations, after the implementation of the historical first optimization measure, the interval between steps was shortened from 8 minutes to 3 minutes, and the historical first optimization magnitude was 40%.

[0137] S412: Based on the multiple sets of historical first-class features and the first historical optimization magnitude, the first linear regression equation is obtained.

[0138] In one embodiment, a set of historical first-class features and the first historical optimization magnitude consist of historical data fusion integrity features, historical job resource scheduling features, historical process execution continuity features, and historical first optimization magnitude corresponding to a given location, forming a set of data. To facilitate subsequent regression analysis, the historical data fusion integrity features, historical job resource scheduling features, historical process execution continuity features, and historical first optimization magnitude are all converted from percentages to values ​​between 0 and 1. When converting the historical first optimization magnitude from percentages to values ​​between 0 and 1, normalization and quantization can also be performed based on a pre-set corresponding maximum optimization magnitude. For example, if the maximum optimization magnitude is 50%, and the optimization magnitude is 30%, the quantization is 30% / 50% = 0.6.

[0139] In one embodiment, a linear regression model is used to analyze multiple sets of data to obtain the correlation between historical data fusion integrity characteristics, historical job resource scheduling characteristics, historical process execution coherence characteristics, and historical first optimization magnitude. Using historical data fusion integrity characteristics, historical job resource scheduling characteristics, and historical process execution coherence characteristics as independent variables, and historical first optimization magnitude as the dependent variable, we obtain Y = a1×X1 + a2×X2 + a3×X3, where Y represents the historical first optimization magnitude, X1, X2, and X3 represent historical data fusion integrity characteristics, historical job resource scheduling characteristics, and historical process execution coherence characteristics, respectively, and a1, a2, and a3 are the coefficients corresponding to each characteristic. The specific values ​​of each characteristic coefficient are calculated by fitting the sample data using the least squares method; for example, a1 = 0.3, a2 = 0.4, and a3 = 0.2. The goodness of fit of the equation reflects the explanatory power of the model.

[0140] S413: Determine the weight coefficient of each sub-feature in the first type of features based on the first linear regression equation.

[0141] Specifically, the sub-features are data fusion integrity, job resource scheduling, and process execution coherence. Once the first linear regression equation Y=a1×X1+a2×X2+a3×X3 is determined, the weight coefficient of each sub-feature in the first category of features is also determined.

[0142] For example, a1=0.3 is the weight coefficient of the data fusion integrity feature, a2=0.4 is the weight coefficient of the job resource scheduling feature, and a3=0.2 is the weight coefficient of the process execution continuity feature.

[0143] S414: Determine the weight of each sub-feature in the first type of features based on the weight coefficient of each sub-feature.

[0144] Specifically, the weight coefficients of each sub-feature are normalized to obtain the weight of each sub-feature. The weight of the data fusion integrity feature = weight coefficient of data fusion integrity feature / (weight coefficient of data fusion integrity feature + weight coefficient of job resource scheduling feature + weight coefficient of process execution coherence feature), the weight of the job resource scheduling feature = weight coefficient of job resource scheduling feature / (weight coefficient of data fusion integrity feature + weight coefficient of job resource scheduling feature + weight coefficient of process execution coherence feature), and the weight of the process execution coherence feature = weight coefficient of process execution coherence feature / (weight coefficient of data fusion integrity feature + weight coefficient of job resource scheduling feature + weight coefficient of process execution coherence feature). This ensures that the weight values ​​are generated entirely by data, objectively reflecting the actual influence of each factor and avoiding biases caused by human settings.

[0145] For example, a1=0.3 is the weight coefficient of the data fusion integrity feature, a2=0.4 is the weight coefficient of the job resource scheduling feature, and a3=0.2 is the weight coefficient of the process execution continuity feature. After normalization, the weight of the data fusion integrity feature is 0.3 / (0.3+0.4+0.2)=0.33, the weight of the job resource scheduling feature is 0.4 / (0.3+0.4+0.2)=0.45, and the weight of the process execution continuity feature is 0.2 / (0.3+0.4+0.2)=0.22.

[0146] S415: Calculate the target correlation coefficient based on the weight coefficient of each sub-feature and the weight of each sub-feature.

[0147] Specifically, the target correlation coefficient is the sum of the product of the weight coefficient of each sub-feature and the weight of each sub-feature. When the target efficiency optimization region is a first-type efficiency optimization region, the target correlation coefficient is the first correlation coefficient. For example, if the weight of the data fusion integrity feature is 0.33, the weight of the job resource scheduling feature is 0.45, the weight of the process execution coherence feature is 0.22, the weight coefficient of the data fusion integrity feature is 0.3, the weight coefficient of the job resource scheduling feature is 0.4, and the weight coefficient of the process execution coherence feature is 0.2, then the first correlation coefficient = 0.3 × 0.33 + 0.4 × 0.45 + 0.2 × 0.22 = 0.323.

[0148] The target correlation coefficient represents the overall correlation strength between the first type of feature and the optimized amplitude.

[0149] In one embodiment, when the target efficiency optimization region is a second type of efficiency optimization region; step S4, based on a pre-set optimization database, determines the target historical optimization data according to the type of the target efficiency optimization region, and determines historical optimization measures and target correlation coefficients based on the target historical optimization data, including: S421: Based on the optimization database, determine the target historical optimization data according to the second type of features corresponding to the second type of efficiency optimization region. The target historical optimization data includes historical second type features, historical second optimization measures, and historical second optimization magnitude.

[0150] Specifically, the second type of features corresponding to the second type of efficiency optimization region are the delay repair timeliness feature and data transmission stability feature extracted in step S3 based on the multi-source data of the on-site operation at this location in step S1. Each location within the second type of efficiency optimization region has its corresponding second type of feature.

[0151] Specifically, at this point, based on the second type of characteristics of the second type of efficiency optimization region, similar data is searched from the optimization database according to the pre-set search rules, and similar historical data is determined as the target historical optimization data.

[0152] Specifically, the optimization database stores full-process data of all past power marketing field operations, including multi-source data of cross-modal field operations for each batch of operations, operation status verification results, first type of features, second type of features, first optimization measures, first optimization effect, and first optimization improvement.

[0153] Specifically, the search rules include pre-setting a search threshold range for each feature. Second-class features in the optimization database that fall within this threshold range are considered target second-class features, and the corresponding optimization data are considered target historical optimization data. The search threshold range for each feature fluctuates by a preset percentage based on the current feature value. For example, if a second-class feature at a certain location in the second-class efficiency optimization region includes a delay repair timeliness feature of 0.8 and a data transmission stability feature of 0.85, with a preset fluctuation percentage of 10% of the search threshold range, then the search threshold range for the delay repair timeliness feature is [0.72, 0.88], and the search threshold range for the data transmission stability feature is [0.77, 0.94]. Historical optimization data in the optimization database that simultaneously satisfies both the historical delay repair timeliness feature (located in [0.72, 0.88]) and the historical data transmission stability feature (located in [0.77, 0.94]) is considered target historical optimization data.

