Method for resource allocation and optimal scheduling of supervisory resources

Abstract

The application provides a method for resource allocation and optimal scheduling of supervision resources. The method matches the attributes of the supervisors with the supervision areas by constructing a three-dimensional visualization model of cable laying, generates an initial allocation scheme, and then dynamically adjusts the allocation of supervisors according to the correlation analysis of historical electrical engineering quality problems and supervisor attributes. The construction process is monitored in real time using Internet of Things technology, and abnormal conditions are warned and located. The quality of key processes is judged through image analysis, and professional supervisors are matched according to the severity of the problem. Skill training is carried out combined with virtual reality technology, and personnel scheduling strategies are optimized. The allocation of supervisors is continuously evaluated and dynamically adjusted through big data analysis, forming a closed-loop optimization. In the application, the matching degree of supervisors and supervision areas is high, and the supervisors can quickly and efficiently complete the supervision work, effectively improving the construction efficiency of the cable laying project and improving the quality of the cable laying project.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, in particular to a method for resource allocation and optimal scheduling of supervision resources. BACKGROUND

[0002] In cable laying projects, the scientific allocation and optimal scheduling of supervisors face technical challenges. The complexity and variability of cable laying paths pose great challenges to supervision work. The pipeline layout in different areas, pipeline layout and other factors differ greatly, resulting in highly nonlinear characteristics of cable laying paths. Supervisors need to analyze the characteristics of each path in detail and dynamically adjust the supervision strategy according to the actual situation, which puts high demands on the professional competence of supervisors. There is a contradiction between the professional qualifications of supervisors and the resource allocation of supervision areas. Since cable laying involves multiple professional fields such as electrical, civil, mechanical, etc., there are differences in technical background and working methods among supervisors of different professions. How to reasonably allocate supervisors of different professions according to the specific circumstances of the supervision area and make them work closely together to form a combined force is a problem that needs to be solved urgently. SUMMARY

[0003] The present application provides a method for resource allocation and optimal scheduling of supervision resources, mainly comprising:

[0004] Obtain cable laying path data, build a three-dimensional visualization model, match the attributes of supervisors with the supervision areas in the model, and dynamically adjust the allocation of supervisors according to the correlation analysis of historical electrical engineering quality problems and supervisor attributes; in the concealed parts of the cable laying path, real-time collection of environmental parameters, equipment status and personnel position data of the cable laying site, visualization of the construction process, sending of early warning to the adapted supervisor in case of abnormality, highlighting of the abnormal position in the three-dimensional model; analysis of the images and videos of cable fireproof plugging and cable insulation resistance testing at the construction site, extraction of key information of the plugging material and resistance tester, comparison of the key information with the cable attributes at the corresponding position in the three-dimensional visualization model, and judgment of whether the construction quality meets the specification standards; if the fireproof plugging material is found to be inconsistent or the insulation resistance test is abnormal, match the adapted supervisor for disposal according to the problem severity, feed back the review situation to the three-dimensional visualization model and update the cable state; receive the update instruction of the cable laying construction, understand the semantics of the update content, evaluate the feasibility of the update in combination with the skill level of the supervisors; digitize the data of the cable laying construction, record the supervision opinions and operation logs, record the supervision of the changes in the cable laying construction into the evidence, and continuously evaluate and dynamically adjust the allocation of supervisors to form a closed loop optimization.

[0005] The technical scheme provided by the embodiment of the present application can include the following beneficial effects:

[0006] The application discloses a method for resource configuration and optimized scheduling of supervision resources. The method matches the attributes of supervision personnel with supervision areas by constructing a three-dimensional visualization model of cable laying to generate an initial configuration scheme. The construction process is monitored in real time using Internet of Things technology to provide early warning and positioning for abnormal conditions. The quality of key procedures is determined through image analysis, and professional supervision personnel are matched according to the severity of the problem to deal with it. Skill training is carried out in combination with virtual reality technology to optimize personnel scheduling strategies. Decision tree methods are used to evaluate the feasibility of design changes to achieve digital evidence storage and responsibility tracing. The application continuously evaluates and dynamically adjusts the configuration of supervision personnel through big data analysis to form a closed-loop optimization, effectively improving the intelligent level and quality control ability of electrical engineering supervision. Because the matching degree of supervision personnel and supervision areas is high, supervision personnel can quickly and efficiently complete supervision work, effectively improving the construction efficiency of cable laying projects and improving the quality of cable laying projects. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 The flowchart of the method for resource configuration and optimized scheduling of supervision resources of the application. DETAILED DESCRIPTION

[0008] The technical solutions of the application will be described below in conjunction with the embodiments. Obviously, the described embodiments are only some of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the application.