[0154] This range-based fuzzy matching method ensures the similarity of matching scenarios while avoiding missing valuable historical optimization cases due to minor differences in feature values. These target historical optimization data represent data from historical operations where there were deviations in the timeliness of delay repairs and the stability of data transmission, and where efficiency improvements were achieved through optimization measures—in other words, the portion of the second type of feature with room for efficiency optimization.

[0155] In one embodiment, the target historical optimization data establishes mapping relationships for location, multi-source on-site operation data, operation status verification results, historical second-type features, historical second-type optimization measures, historical second-type optimization effects, and historical second-type optimization magnitude. The target historical optimization data can be determined based on the historical second-type features, and is a complete dataset including the historical second-type features, historical second-type optimization measures, and historical second-type optimization magnitude. Once the target historical optimization data is determined, the historical second-type optimization measures and historical second-type optimization magnitude can be determined.

[0156] In one embodiment, the historical second optimization measure is a targeted measure that has already been implemented and can be used to generate corresponding optimization implementation guidance information later. For example, adjusting the transmission link bandwidth, optimizing the data retransmission mechanism, and adding node caching strategies.

[0157] In one embodiment, the historical second optimization magnitude is a quantification of the optimization effect achieved after implementing the historical second optimization measure. For example, the historical second optimization magnitude is 40%.

[0158] S422: Based on the multiple sets of historical second-class features and the historical second-optimization magnitude, the second linear regression equation is obtained.

[0159] In one embodiment, a set of historical second-type features and second-historical optimization magnitudes consists of historical delay repair timeliness features, historical data transmission stability features, and historical second-optimization magnitudes corresponding to a given location, forming a set of data. To facilitate subsequent regression analysis, the historical delay repair timeliness features, historical data transmission stability features, and historical second-optimization magnitudes are all converted from percentages to values ​​between 0 and 1. When converting all historical second-optimization magnitudes from percentages to values ​​between 0 and 1, normalization and quantization can also be performed based on a pre-set corresponding maximum optimization magnitude. For example, if the maximum optimization magnitude is 50%, and the optimization magnitude is 30%, the quantization is 30% / 50% = 0.6.

[0160] In one embodiment, a linear regression model is used to analyze multiple sets of data to obtain the correlation between historical delay repair timeliness characteristics, historical data transmission stability characteristics, and historical second optimization magnitude. Using historical delay repair timeliness characteristics and historical data transmission stability characteristics as independent variables, and historical second optimization magnitude as the dependent variable, we obtain Z = b1 × M1 + b2 × M2, where Z represents the historical second optimization magnitude, M1 and M2 represent historical delay repair timeliness characteristics and historical data transmission stability characteristics, respectively, and b1 and b2 are the coefficients corresponding to each characteristic. The specific values ​​of each characteristic coefficient are calculated by fitting the sample data using the least squares method. For example, b1 = 0.5, meaning that for every unit of optimization in delay repair timeliness, the transmission efficiency improvement increases by 0.5 units; b2 = 0.3, meaning that for every unit of optimization in data transmission stability, the transmission efficiency improvement increases by 0.3 units.

[0161] S423: Determine the weight coefficient of each sub-feature in the second type of features based on the second linear regression equation.

[0162] Specifically, the sub-features are the timeliness of delayed repair and the stability of data transmission. Once the second linear regression equation Z=b1×M1+b2×M2 is determined, the weight coefficient of each sub-feature in the second category of features is also determined.

[0163] For example, b1=0.5 is the weighting coefficient for the timeliness of delayed repair, and b2=0.3 is the weighting coefficient for the stability of data transmission.

[0164] S424: Determine the weight of each sub-feature in the second type of features based on the weight coefficient of each sub-feature.

[0165] Specifically, the weight coefficients of each sub-feature are normalized to obtain the weight of each sub-feature. The weight of the delayed repair timeliness feature = weight coefficient of delayed repair timeliness feature / (weight coefficient of delayed repair timeliness feature + weight coefficient of data transmission stability feature), and the weight of the data transmission stability feature = weight coefficient of data transmission stability feature / (weight coefficient of delayed repair timeliness feature + weight coefficient of data transmission stability feature). This ensures that the weight values ​​are generated entirely by data, objectively reflecting the actual impact of each factor and avoiding biases caused by human intervention.

[0166] For example, b1=0.5 is the weight coefficient of the delay repair timeliness feature, and b2=0.3 is the weight coefficient of the data transmission stability feature. After normalization, the weight of the delay repair timeliness feature is 0.5 / (0.5+0.3)=0.63, and the weight of the data transmission stability feature is 0.3 / (0.5+0.3)=0.37.

[0167] S425: Calculate the target correlation coefficient based on the weight coefficient of each sub-feature in the second type of features and the weight of each sub-feature.

[0168] Specifically, the target correlation coefficient is the sum of the product of the weight coefficient of each sub-feature and the weight of each sub-feature. When the target efficiency optimization region is a second type of efficiency optimization region, the target correlation coefficient is the second correlation coefficient. For example, if the weight of the delay repair timeliness feature is 0.63, the weight of the data transmission stability feature is 0.37, the weight coefficient of the delay repair timeliness feature is 0.5, and the weight coefficient of the data transmission stability feature is 0.3, then the second correlation coefficient = 0.63 × 0.5 + 0.37 × 0.3 = 0.426.

[0169] At this point, the target correlation coefficient represents the overall correlation strength between the second type of feature and the optimized amplitude.

[0170] S5: Based on the pre-set optimization generation model, generate optimization implementation guidance information according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0171] In one embodiment, when the target efficiency optimization region is a first-type efficiency optimization region; step S5 generates optimization implementation guidance information based on a pre-set optimization generation model, according to the multi-source data of on-site operations corresponding to the target efficiency optimization region, the historical optimization measures, and the target correlation coefficient, including: S511: Determine the quantification value of the first influence range based on the multi-source data of on-site operations in the target efficiency optimization area.

[0172] Specifically, the first impact range quantification value refers to a specific numerical representation of the number of devices, personnel, and work batches affected by a problem in the area's operations. The first impact range quantification value is obtained by weighted summation of these three factors. For example, if a problem in the meter installation work area affects the installation progress of 300 meters, 30 people, and 8 work batches, and the preset weights for the three factors are 0.4, 0.3, and 0.3 respectively, then the first impact range quantification value = 0.4 × 300 + 0.3 × 30 + 0.3 × 8 = 131.4.