[0009] As Figure 1 , the method for resource configuration and optimized scheduling of supervision resources of the embodiment can specifically include:

[0010] S101, acquire cable laying path data, construct a three-dimensional visualization model, match the attributes of supervision personnel with supervision areas in the model, and dynamically adjust the configuration of supervision personnel according to the association analysis of historical electrical engineering quality problems and supervision personnel attributes.

[0011] In the embodiment, cable laying path data is acquired, a three-dimensional visualization model is constructed, the attributes of the supervisors are matched with the supervision areas in the model, and an initial supervisor configuration scheme is generated; based on the correlation analysis of historical electrical engineering quality problems and the attributes of the supervisors, the configuration scheme is optimized, and dynamic adjustment of the supervisor configuration is realized. Specifically, cable laying path data is acquired according to electrical engineering design drawings, a three-dimensional visualization model is constructed using AutoCAD software, and the cable laying path data includes cable laying paths, cable specifications and models, pipeline layout, and grounding device arrangement information in different areas. Risk level evaluation is performed on the three-dimensional visualization model, the weights of the engineering difficulty coefficients and the equipment complexity factors are calculated using the analytic hierarchy process, and the risk scores of each supervision area are obtained. According to the risk scores, the historical electrical engineering quality problems are classified using the K-means clustering algorithm, and a correlation model of the quality problems and the attributes of the supervisors is established. After the correlation model is established, an optimization strategy for the supervisor configuration is obtained by training a BP neural network, the input layer of the BP neural network includes the attributes of the supervisors such as education, working years, and professional certificates, and the output layer is the probability of the quality problems. The attributes of the supervisors are acquired, the attributes of the supervisors are matched with the supervision areas in the three-dimensional visualization model, and an initial supervisor configuration scheme is generated. According to the initial supervisor configuration scheme, the frequency of the quality problems in each supervision area is counted using a sliding time window, and it is determined whether the frequency of the quality problems exceeds a preset threshold. If the frequency of the quality problems exceeds the preset threshold, a dynamic adjustment process of the supervisors is started.

[0012] Specifically, the cable laying path data is obtained from the electrical engineering design drawings, the pipeline layout and grounding device arrangement information is extracted, and a three-dimensional visualization model is constructed using AutoCAD software. The model is labeled with different regional cable specifications, models, and laying methods. Risk level assessment is performed on the three-dimensional visualization model, and factors such as engineering difficulty coefficient and equipment complexity are considered. The weight of each factor is calculated using the analytic hierarchy process, and the risk score of each supervision area is obtained. The historical electrical engineering quality problems are classified using the K-means clustering algorithm, and a correlation model between quality problems and supervisor attributes is established. The optimization strategy for supervisor allocation is obtained through BP neural network training. The input layer includes supervisor education, work experience, and professional certificates, and the output layer is the probability of quality problems. The supervisor attributes are obtained, and the knowledge graph technology is used to match the supervisor attributes with the supervision areas in the three-dimensional visualization model to generate an initial supervisor allocation scheme. The correlation between the engineering characteristics of the supervision area and the supervisor attributes is calculated, and the minimum support and confidence thresholds are set to filter out strong correlation rules. If the matching degree of a supervisor and their assigned area is below the preset threshold, the adjustment mechanism is triggered, and experienced supervisors are added to high-risk areas, while new supervisors are assigned to low-risk areas for training. According to the real-time engineering progress and quality feedback data, the quality problem frequency of each supervision area is calculated using a sliding time window. When the frequency exceeds the preset threshold, the supervisor dynamic adjustment process is started, and the configuration optimization algorithm is re-executed. In the electrical engineering design drawings, image recognition technology is used to extract cable laying path data, with an identification rate of over 95%. Pipeline layout and grounding device arrangement information are extracted from the drawing text description using OCR technology, with an accuracy rate of 98%. The extracted data is converted into a three-dimensional visualization model using the API interface of AutoCAD software, with a model precision error controlled within ±5mm. When performing risk assessment on the three-dimensional visualization model, factors such as engineering difficulty coefficient (weight 0.4), equipment complexity (weight 0.3), and environmental factors (weight 0.3) are considered. The weight is calculated using the analytic hierarchy process, a judgment matrix is constructed, and the weight vector is solved using the eigenvalue method. The risk score of each supervision area ranges from 0 to 100, with scores above 80 indicating high-risk areas and scores below 50 indicating low-risk areas. The historical quality problems are classified using the K-means clustering algorithm, with a clustering number K=5 and Euclidean distance as the similarity measure. The BP neural network model includes an input layer (10 nodes corresponding to supervisor attributes), a hidden layer (20 nodes), and an output layer (5 nodes corresponding to quality problem types). The Sigmoid activation function is used, the learning rate is set to 0.01, and the iteration number is 1000. The knowledge graph technology is used to construct the association network between supervisor attributes and supervision areas. The graph includes node types, such as supervisors, supervision areas, and relationship types, including suitable supervision and having qualifications.The node similarity is calculated using a graph embedding algorithm, and the cosine similarity is selected as the matching metric. The minimum support is set to 0.1, and the minimum confidence is set to 0.7. The Apriori algorithm is used to mine association rules. The matching degree threshold of the supervisor and the assigned area is set to 0.8, and below this value triggers the adjustment mechanism. The real-time engineering progress and quality feedback data uses a sliding time window of 1 week, and the frequency of quality problems is counted. When the frequency exceeds 3 times per week, the supervisor dynamic adjustment process is started, and the configuration optimization algorithm is re-executed. The frequency of adjustment is not more than 1 time per month to ensure the stability of the project.

[0013] The supervisor skill training is carried out through an online learning platform. The cable laying and grounding device construction process are simulated through virtual reality, and fault scenarios are set for practical operation exercises. The operation of the supervisor is scored, and the training results are associated with the supervisor attributes to optimize the personnel scheduling strategy.

[0014] The supervisor skill training course content is obtained, and multiple difficulty levels of learning modules are set according to the course content. The course content association matrix is used to construct the association relationship between the learning content. The three-dimensional modeling software is used to convert the engineering design drawings into three-dimensional visual models, and the cable laying and grounding device construction scenes are constructed according to the three-dimensional visual models. The construction specifications are converted into executable virtual operation processes. The operation behavior of the supervisor in the virtual operation process is recorded, and the operation similarity between the operation behavior and the preset standard operation process is calculated. The training results of the supervisor are obtained according to the operation similarity, and the training results are associated with the supervisor attribute data to construct the supervisor skill evaluation index system. The genetic algorithm is used to calculate the optimal personnel scheduling scheme, and if the engineering progress and quality feedback change, the personnel allocation is dynamically adjusted according to the changes to realize the continuous optimization of the supervisor configuration.

[0015] Specifically, an online learning management system is used to create a supervision personnel skill training course, multiple difficulty levels of learning modules are set for cable laying and grounding device construction process, the correlation between learning content is constructed through course content correlation matrix, the correlation weight is calculated according to the similarity and difficulty level of learning content, and the adaptive learning path recommendation is realized. A three-dimensional modeling software is used to build a cable laying and grounding device construction scene, engineering design drawings are converted into three-dimensional visual models, construction specifications are converted into executable virtual operation processes, a variety of common fault scene databases are imported, real operation practice tasks are generated by random combination, fault scenes are integrated with virtual environment to form a complete virtual training environment. Action capture technology is used to record the operation behavior of supervision personnel in the virtual environment, combined with the preset standard operation process, operation sequence comparison algorithm is used to calculate the operation similarity, and the real operation performance of supervision personnel is quantitatively scored from three dimensions of time, space and operation sequence. The training results are associated with the attribute data of supervision personnel, a supervision personnel skill evaluation index system is constructed, a genetic algorithm is used to calculate the optimal personnel scheduling scheme, the training results, personnel attributes and engineering requirements are converted into optimization objectives, an adaptability function is designed to evaluate the pros and cons of the scheduling scheme, and the personnel allocation is dynamically adjusted according to the engineering progress and quality feedback to realize the continuous optimization of supervision personnel configuration. The online learning management system builds a course containing cable laying and grounding device construction process, sets three difficulty levels of primary, intermediate and advanced, and each level contains 10 learning modules. The course content correlation matrix uses a 50x50 two-dimensional array, the similarity of learning content is calculated by cosine similarity, and the difficulty level difference is used as a weight factor with a weight range of 0.1-1.0. The adaptive learning path recommendation uses a collaborative filtering algorithm, based on user historical learning records and content similarity, and recommends courses with a similarity greater than 0.8. The three-dimensional modeling software imports CAD format engineering drawings with a conversion accuracy of ±1mm, and generates a construction scene containing 500 virtual objects. The construction specifications are converted into 100 standardized operation processes, each process contains 10-20 operation steps. The fault scene database contains 200 common faults, and 1000 real operation practice tasks are generated by random combination. The action capture system has a sampling frequency of 60Hz, and records the spatial coordinates of 18 key nodes of the supervision personnel. The operation sequence comparison algorithm uses a dynamic programming method to calculate the edit distance between the standard operation process and the actual operation, the time dimension error threshold is set to ±5 seconds, the space dimension error threshold is set to ±10cm, and the operation sequence consistency requirement is greater than 80%. The quantitative scoring uses a percentage system, and the weights of time, space and sequence dimensions are 0.3, 0.3 and 0.4 respectively. The supervision personnel skill evaluation index system includes professional knowledge, operation skill and emergency disposal, and their weights are 0.4, 0.4 and 0.2 respectively, and each aspect has 5 secondary indicators. The population size of genetic algorithm is set to 100, the crossover probability is 0.8, the mutation probability is 0.1, and the iteration number is 1000.The fitness function comprehensively considers the personnel skill matching degree, workload balance and cost factors, and the weights thereof are 0.5, 0.3 and 0.2 respectively. According to the weekly engineering progress report and quality feedback data, the personnel allocation is dynamically adjusted, the adjustment cycle is 2 weeks, the adjustment amplitude is not more than 20% of the total number of personnel each time, so as to maintain the continuity of work.