[0173] Specifically, the method includes: determining all equipment numbers in the target efficiency optimization area based on multi-source data of on-site operations in the target efficiency optimization area, and determining the number of equipment based on the equipment numbers; finding all operation batch numbers based on the equipment numbers in the target efficiency optimization area to obtain the number of operation batches; and finding personnel information based on the operation batch numbers in the target efficiency optimization area to determine the number of personnel based on the personnel information.

[0174] Specifically, the equipment number in the work log is linked to the work batch number. The work batch number can be obtained from the equipment number, and the number of work batches can be determined from the work batch number. One work batch number corresponds to one work batch.

[0175] Specifically, the work batch number in the work log is linked to the personnel schedule. The personnel schedule can be obtained based on the work batch number, the personnel information can be obtained based on the personnel schedule, and the number of personnel can be determined based on the personnel information.

[0176] S512: Calculate the optimization magnitude of each feature based on the first type of feature of the target efficiency optimization region and the benchmark values ​​of each feature in the optimization database.

[0177] Specifically, the optimization magnitude of each feature is obtained by calculating the deviation rate between each sub-feature in the first type of feature and the benchmark value of each feature in the corresponding optimization database.

[0178] Specifically, the deviation rate of the data fusion integrity feature = (data fusion integrity feature benchmark value - data fusion integrity feature of the target efficiency optimization area) / data fusion integrity feature benchmark value × 100%, and the optimization range of the data fusion integrity feature = deviation rate of the data fusion integrity feature × weight of the data fusion integrity feature.

[0179] Specifically, the deviation rate of the task resource scheduling feature = (benchmark value of task resource scheduling feature - task resource scheduling feature of the target efficiency optimization area) / benchmark value of task resource scheduling feature × 100%, and the optimization range of the task resource scheduling feature = deviation rate of task resource scheduling feature × weight of task resource scheduling feature.

[0180] Specifically, the deviation rate of process execution coherence characteristics = (benchmark value of process execution coherence characteristics - process execution coherence characteristics of the target efficiency optimization area) / benchmark value of process execution coherence characteristics × 100%, and the optimization range of process execution coherence characteristics = deviation rate of process execution coherence characteristics × weight of process execution coherence characteristics.

[0181] For example, the data fusion integrity feature of the target efficiency optimization area is 0.7, and the baseline value of the data fusion integrity feature in the optimization database is 0.875; the job resource scheduling feature of the target efficiency optimization area is 0.85, and the baseline value of the job resource scheduling feature in the optimization database is 1.0; the process execution coherence feature of the target efficiency optimization area is 0.9, and the baseline value of the process execution coherence feature in the optimization database is 1.0. Therefore, the optimization margin of the data fusion integrity feature = (0.875 - 0.7) ÷ 0.875 × 100% × 0.33 = 6.6%; the optimization margin of the job resource scheduling feature = (1.0 - 0.85) ÷ 1.0 × 100% × 0.45 = 6.75%; and the optimization margin of the process execution coherence feature = (1.0 - 0.9) ÷ 1.0 × 100% × 0.22 = 2.2%.

[0182] S513: Calculate the process optimization quantization value based on the target correlation coefficient, the first influence range quantization value, and the optimization magnitude of each feature.

[0183] Specifically, when the target efficiency optimization region is a first type of efficiency optimization region, the target correlation coefficient is the first correlation coefficient. The process optimization quantification value = the average optimization magnitude of each feature × the first correlation coefficient × the first influence range quantification value.

[0184] For example, if the optimization margin for data fusion integrity is 6.6%, the optimization margin for job resource scheduling is 6.75%, the optimization margin for process execution coherence is 2.2%, the first correlation coefficient is 0.323, and the quantified value for the first influence range is 131.4, then the quantified value for process optimization = (6.6% + 6.75% + 2.2%) / 3 × 0.323 × 131.4 = 2.2. The quantified value for process optimization represents the potential scale of process optimization effects in that region.

[0185] In this application, the first correlation coefficient is a comprehensive quantitative index obtained by linear regression fitting based on the first type of feature, which integrates the correlation strength information between each feature and the optimization effect. When calculating the optimization magnitude of each feature in step S512, the weight of each feature participates in the calculation of the optimization magnitude of a single feature, reflecting the contribution ratio of a single type of feature to the overall optimization effect. When calculating the process optimization quantitative value in step S513, the first correlation coefficient is used as a global correction coefficient to perform overall scaling and calibration on the mean of the optimization magnitude of all features. The first correlation coefficient carries the comprehensive correlation level of all features at the current point, which can correct the problem that the mean of the magnitude of a single feature cannot reflect the severity of the overall defect of the point. When calculating a single feature, the subdivided weights of each feature are used separately, and when quantifying globally, the integrated comprehensive correlation coefficient is used to distinguish the magnitude of the influence of a single feature. The latter represents the severity of the overall data fusion defect of the point. The dual application can simultaneously take into account the refined measurement of single features and the overall optimization potential assessment of the entire region, so that the final calculated process optimization quantitative value retains the differentiated contribution of each subdivided feature and fits the overall fault correlation degree of the point, improving the accuracy and fit of the quantitative results.

[0186] S514: Based on the optimization database, and according to the historical optimization measures, determine the historical process simplification index and its corresponding historical process efficiency optimization range, and determine the first process optimization coefficient based on multiple sets of historical process simplification indices and their corresponding historical process efficiency optimization ranges.

[0187] Specifically, the historical process simplification index in the optimization database refers to the degree of process simplification in historical optimization measures. For example, if 10 steps are reduced to 7, the historical process simplification index is 0.7, corresponding to a historical process efficiency improvement of 35%. The historical process simplification index and the historical process efficiency improvement are recorded in the optimization database and have a mapping relationship with historical optimization measures.

[0188] Specifically, linear regression analysis was conducted using multiple sets of historical process simplification indices and historical process efficiency optimization magnitudes to obtain a regression equation with the process simplification index as the independent variable and the process efficiency optimization magnitude as the dependent variable. The first process optimization coefficient is the regression coefficient of the process simplification index. For example, the analysis showed that the first process optimization coefficient was 0.5, meaning that for every 0.1 increase in the historical process simplification index, the historical process efficiency optimization magnitude increased by 5%.

[0189] S515: Calculate the target process simplification index based on the process optimization quantification value and the first process optimization coefficient.

[0190] Specifically, the target process simplification index is the product of the process optimization quantification value and the first process optimization coefficient. It is a quantitative indicator that represents the degree to which the process can be simplified within the current first-class efficiency optimization area. The higher the index value, the greater the space for simplification.

[0191] For example, if the optimization coefficient of the first process is 0.5 and the quantitative value of process optimization is 2.2, then the target process simplification index = 2.2 × 0.5 = 1.1.

[0192] S516: Based on the optimization database, and according to the historical optimization measures, determine the historical resource configuration data and its corresponding historical resource configuration optimization index, and determine the first resource optimization coefficient based on multiple sets of historical resource configuration data and their corresponding historical resource configuration optimization indices.

[0193] Specifically, optimizing historical resource allocation data in the database refers to the allocation ratio of resources (such as personnel and tools) in historical operations. The historical resource allocation optimization index is a quantitative value of the rationality of resource allocation (for example, an increase from 0.6 to 0.9 represents an improvement in resource utilization). Historical resource allocation data and historical resource allocation optimization index are recorded in the optimization database and have a mapping relationship with historical optimization measures.

[0194] Specifically, linear regression analysis is performed using multiple sets of historical resource allocation data and historical resource allocation optimization indices to obtain a regression equation with resource allocation data as the independent variable and the resource allocation optimization index as the dependent variable. The first resource optimization coefficient is the regression coefficient of the historical resource allocation data. For example, the analysis shows that the first resource optimization coefficient is 0.4, which means that for every 10% optimization of resource allocation data, the resource allocation optimization index can be increased by 0.4.

[0195] S517: Calculate the target resource allocation optimization index based on the process optimization quantification value and the first resource optimization coefficient.

[0196] Specifically, the target resource allocation optimization index is the product of the process optimization quantification value and the first resource optimization coefficient. It represents the degree of improvement in the rationality of resource allocation within the current first-type efficiency optimization area. The higher the index value, the greater the potential for improvement.

[0197] For example, if the first resource optimization coefficient is 0.4 and the process optimization quantification value is 2.2, then the target resource allocation optimization index = 2.2 × 0.4 = 0.88.

[0198] S518: Based on the process simplification baseline value and resource configuration baseline value in the optimization database, determine whether it is feasible according to the target process simplification index and the target resource configuration optimization index. When it is determined that it is feasible, generate the optimization implementation guidance information according to the historical optimization measures.

[0199] Specifically, when the target process simplification index is greater than or equal to the process simplification benchmark value, and the target resource allocation optimization index is greater than or equal to the resource allocation benchmark value, it is determined to be feasible; when the target process simplification index is less than the process simplification benchmark value, or the target resource allocation optimization index is less than the resource allocation benchmark value, it is determined to be unfeasible.

[0200] Specifically, when determining feasibility, based on a pre-set prompt template, optimization implementation guidance information is generated according to historical optimization measures, the target process simplification index, and the target resource allocation optimization index. For example, the prompt template might read: "Optimization measures are: ...; Target process simplification index is: ...; Target resource allocation optimization index is: ...; Both the process simplification plan and the resource allocation plan meet the standards and can be implemented directly." The historical optimization measures, target process simplification index, and target resource allocation optimization index are then filled into the corresponding positions to generate the optimization implementation guidance information.

[0201] In one embodiment, when the target efficiency optimization region is a second type of efficiency optimization region; step S5 generates optimization implementation guidance information based on a pre-set optimization generation model, according to the multi-source data of on-site operations corresponding to the target efficiency optimization region, the historical optimization measures, and the target correlation coefficient, including: S521: Determine the quantification value of the second influence range based on the multi-source data of on-site operations in the target efficiency optimization area.

[0202] Specifically, the second impact range quantification value refers to a specific numerical representation of the number of data transmission terminals affected and the amount of data transmitted synchronously in the region due to operational problems. The second impact range quantification value is obtained by weighted summation of the number of transmission terminals and the amount of data transmitted synchronously. For example, if a data transmission problem at a certain location in a second-category efficiency optimization region affects 5 data transmission terminals and 200 synchronously transmitted data, with preset weights of 0.6 and 0.4 for the number of data transmission terminals and the amount of data transmitted synchronously, respectively, then the second impact range quantification value = 5 × 0.6 + 200 × 0.4 = 83.

[0203] Specifically, the method includes: determining the IP addresses of devices that generate delay alarm signals in the target efficiency optimization area based on multi-source data of field operations in the target efficiency optimization area; determining the number of data transmission terminals based on the device IP addresses; and determining the number of data to be initiated for synchronous transmission based on the device IP addresses.

[0204] For example, a data transmission problem at a certain location in the second type of efficiency optimization area affects the synchronization efficiency of 5 data transmission terminals and 200 data entries, and the quantification value of the second impact range is set to 200.

[0205] S522: Calculate the optimization magnitude of each feature based on the second type of feature of the target efficiency optimization region and the benchmark values ​​of each feature in the optimization database.

[0206] Specifically, the optimization magnitude of each feature is obtained by calculating the deviation rate between each sub-feature in the second type of feature and the benchmark value of each feature in the corresponding optimization database.

[0207] Specifically, the deviation rate of delayed repair timeliness feature = (delayed repair timeliness feature benchmark value - delayed repair timeliness feature of target efficiency optimization area) / delayed repair timeliness feature benchmark value × 100%, and the optimization range of delayed repair timeliness feature = deviation rate of delayed repair timeliness feature × weight of delayed repair timeliness feature.

[0208] Specifically, the deviation rate of data transmission stability feature = (data transmission stability feature baseline value - data transmission stability feature of target efficiency optimization area) / data transmission stability feature baseline value × 100%, and the optimization range of data transmission stability feature = deviation rate of data transmission stability feature × weight of data transmission stability feature.

[0209] For example, the latency repair timeliness characteristic of the target efficiency optimization region is 0.7, and the baseline value of the latency repair timeliness characteristic in the optimization database is 0.875; the data transmission stability characteristic of the target efficiency optimization region is 0.85, and the baseline value of the data transmission stability characteristic in the optimization database is 1.0. Therefore, the optimization margin of the latency repair timeliness characteristic = (0.875 - 0.7) ÷ 0.875 × 100% × 0.63 = 12.6%; the optimization margin of the data transmission stability characteristic = (1.0 - 0.85) ÷ 1.0 × 100% × 0.37 = 5.55%.

[0210] S523: Calculate the transmission optimization potential quantification value based on the target correlation coefficient, the second influence range quantification value, and the optimization magnitude of each feature.

[0211] Specifically, when the target efficiency optimization region is a second type of efficiency optimization region, the target correlation coefficient is the second correlation coefficient. The transmission optimization potential quantification value = the average optimization magnitude of each feature × the second correlation coefficient × the second influence range quantification value.

[0212] For example, if the optimization margin for delay repair timeliness is 12.6%, the optimization margin for data transmission stability is 5.55%, the second correlation coefficient is 0.426, and the quantified value for the second influence range is 83, then the quantified value for transmission optimization potential is (12.6% + 5.55%) / 2 × 0.426 × 83 = 3.2. The quantified value for transmission optimization potential represents the potential scale of transmission optimization effects in this region.

[0213] S524: Obtain the pre-processing transmission status data and the post-processing transmission status data of the target efficiency optimization region, obtain the transmission status benchmark value based on the optimization database, and obtain the pre-processing transmission parameter value according to the pre-processing transmission status data, the post-processing transmission status data and the transmission status benchmark value.