[0016] According to the specific situation of the supervision area, the matching degree of the supervision personnel attribute and the supervision task is analyzed, the importance weight of different professional fields, the working experience and skill level of the supervision personnel are combined, the objective function and the constraint condition are constructed, and the optimal supervision personnel allocation scheme is solved.

[0017] The expert questionnaire survey data are acquired, the importance weight of each professional field is calculated, the supervision area professional demand score vector is obtained, the data are acquired from the professional qualification, working experience and skill level database of the supervision personnel according to the supervision area professional demand score vector, the multi-index evaluation method is used to quantize the comprehensive ability of each supervision personnel, the supervision personnel ability score vector is obtained, the weighted cosine similarity algorithm is used to calculate the similarity between the supervision personnel ability score vector and the supervision area professional demand score vector, the matching degree score matrix of the supervision personnel to each supervision area is obtained, after the matching degree score matrix is obtained, the integer programming model with the maximum overall matching degree as the objective function is established, the allocation relationship between the supervision personnel and the supervision area is the decision variable of the integer programming model, based on the integer programming model, the branch and bound algorithm is used to solve the optimal supervision personnel allocation scheme, and the personnel allocation result of each supervision area is output.

[0018] Specifically, according to the geographical location, engineering scale, technical difficulty and other characteristics of the supervision area, the importance weight of each professional field is calculated by Delphi method combined with entropy weight method. The opinions are collected through multiple rounds of expert questionnaire survey, and the objective weight of electrical, civil, water supply and drainage and other professions is calculated by entropy weight method to obtain the professional demand score vector of the supervision area. The professional qualification, work experience and skill level data of the supervision personnel are obtained, and the comprehensive ability of each supervision personnel is quantified by using multi-index evaluation method. The specific indexes include education background, professional title, project experience, skill certification, etc. The weighted score is calculated by setting the evaluation index set and standardizing the processing to obtain the ability score vector of the supervision personnel. The matching degree evaluation function of the supervision personnel and the supervision task is constructed, the similarity between the ability score vector of the supervision personnel and the professional demand score vector of the supervision area is calculated by using weighted cosine similarity algorithm, and the matching degree score matrix of each supervision personnel to each supervision area is obtained by considering the importance weight of different professional fields. An integer programming model is established, the objective function is to maximize the overall matching degree, the decision variable is the allocation relationship between the supervision personnel and the supervision area, the constraint conditions include personnel quantity limitation, professional qualification requirement, workload balance, etc. The optimal supervision personnel allocation scheme is solved by using branch and bound algorithm, the solving efficiency is improved by setting the branch strategy and pruning rule, and the personnel allocation result of each supervision area is output. In the supervision project management, the supervision area is evaluated first, considering the geographical location such as city center, suburb, remote area, engineering scale such as small, medium and large, and technical difficulty such as low, medium and high. Delphi method is used for three rounds of expert questionnaire survey, 15 industry experts are invited to participate, and importance score is given to 10 professional fields such as electrical, civil, water supply and drainage. The questionnaire uses Likert 5-level scale, 1-5 represents unimportant to extremely important. After collecting the opinions of experts, the weight of each profession is calculated by using entropy weight method to obtain the professional demand score vector of the supervision area. For example, the professional demand score vector of a supervision area is [0.3, 0.25, 0.2, 0.15, 0.1], which corresponds to electrical, civil, water supply and drainage, heating and ventilation and fire protection respectively. For supervision personnel evaluation, the data of education background (1 point for bachelor degree, 2 points for master degree, 3 points for doctor degree), professional title (1 point for assistant, 2 points for intermediate, 3 points for senior), project experience (0.5 points per year, upper limit 5 points) and skill certification (0.5 points per one, upper limit 2 points) are collected. The comprehensive score is calculated by using weighted summation method, and the weights are 0.2, 0.3, 0.3 and 0.2 respectively. The score is standardized to obtain the ability score vector of the supervision personnel. For example, the ability score vector of a supervision personnel is [0.8, 0.7, 0.6, 0.5, 0.4], which corresponds to the above five professions. When constructing the matching degree evaluation function, the weighted cosine similarity algorithm is used to calculate the similarity between the ability score vector of the supervision personnel and the professional demand score vector of the supervision area. The weight uses the value in the professional demand score vector to obtain the matching degree score matrix of each supervision personnel to each supervision area.Finally, an integer programming model is established, where the decision variable xij represents whether the i-th supervisor is assigned to the j-th supervision area (0 or 1). The objective function is to maximize the overall matching score. Constraints include that at least one supervisor is assigned to each supervision area, each supervisor is assigned to at most one area, and specific professional qualification requirements must be met. A branch and bound algorithm is used to solve the problem, with the branching strategy prioritizing the minimum lower bound and the pruning rule using the local upper bound method. Through iterative calculation, the optimal supervisor allocation scheme is obtained.