[0214] Specifically, preprocessing refers to optimization measures that staff can immediately take on-site. For example, when expanding the capacity of a transmission link, the transmission status data of this link differs before and after the expansion.

[0215] Specifically, transmission status data includes key performance indicators of the transmission link, such as bandwidth utilization, transmission latency, and packet loss rate.

[0216] Specifically, the method includes: determining the original bandwidth configuration information of each transmission link based on a pre-set link configuration management database; when the collected bandwidth occupancy rate and transmission rate are compared with the original bandwidth configuration information, if it is found that the bandwidth occupancy rate of a link has exceeded the original bandwidth occupancy rate configuration value, and the transmission delay time and packet loss rate decrease within a preset time period in accordance with the preset decrease range, then it is determined that the transmission link has completed preprocessing.

[0217] Specifically, the preprocessing transmission parameter value quantifies the effectiveness of the current transmission repair measures. Preprocessing transmission parameter value = Bandwidth occupancy rate weight × (1 - (Preprocessed bandwidth occupancy rate - Baseline bandwidth occupancy rate) / (Preprocessed bandwidth occupancy rate - Baseline bandwidth occupancy rate)) + Transmission delay time weight × (1 - (Preprocessed transmission delay time - Baseline transmission delay time) / (Preprocessed transmission delay time - Baseline transmission delay time)) + Packet loss rate weight × (1 - (Preprocessed packet loss rate - Baseline packet loss rate) / (Preprocessed packet loss rate - Baseline packet loss rate)), where 1 represents the degree of complete repair. For example, if the weight of the bandwidth utilization rate is 0.4, the weight of the transmission delay time is 0.3, the weight of the packet loss rate is 0.3, the bandwidth utilization rate before preprocessing is 99%, the baseline value of bandwidth utilization is 80%, and the bandwidth utilization rate after preprocessing is 81%, the transmission delay time before preprocessing is 150ms, the baseline value of transmission delay time is 100ms, and the transmission delay time after preprocessing is 110ms, the packet loss rate before preprocessing is 5%, the baseline value of packet loss rate is 1%, and the packet loss rate after preprocessing is 1.2%, then the preprocessing transmission parameter value = 0.4 × (1 - (81% - 80%) / (99% - 80%)) + 0.3 × (1 - (110 - 100) / (150 - 100)) + 0.3 × (1 - (1.2% - 1%) / (5% - 1%)) = 0.9.

[0218] When the preprocessed transmission status data is less than the transmission status reference value, the preprocessed transmission status data is taken as the transmission status reference value when calculating the preprocessed transmission parameter value.

[0219] S525: Based on the optimization database, and according to the historical optimization measures, determine the historical transmission efficiency improvement rate and its corresponding historical operation efficiency improvement rate, and determine the correlation optimization coefficient based on multiple sets of historical transmission efficiency improvement rates and their corresponding historical operation efficiency improvement rates.

[0220] Specifically, the historical transmission efficiency improvement is a precise quantification of the optimization effect of a single indicator. It refers to the rate of change of a specific technical indicator before and after optimization for a single transmission link or a single data record, and is a precise measure of the optimization effect of transmission technology. For example, by obtaining the key indicators of transmission delay time and transmission success rate for each transmission link before and after historical optimization measures, the historical transmission efficiency improvement of each link is calculated according to a pre-set calculation formula. For example, historical transmission efficiency improvement = transmission delay time weight × (1 - transmission delay time after optimization / transmission delay time before optimization) + transmission success rate weight × (1 - transmission success rate after optimization / transmission success rate before optimization).

[0221] Specifically, the historical operational efficiency improvement is a comprehensive assessment of the overall operational efficiency, referring to the rate of change in operational efficiency before and after optimization throughout the entire operational process. It is a comprehensive measure of the operational efficiency brought about by transmission optimization.

[0222] Specifically, the steps for determining the magnitude of historical operation efficiency improvement are as follows: The target optimization data includes two sets of historical field operation multi-source data under the same device IP address and the same batch of optimization events, before and after optimization. This encompasses three types of timestamped raw data: video operation data, equipment operation data, and operation log data. Two-layer verification (independent verification and cross-verification) is performed on the two sets of field operation multi-source data using verification standards. After verification, the average operation time T before optimization is extracted from the two sets of historical field operation multi-source data. 前 Optimized average operation time T for the entire process 后 The average processing time H for anomaly points before optimization 前 Optimized average processing time H for abnormal locations 后 Before optimization, the time consumption of multi-source data alignment was A. 前 Optimized multi-source data alignment time A 后 Three sets of core efficiency quantification indicators; extracting the timeliness characteristics of historical delay repair and the stability characteristics of historical data transmission before optimization; based on the pre-set mapping relationship between features and business weights, determining the business weight corresponding to each core efficiency quantification indicator, including: the business weight W1 for the average operation time of the entire process, the business weight W2 for the average handling time of anomalies, and the business weight W3 for the time spent aligning multi-source data, and W1+W2+W3=1; calculating the change rate of the average operation time of the entire process, the change rate of the average handling time of anomalies, and the change rate of the time spent aligning multi-source data, the change rate of the average operation time of the entire process = (T 前 -T 后 )÷T 前 ×100%, the rate of change in average handling time for abnormal locations = (H 前 -H 后 )÷H前 ×100%, the rate of change in multi-source data alignment time = (A 前- A 后 )÷A 前 ×100%, and then obtain the historical operation efficiency improvement value by weighted summation. Historical operation efficiency improvement value = average operation time change rate of the whole process ×W1 + average handling time change rate of abnormal points ×W2 + multi-source data alignment time change rate ×W3.

[0223] Specifically, the historical improvement in transmission efficiency and the historical improvement in operation efficiency are recorded one-to-one in the optimization database, and there is a mapping relationship between them and the historical optimization measures.

[0224] Specifically, linear regression analysis was conducted using multiple sets of historical transmission efficiency improvement rates and historical operation efficiency improvement rates to obtain a regression equation with transmission efficiency improvement rate as the independent variable and operation efficiency improvement rate as the dependent variable. The correlation optimization coefficient is the regression coefficient for transmission efficiency improvement rate. For example, the analysis yielded a correlation optimization coefficient of 0.5, meaning that for every 0.1 increase in transmission efficiency improvement rate, the operation efficiency improvement rate increases by 5%.

[0225] S526: Calculate the transmission efficiency threshold based on the quantified value of the transmission optimization potential, the preprocessed transmission parameter value, and the correlation optimization coefficient.

[0226] Specifically, the transmission efficiency threshold is the product of the quantified value of the transmission optimization potential, the value of the preprocessed transmission parameters, and the correlation optimization coefficient. For example, if the quantified value of the transmission optimization potential is 3.2, the value of the preprocessed transmission parameters is 0.9, and the correlation optimization coefficient is 0.5, then the transmission efficiency threshold = 3.2 × 0.9 × 0.5 = 1.44.

[0227] S527: Based on the transmission efficiency qualification threshold in the optimization database, determine whether it is feasible according to the transmission efficiency qualification threshold. When it is determined that it is feasible, generate the optimization implementation guidance information according to the historical optimization measures.

[0228] Specifically, if the transmission efficiency compliance threshold is greater than or equal to the transmission efficiency qualification threshold, it is determined that the procedure is feasible; if the transmission efficiency compliance threshold is less than the transmission efficiency qualification threshold, it is determined that the procedure is not feasible.

[0229] Specifically, when determining feasibility, based on a pre-set prompt template, optimization implementation guidance information is generated according to historical optimization measures and transmission efficiency achievement thresholds. For example, the prompt template might read, "Optimization measures are: ...; Transmission efficiency achievement thresholds are: ...; Transmission efficiency has been achieved, it is recommended to implement the plan." The historical optimization measures and transmission efficiency achievement thresholds are then filled into the corresponding positions to generate the optimization implementation guidance information.

[0230] As can be seen, in the above scheme, this application achieves cross-modal data integration and correlation analysis by independently verifying and cross-verifying multi-source data of on-site operations, which can improve the accuracy of efficiency optimization analysis. By further analyzing the verification results, the target efficiency optimization area is screened out, and targeted optimization measures and optimization implementation guidance information are generated. Different adjustment methods are matched to problems in different areas, thereby improving resource utilization and optimization effect.

[0231] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0232] In one embodiment, a device for optimizing the efficiency of on-site inspections in electricity marketing is provided, which corresponds one-to-one with the method for optimizing the efficiency of on-site inspections in electricity marketing described in the above embodiments. For example... Figure 3 As shown, the power marketing field operation inspection efficiency optimization device includes an acquisition module 101, a verification module 102, a region filtering module 103, a query module 104, and a generation module 105. Detailed descriptions of each functional module are as follows: The acquisition module 101 is used to acquire multi-source data of field operations at multiple locations; The verification module 102 is used to perform independent and cross-verification on the multi-source data of the on-site operation at each location based on a pre-set verification standard, and generate an operation status verification result. The region filtering module 103 is used to filter out target efficiency optimization regions from multiple locations and determine the type of the target efficiency optimization regions based on pre-set region type determination rules, according to the multi-source data of the on-site operations at multiple locations and their corresponding operation status verification results; The query module 104 is used to determine the target historical optimization data based on the pre-set optimization database and the type of the target efficiency optimization region, and to determine the historical optimization measures and target correlation coefficient based on the target historical optimization data; The generation module 105 is used to generate optimization implementation guidance information based on a pre-set optimization generation model, according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0233] Specifically, the verification module 102 is also used to perform spatiotemporal alignment of the video operation data, the equipment operation data, and the operation log data based on a pre-set alignment model; Based on the first identification model and video operation standard in the verification standard, the video clothing, video dialogue, video actions, and video flow in the video operation data are identified according to the first identification model, and the video clothing, video dialogue, and video actions are independently verified according to the video operation standard. Based on the second identification model and the device parameter standards in the verification standard, various types of device parameters in the device operation data are identified according to the second identification model, and the various types of device parameters and parameter volatility are independently verified according to the device parameter standards. Based on the third identification model and the work log standard in the verification standard, the text fields, text actions, and operation processes in the work log data are identified according to the third identification model, and the text fields and text actions are independently verified according to the work log standard. Based on the cross-validation criteria in the verification standard, the consistency of the video process and the operation process is verified, the consistency of the video actions and the various device parameters is verified, and the consistency of the text fields and the various device parameters is verified. The job status verification result is generated based on the independent verification results and the cross-verification results.

[0234] Specifically, the region filtering module 103 is also used to calculate the data association strength of each location based on the association strength calculation rules and the operation status verification results of each type of data at each location; Based on the aforementioned association strength assessment threshold determination rule, the association strength assessment threshold is determined according to the data association strength at multiple locations; Based on the first optimization region division rule, the locations where the data association strength is greater than the association strength evaluation threshold are selected as the first optimization region; Based on the priority calculation rules, the priority coefficient of each location is calculated according to the operation status verification results of the multi-source data of the on-site operation at each location in the first optimization area, and the multiple locations are sorted by priority. Based on the priority threshold determination rule, the priority threshold is determined according to the priority coefficient of each position in the first optimization region; Based on the second optimization region division rule, positions with priority coefficients greater than the priority threshold are selected as the second optimization region; Based on the region classification rules, first-type features and second-type features are extracted from the multi-source data of on-site operations corresponding to each location in the second optimization region. The locations where the first-type features or the second-type features are extracted are selected as target efficiency optimization regions. The target efficiency optimization region type where the first-type features are extracted is the first-type efficiency optimization region, and the target efficiency optimization region type where the second-type features are extracted is the second-type efficiency optimization region.

[0235] Specifically, the query module 104 is also used to determine target historical optimization data based on the optimization database and according to the first type of features corresponding to the first type of efficiency optimization region. The target historical optimization data includes historical first type of features, historical first optimization measures, and historical first optimization magnitude. Based on the multiple sets of historical first-class features and the historical first optimization magnitude, the first linear regression equation is obtained; Based on the first linear regression equation, determine the weight coefficient of each sub-feature in the first type of features; The weight of each sub-feature in the first type of features is determined based on the weight coefficient of each sub-feature. The target correlation coefficient is calculated based on the weight coefficient of each sub-feature and the weight of each sub-feature.

[0236] Specifically, the query module 104 is also used to determine target historical optimization data based on the optimization database and according to the second type of features corresponding to the second type of efficiency optimization region. The target historical optimization data includes historical second type features, historical second optimization measures, and historical second optimization magnitude. Based on the multiple sets of historical second-type features and the second historical optimization magnitude, the second linear regression equation is obtained; Based on the second linear regression equation, determine the weight coefficient of each sub-feature in the second type of features; The weight of each sub-feature in the second type of features is determined based on the weight coefficient of each sub-feature. The target correlation coefficient is calculated based on the weight coefficient of each sub-feature in the second type of features and the weight of each sub-feature.