[0019] S102. In the concealed parts of the cable laying path, collect environmental parameters, equipment status and personnel location data of the cable laying site in real time, visualize and monitor the construction process, send early warnings to the appropriate supervisors when abnormalities occur, and highlight the abnormal locations in the three-dimensional model.

[0020] In this embodiment, environmental parameters, equipment status, and personnel location data of the cable laying site are collected in real time in concealed locations including cable trenches and cable wells, enabling visual monitoring and early warning of the construction process. If an abnormal state is detected, an early warning message is sent to the supervisors with the corresponding professional qualifications and work experience, and the abnormal location is highlighted in the 3D visualization model to guide the supervisors to quickly locate the problem. Specifically, the process involves acquiring environmental parameters and equipment operating status data collected by a wireless sensor network deployed in concealed locations, including cable trenches and cable wells. Data packets containing sensor IDs, timestamps, measured values, and battery levels are generated based on the collected data. These data packets are then cleaned, normalized, and feature-extracted to obtain processed sensor data. The mean, standard deviation, and rate of change of the processed sensor data are calculated using a sliding time window. Abnormal states are identified using preset threshold rules and an isolated forest algorithm. Based on the identified abnormal states, supervisor information is retrieved from the Neo4j graph database, which stores supervisors' professional qualifications, work experience, and skill ratings. Matching supervisors are selected using the Cypher query language, and warning information containing the abnormality type, location, severity, and occurrence time is pushed to them. The abnormal state data is then associated with the spatial location information in a 3D visualization model to render the 3D scene. The world coordinates of the abnormal location are calculated using a vertex shader, and the intensity of red highlighting is set in the fragment shader based on the distance to the abnormal point, achieving precise location and visualization of the abnormal position.

[0021] Specifically, a wireless sensor network is deployed within cable trenches and wells. Environmental parameters are collected using temperature, humidity, and gas concentration sensors; equipment operating status is monitored using vibration and current sensors; and the location data of construction personnel is obtained using Bluetooth positioning beacons. All sensor data is transmitted to a data acquisition gateway via the LoRaWAN protocol. The sensor sampling frequency is set to once per minute, and the data packet format includes sensor ID, timestamp, measured value, and battery level. A real-time data processing platform is built, using Apache Flink to clean, normalize, and extract features from the sensor data. The mean, standard deviation, and rate of change of various indicators are calculated using a 5-minute sliding time window. Anomalies are identified using preset threshold rules and the isolated forest algorithm. Based on the detected anomalies, warning information including anomaly type, location, severity, and occurrence time is generated. A knowledge base for supervisors is established, recording the professional qualifications, work experience, and skill ratings of each supervisor. The Neo4j graph database is used to store supervisor information, and Cypher query language is used to filter supervisors based on professional matching and geographical location. The most suitable supervisor is selected, and warning information is pushed through a message queue. The 3D visualization model was updated, associating abnormal status data with spatial location information within the model. WebGL technology was used to render the 3D scene in the browser. The world coordinates of the abnormal location were calculated using a vertex shader, and the intensity of red highlighting was adjusted in the fragment shader based on the distance to the abnormal point. Combined with real-time location data, the movement trajectory of the supervisors was displayed in the model, achieving precise location and visualization of the abnormality. 100 wireless sensor nodes were deployed in the cable trenches and cable wells, including 30 temperature sensors (accuracy ±0.5℃), 30 humidity sensors (accuracy ±3%RH), 20 gas concentration sensors (detection range 0-100ppm), 10 vibration sensors (frequency range 0-1000Hz), and 10 current sensors (range 0-100A). Each sensor collected data every 60 seconds, transmitting it to the gateway via the LoRaWAN protocol (frequency band 470-510MHz, transmission rate 0.3-50kbps). The data consists of 20 bytes, including a 4-byte sensor ID, an 8-byte timestamp, a 4-byte measurement value, and a 4-byte battery level. The Apache Flink real-time processing platform receives the data stream and uses a 5-minute sliding time window with a 1-minute step size to calculate the mean, standard deviation, and rate of change. Anomaly detection employs the Isolation Forest algorithm with 100 decision trees, a sampling size of 256, and an anomaly threshold of -0.5. Upon an anomaly detection, a JSON-formatted alert is generated, containing the anomaly type (e.g., excessively high temperature, abnormal humidity), location coordinates (x, y, z), severity level (1-5), and the time of occurrence.The supervisor knowledge base is stored using the Neo4j graph database, containing 50 supervisor nodes. Each node has 10 attributes, such as professional qualifications, years of work experience, and skill rating. The Cypher query language is used to filter the most suitable supervisors based on professional matching (weight 0.6) and geographical location (weight 0.4), with query time controlled within 100ms. Selected supervisor information is pushed to mobile terminals via a RabbitMQ message queue with a throughput of 10,000 messages / second. The 3D visualization model uses WebGL technology, calculating world coordinates through vertex shaders and setting red highlighting effects in the fragment shader. RGB values ​​range from 255,0,0 to 255,200,200, varying with distance from anomaly points. The model updates at 10 frames / second and supports 30 simultaneous online users viewing in real time. Supervisor locations are updated every 5 seconds, marked with blue dots in the model, and their movement trajectories over the past 10 minutes are displayed.

[0022] S103. Analyze the images and videos of cable fireproofing and cable insulation resistance testing at the construction site, extract key information of the sealing materials and resistance testing instruments, compare the key information with the cable attributes at the corresponding locations in the three-dimensional visualization model, and determine whether the construction quality meets the specifications and standards.

[0023] In this embodiment, images and videos of key procedures such as cable fireproofing and cable insulation resistance testing at the construction site are analyzed to extract key information such as the type, specifications, quantity, construction process, and model, range, accuracy, and calibration status of the resistance testing instrument. This information is then compared with the cable attributes at the corresponding locations in the 3D visualization model to determine whether the construction quality meets the standards. Specifically, video data of the cable fireproofing and insulation resistance testing process is acquired; keyframes are extracted from the video data to obtain representative images at one frame per second; target detection algorithms are used to process the representative images to identify the sealing materials and testing instruments, resulting in labeled images; optical character recognition (OCR) methods are used to process the labeled images to extract text information, including the type, specifications, and quantity of sealing materials, as well as the model, range, accuracy, and calibration date of the resistance testing instruments; based on the spatiotemporal information in the labeled images, the text information is correlated with the cable location in the 3D visualization model to obtain associated data; a rule engine based on the Rete algorithm is constructed, and the associated data is input into a rule base containing rule sets for material specification matching, process parameter compliance, and instrument accuracy requirements; based on the output of the rule engine, the compliance of the construction quality is determined, and a quality assessment report is generated.

[0024] Specifically, high-definition cameras are used to capture real-time footage of the cable fireproofing and insulation resistance testing process. Image preprocessing algorithms are used to denoise, enhance, and correct the acquired images. Keyframes are extracted from the video data, and a representative frame is selected every second. The YOLOv5 object detection algorithm is used to identify the sealing materials and testing instruments, and the detection results are marked on the original images to form labeled images. Optical character recognition technology is used to extract text information from the labeled images, including the type, specifications, and quantity of sealing materials, as well as the model, range, accuracy, and calibration date of the resistance testing instruments. The BERT-CRF named entity recognition model is used to perform semantic analysis on the extracted text, identify key information, and store it in a structured manner. The construction process is analyzed to identify the sequence of procedures and key parameters, and process parameters such as sealing depth, compaction degree, and test voltage are extracted. Based on the spatiotemporal information in the images, the extracted construction process data is associated with the cable location in the 3D visualization model. The R-tree spatial indexing algorithm is used to quickly locate the construction area, establish a mapping relationship between image coordinates and 3D model coordinates, and obtain the cable attribute information at the corresponding location from the model database, including cable model, cross-sectional area, and length. A rule engine based on the Rete algorithm was constructed. Extracted construction data and cable attribute information were input into a rule base, which included rule sets for material specification matching, process parameter compliance, and instrument accuracy requirements. Through rule matching and inference, the system compared whether the sealing materials and test instrument parameters met the specifications, determined the compliance of the construction quality, and generated a quality assessment report containing the judgment results and detailed comparison data. At the cable construction site, 10 4K high-definition cameras were deployed, each capturing 1080p resolution video streams at 30fps. In the video preprocessing stage, Gaussian filtering was used for noise reduction, histogram equalization was used to enhance image contrast, and affine transformation was used to correct image tilt. The keyframe extraction algorithm selected one frame per second and used the inter-frame difference method to calculate the difference between adjacent frames, with a difference threshold set to 0.1. The YOLOv5 object detection model was trained on 5000 labeled images, including 20 types of sealing materials and 10 types of test instruments, achieving a detection accuracy of 95%. Optical character recognition used the Tesseract engine, with a character recognition accuracy of 98%. The BERT-CRF named entity recognition model was trained using 10,000 annotated corpora containing 15 entity types, achieving an F1 score of 0.92. Construction process analysis employed a rule-based sequence labeling algorithm to identify 8 standard procedures and extract 12 key process parameters. The R-tree spatial indexing algorithm constructed a 5-level tree structure, with each node containing a maximum of 50 child nodes, achieving a query time complexity of O(logn). 3D model coordinate mapping utilized an affine transformation matrix, with mapping errors controlled within ±5cm. The rule engine, based on the Rete algorithm, included 100 rules covering 5 major categories of construction specifications. Rule matching speed reached 10,000 times / second, with a single quality assessment taking no more than 100ms.The final quality assessment report contains 20 assessment indicators. The pass / fail standards and actual values ​​of each indicator are compared, and quantitative scores and qualitative recommendations are given.