[0237] Specifically, the generation module 105 is also used to determine the first influence range quantification value based on the multi-source data of the on-site operations in the target efficiency optimization area; Based on the first type of feature of the target efficiency optimization region and the benchmark values ​​of each feature in the optimization database, the optimization magnitude of each feature is calculated; The process optimization quantification value is calculated based on the target correlation coefficient, the first influence range quantification value, and the optimization magnitude of each feature; Based on the optimization database, and according to the historical optimization measures, the historical process simplification index and its corresponding historical process efficiency optimization range are determined, and the first process optimization coefficient is determined based on multiple sets of historical process simplification indices and their corresponding historical process efficiency optimization ranges. Calculate the target process simplification index based on the process optimization quantification value and the first process optimization coefficient; Based on the optimization database, and according to the historical optimization measures, historical resource allocation data and their corresponding historical resource allocation optimization indices are determined, and a first resource optimization coefficient is determined based on multiple sets of historical resource allocation data and their corresponding historical resource allocation optimization indices. Calculate the target resource allocation optimization index based on the process optimization quantification value and the first resource optimization coefficient; Based on the process simplification baseline value and resource configuration baseline value in the optimization database, and according to the target process simplification index and the target resource configuration optimization index, it is determined whether it is feasible. When it is determined to be feasible, the optimization implementation guidance information is generated based on the historical optimization measures.

[0238] Specifically, the generation module 105 is also used to determine the second influence range quantification value based on the multi-source data of the on-site operations in the target efficiency optimization area; Based on the second type of features of the target efficiency optimization region and the baseline values ​​of each feature in the optimization database, the optimization magnitude of each feature is calculated; Based on the target correlation coefficient, the second influence range quantification value, and the optimization magnitude of each feature, calculate the transmission optimization potential quantification value; Obtain the pre-processing transmission status data and the post-processing transmission status data of the target efficiency optimization region, obtain the transmission status benchmark value based on the optimization database, and obtain the pre-processing transmission parameter value according to the pre-processing transmission status data, the post-processing transmission status data and the transmission status benchmark value. Based on the optimization database, and according to the historical optimization measures, the historical transmission efficiency improvement rate and its corresponding historical operation efficiency improvement rate are determined, and the correlation optimization coefficient is determined based on multiple sets of historical transmission efficiency improvement rates and their corresponding historical operation efficiency improvement rates. Based on the quantitative value of transmission optimization potential, the value of preprocessed transmission parameters, and the correlation optimization coefficient, calculate the threshold for achieving transmission efficiency. Based on the transmission efficiency qualification threshold in the optimization database, it is determined whether the implementation is feasible. If it is determined to be feasible, the optimization implementation guidance information is generated based on the historical optimization measures.

[0239] This invention provides a device for optimizing the efficiency of on-site operations in power marketing. By independently verifying and cross-verifying multi-source data from on-site operations, it achieves the integration and correlation analysis of cross-modal data, which can improve the accuracy of efficiency optimization analysis. Through further analysis of the verification results, it identifies target efficiency optimization areas and generates targeted optimization measures and implementation guidance information. This enables different adjustment methods to be matched to problems in different areas, thereby improving resource utilization and optimization effect.

[0240] Specific limitations regarding the efficiency optimization device for on-site power marketing operations can be found in the limitations of the efficiency optimization method for on-site power marketing operations described above, and will not be repeated here. Each module in the aforementioned efficiency optimization device for on-site power marketing operations can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0241] In one embodiment, a computer device is provided, which may be a device terminal, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a power marketing field operation inspection efficiency optimization method on the device side.

[0242] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Acquire multi-source data on field operations at multiple locations; Based on pre-set verification standards, the multi-source data of the on-site operations at each location are independently and cross-verified to generate operation status verification results. Based on pre-set area type determination rules, and according to the multi-source data of field operations at multiple locations and their corresponding operation status verification results, target efficiency optimization areas are selected from multiple locations and the type of the target efficiency optimization areas is determined. Based on a pre-set optimization database, historical optimization data of the target is determined according to the type of the target efficiency optimization region, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data of the target. Based on a pre-set optimization generation model, optimization implementation guidance information is generated according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0243] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Acquire multi-source data on field operations at multiple locations; Based on pre-set verification standards, the multi-source data of the on-site operations at each location are independently and cross-verified to generate operation status verification results. Based on pre-set area type determination rules, and according to the multi-source data of field operations at multiple locations and their corresponding operation status verification results, target efficiency optimization areas are selected from multiple locations and the type of the target efficiency optimization areas is determined. Based on a pre-set optimization database, historical optimization data of the target is determined according to the type of the target efficiency optimization region, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data of the target. Based on a pre-set optimization generation model, optimization implementation guidance information is generated according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

[0244] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and device side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0245] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0246] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0247] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0248] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for optimizing the efficiency of on-site inspection in electricity marketing, characterized in that, include: Acquire multi-source data on field operations at multiple locations; Based on pre-set verification standards, the multi-source data of the on-site operations at each location are independently and cross-verified to generate operation status verification results. Based on pre-set area type determination rules, and according to the multi-source data of field operations at multiple locations and their corresponding operation status verification results, target efficiency optimization areas are selected from multiple locations and the type of the target efficiency optimization areas is determined. Based on a pre-set optimization database, historical optimization data of the target is determined according to the type of the target efficiency optimization region, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data of the target. Based on a pre-set optimization generation model, optimization implementation guidance information is generated according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

2. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 1, characterized in that, The multi-source data for on-site operations includes video operation data, equipment operation data, and operation log data; Based on pre-set verification standards, the multi-source data of the on-site operations at each location are verified to generate operation status verification results, including: Based on a pre-set preprocessing model, the video operation data, the equipment operation data, and the operation log data are subjected to spatiotemporal alignment and structured processing; Based on the first identification model and video operation standard in the verification standard, the video clothing, video dialogue, video actions, and video flow in the video operation data are identified according to the first identification model, and the video clothing, video dialogue, and video actions are independently verified according to the video operation standard. Based on the second identification model and the device parameter standards in the verification standard, various types of device parameters in the device operation data are identified according to the second identification model, and the various types of device parameters and parameter volatility are independently verified according to the device parameter standards. Based on the third identification model and the work log standard in the verification standard, the text fields, text actions, and operation processes in the work log data are identified according to the third identification model, and the text fields and text actions are independently verified according to the work log standard. Based on the cross-validation criteria in the verification standard, the consistency of the video process and the operation process is verified, the consistency of the video actions and the various device parameters is verified, and the consistency of the text fields and the various device parameters is verified. The job status verification result is generated based on the independent verification results and the cross-verification results.

3. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 1, characterized in that, The region type determination rules include association strength calculation rules, association strength evaluation threshold determination rules, first optimized region division rules, priority calculation rules, priority threshold determination rules, second optimized region division rules, and region classification rules. Based on pre-set area type determination rules, and according to the multi-source data of on-site operations at multiple locations and their corresponding operation status verification results, target efficiency optimization areas are selected from multiple locations, and the type of the target efficiency optimization area is determined, including: Based on the association strength calculation rules, the data association strength of each location is calculated according to the job status verification results of each type of data at each location; Based on the aforementioned association strength assessment threshold determination rule, the association strength assessment threshold is determined according to the data association strength at multiple locations; Based on the first optimization region division rule, the locations where the data association strength is greater than the association strength evaluation threshold are selected as the first optimization region; Based on the priority calculation rules, the priority coefficient of each location is calculated according to the operation status verification results of the multi-source data of the on-site operation at each location in the first optimization area, and the multiple locations are sorted by priority. Based on the priority threshold determination rule, the priority threshold is determined according to the priority coefficient of each position in the first optimization region; Based on the second optimization region division rule, positions with priority coefficients greater than the priority threshold are selected as the second optimization region; Based on the region classification rules, first-type features and second-type features are extracted from the multi-source data of on-site operations corresponding to each location in the second optimization region. The locations where the first-type features or the second-type features are extracted are selected as target efficiency optimization regions. The target efficiency optimization region type where the first-type features are extracted is the first-type efficiency optimization region, and the target efficiency optimization region type where the second-type features are extracted is the second-type efficiency optimization region.

4. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 1, characterized in that, The first category of features includes data fusion integrity features, job resource scheduling features, and process execution coherence features; The second category of features includes features related to delay repair timeliness and data transmission stability.

5. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 1, characterized in that, When the target efficiency optimization region is a first type of efficiency optimization region; Based on a pre-set optimization database, and according to the type of the target efficiency optimization region, historical optimization data is determined, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data, including: Based on the optimization database, target historical optimization data is determined according to the first type of features corresponding to the first type of efficiency optimization region. The target historical optimization data includes historical first type features, historical first optimization measures, and historical first optimization magnitude. Based on the multiple sets of historical first-class features and the historical first optimization magnitude, the first linear regression equation is obtained; Based on the first linear regression equation, determine the weight coefficient of each sub-feature in the first type of features; The weight of each sub-feature in the first type of features is determined based on the weight coefficient of each sub-feature. The target correlation coefficient is calculated based on the weight coefficient of each sub-feature and the weight of each sub-feature.

6. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 5, characterized in that, When the target efficiency optimization region is a first type of efficiency optimization region; Based on a pre-set optimization generation model, and according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient, optimization implementation guidance information is generated, including: Based on the multi-source data of on-site operations in the target efficiency optimization area, determine the quantification value of the first influence range; Based on the first type of feature of the target efficiency optimization region and the benchmark values ​​of each feature in the optimization database, the optimization magnitude of each feature is calculated; The process optimization quantification value is calculated based on the target correlation coefficient, the first influence range quantification value, and the optimization magnitude of each feature; Based on the optimization database, and according to the historical optimization measures, the historical process simplification index and its corresponding historical process efficiency optimization range are determined, and the first process optimization coefficient is determined based on multiple sets of historical process simplification indices and their corresponding historical process efficiency optimization ranges. Calculate the target process simplification index based on the process optimization quantification value and the first process optimization coefficient; Based on the optimization database, and according to the historical optimization measures, historical resource allocation data and their corresponding historical resource allocation optimization indices are determined, and a first resource optimization coefficient is determined based on multiple sets of historical resource allocation data and their corresponding historical resource allocation optimization indices. Calculate the target resource allocation optimization index based on the process optimization quantification value and the first resource optimization coefficient; Based on the process simplification baseline value and resource configuration baseline value in the optimization database, and according to the target process simplification index and the target resource configuration optimization index, it is determined whether it is feasible. When it is determined to be feasible, the optimization implementation guidance information is generated based on the historical optimization measures.

7. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 1, characterized in that, When the target efficiency optimization region is a second type of efficiency optimization region; Based on a pre-set optimization database, and according to the type of the target efficiency optimization region, historical optimization data is determined, and historical optimization measures and target correlation coefficients are determined based on the historical optimization data, including: Based on the optimization database, target historical optimization data is determined according to the second type of features corresponding to the second type of efficiency optimization region. The target historical optimization data includes historical second type features, historical second optimization measures, and historical second optimization magnitude. Based on the multiple sets of historical second-type features and the second historical optimization magnitude, the second linear regression equation is obtained; Based on the second linear regression equation, determine the weight coefficient of each sub-feature in the second type of features; The weight of each sub-feature in the second type of features is determined based on the weight coefficient of each sub-feature. The target correlation coefficient is calculated based on the weight coefficient of each sub-feature in the second type of features and the weight of each sub-feature.

8. The method for optimizing the efficiency of on-site inspection in power marketing according to claim 7, characterized in that, When the target efficiency optimization region is a second type of efficiency optimization region; Based on a pre-set optimization generation model, and according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient, optimization implementation guidance information is generated, including: Based on the multi-source data of on-site operations in the target efficiency optimization area, determine the quantification value of the second influence range; Based on the second type of features of the target efficiency optimization region and the baseline values ​​of each feature in the optimization database, the optimization magnitude of each feature is calculated; Based on the target correlation coefficient, the second influence range quantification value, and the optimization magnitude of each feature, calculate the transmission optimization potential quantification value; Obtain the pre-processing transmission status data and the post-processing transmission status data of the target efficiency optimization region, obtain the transmission status benchmark value based on the optimization database, and obtain the pre-processing transmission parameter value according to the pre-processing transmission status data, the post-processing transmission status data and the transmission status benchmark value. Based on the optimization database, and according to the historical optimization measures, the historical transmission efficiency improvement rate and its corresponding historical operation efficiency improvement rate are determined, and the correlation optimization coefficient is determined based on multiple sets of historical transmission efficiency improvement rates and their corresponding historical operation efficiency improvement rates. Based on the quantitative value of transmission optimization potential, the value of preprocessed transmission parameters, and the correlation optimization coefficient, calculate the threshold for achieving transmission efficiency. Based on the transmission efficiency qualification threshold in the optimization database, it is determined whether the implementation is feasible. If it is determined to be feasible, the optimization implementation guidance information is generated based on the historical optimization measures.

9. A device for optimizing the efficiency of on-site inspection in power marketing, characterized in that, include: The acquisition module is used to acquire multi-source data on field operations at multiple locations; The verification module is used to perform independent and cross-verification of the multi-source data of the on-site operation at each location based on the pre-set verification standards, and generate operation status verification results. The region filtering module is used to filter out target efficiency optimization regions from multiple locations and determine the type of the target efficiency optimization regions based on pre-set region type determination rules, according to the multi-source data of the on-site operations at multiple locations and their corresponding operation status verification results. The query module is used to determine the target historical optimization data based on the pre-set optimization database and the type of the target efficiency optimization region, and to determine the historical optimization measures and target correlation coefficient based on the target historical optimization data; The generation module is used to generate optimization implementation guidance information based on a pre-set optimization generation model, according to the multi-source data of on-site operations corresponding to the target efficiency optimization area, the historical optimization measures, and the target correlation coefficient.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for optimizing the efficiency of on-site operation inspection in power marketing as described in any one of claims 1 to 8.

Images