[0025] S104. If the fireproof sealing material is found to be inconsistent or the insulation resistance test is abnormal, the appropriate supervisor shall handle the matter according to the severity of the problem, and the verification results shall be fed back to the three-dimensional visualization model and the cable status shall be updated.

[0026] In this embodiment, if the fire-stopping material is found to be inconsistent with the design or the insulation resistance test results are abnormal, a highly skilled supervisor will be assigned to handle the issue based on its severity. The on-site verification results will be fed back to the 3D visualization model to update the cable status. Specifically, a convolutional neural network is used to perform image recognition of the fire-stopping material and obtain material specification information from the design database. A similarity score is calculated based on the image recognition results and the material specification information. If the similarity score is lower than a preset threshold, the fire-stopping material is determined to be inconsistent. Based on the test data collected by the insulation resistance tester, the test data is analyzed using an ARIMA model. The deviation between the test data and the standard value is calculated using the ARIMA model. The degree of deviation is determined based on the deviation to obtain a severity index. A supervisor competency assessment model is constructed. The input features of the competency assessment model include the supervisor's professional qualifications, work experience, and historical handling results. The supervisor's competency assessment model outputs a professional competency score for the supervisor. Based on the severity index of the abnormal situation information, match the supervisors with corresponding professional competence scores; push the abnormal situation information to the matched supervisors; use real-time communication protocols to synchronize the on-site verification data to the three-dimensional visualization model and update the cable status information.

[0027] Specifically, a ResNet50 convolutional neural network is used for image recognition of fireproof sealing materials. Material specification information is obtained from a design database, and feature extraction is performed to compare the material specifications with those in the database. A similarity score is calculated, and if the similarity score is lower than a preset threshold of 0.85, the material is considered non-compliant. Based on data collected by an insulation resistance tester, the test results are analyzed using an ARIMA model to calculate the deviation between the test value and the standard value. A five-level warning threshold is set, quantifying the degree of deviation into a severity index of 1-5. A supervisory personnel competency assessment model is constructed, using the supervisor's professional qualifications, work experience, and historical handling results as input features. A gradient boosting decision tree algorithm is implemented using the XGBoost library to train a regression model that outputs a professional competency score for the supervisor. Supervisors with corresponding competency levels are matched according to the severity index of the problem. A mobile application was developed to push notifications of anomalies to selected supervisors. Augmented reality guidance via GPS and ARKit directs supervisors to the problem location, collects on-site verification data, and uploads it to a central server. After data verification and conflict resolution, the verification results are synchronized to a 3D visualization model using the WebSocket real-time communication protocol, updating cable status information, including material type, construction quality, and insulation performance. At the cable construction site, five high-definition cameras were deployed to capture real-time images of fire-resistant sealing materials. The ResNet50 network, trained on 10,000 labeled images, achieved a 98% recognition accuracy. The design database contains 500 standard material specifications. Feature vector matching was calculated using cosine similarity, with a threshold set at 0.85. An insulation resistance tester collects data every 5 minutes. The ARIMA model uses data from the past 24 hours for prediction, setting five warning thresholds: ±5%, ±10%, ±15%, ±20%, and ±25% of the standard value. The supervisor's competency assessment model is based on the XGBoost algorithm, using 100 decision trees with a learning rate of 0.1 and a maximum depth of 6. Parameters are optimized through 5-fold cross-validation. Input features include professional qualifications (1-5 points), work experience (years), and historical handling results (success rate), outputting a competency score from 0-100. The mobile application is developed based on the Flutter framework and supports Android and iOS systems. GPS positioning accuracy is controlled within ±3 meters, and ARKit achieves centimeter-level spatial positioning accuracy. On-site verification data includes 20 key parameters, and data integrity is verified using the SHA256 algorithm. WebSocket communication latency is controlled within 100ms. The 3D visualization model is built on the Unity engine, supporting 1000 concurrent users. The model updates 10 times per second, and the average response time from anomaly detection to supervisor arrival on-site is controlled within 15 minutes, enabling real-time monitoring and rapid handling of cable construction quality.

[0028] Obtain the schedule and milestones of the cable laying project, and use the critical path method and resource balance method to optimize the allocation of work tasks and resources for the supervisors, respectively, and arrange the on-site time and rest schedule of the supervisors accordingly, and make dynamic adjustments based on the actual progress.

[0029] Acquire the schedule and milestone data for the cable laying project. The schedule includes task names and estimated durations. Generate a Gantt chart to visualize the project timeline based on the schedule and milestone data, displaying task start and end times. Calculate the critical path using the critical path algorithm to obtain the sequence of tasks with the greatest impact on the overall project duration, determining the earliest start and latest finish times for each task. Optimize the task allocation for supervisors using a genetic algorithm. The genetic algorithm uses binary encoding to represent task allocation schemes and generates new schemes through single-point crossover and bit-flip mutation operations. Collect actual project progress data through a real-time progress tracking system. The real-time progress tracking system records task completion percentages and actual cost expenditures. Calculate the deviation between the plan and the actual progress using earned value analysis. If the progress deviation exceeds a preset threshold, a rescheduling mechanism is triggered to dynamically adjust the work arrangements for supervisors.

[0030] Specifically, the project schedule and milestone data for the cable laying project are imported from the project management system. A Gantt chart visualization tool is used to display the project timeline. The critical path algorithm is used to calculate the critical path, identifying the task sequence with the greatest impact on the overall project duration, and determining the earliest start and latest finish times for each task. PERT technology is employed to handle complex dependencies, identify parallel tasks, and construct a complete task network diagram. A resource pool for supervisors is built, recording each supervisor's professional skills, work experience, and current workload. A resource histogram balancing method is used for initial task allocation, calculating the resource requirements and available resources for each task. Resource usage is smoothed by adjusting the start times of tasks on non-critical paths. A genetic algorithm is used to optimize the task allocation for supervisors. Binary encoding is used to represent the task allocation scheme, and single-point crossover and bit-flip mutation operations are used to generate new schemes. Task completion time and resource utilization rate are used as fitness functions, iteratively optimizing until the optimal solution is found, resulting in the specific work arrangement for each supervisor. Based on the optimized task allocation scheme, a schedule and rotation plan for the on-site supervision personnel were developed, taking into account statutory working hours and continuous working day limits. Real-time progress tracking data was collected, and earned value analysis was used to calculate the deviation between the planned and actual progress. When the schedule deviation exceeded 10% or the cost deviation exceeded 5%, a rescheduling mechanism was triggered to dynamically adjust the work arrangements of the supervision personnel. If the actual progress was severely behind schedule, an emergency scheduling plan was activated to temporarily increase human resources or adjust work priorities. In the cable laying project, the project management system imported a schedule plan containing 500 task nodes, including 20 key milestones. The Gantt chart tool displayed the 180-day project duration in days, and the critical path algorithm identified 78 key tasks, accounting for 15.6% of the total tasks. PERT technology handled 50 complex dependencies and identified 30 task groups that could be executed in parallel. The resource pool included 50 supervision personnel, each possessing an average of 3 professional skills, with work experience ranging from 1 to 20 years. The resource histogram balancing method controls daily resource demand fluctuations within ±10%, increasing average resource utilization to 85%. The genetic algorithm uses 100-bit binary encoding to represent task allocation schemes, with a population size of 200, a crossover probability of 0.8, a mutation probability of 0.05, and convergence after 500 generations. The fitness function has a work option weight of 0.6 and a resource utilization weight of 0.4. The optimized scheme shortens the total project duration by 5 days and increases resource utilization by 7%. The shift work plan adheres to the principle of no more than 44 hours of work per week and no more than 6 consecutive days of work. The real-time progress tracking system updates progress data every 4 hours, and earned value analysis calculates the Schedule Performance Index (SPI) and Cost Performance Index (CPI). When the SPI is less than 0.9 or the CPI is less than 0.95, rescheduling is triggered, with an average response time of 2 hours.The emergency dispatch plan has five preset resource allocation modes, which can mobilize an additional 20% of human resources within 24 hours, realizing intelligent scheduling and optimization of the entire process of cable laying project with a team of 50 people and a construction period of 180 days.

[0031] S105. Receive the update instruction for cable laying construction, perform semantic understanding of the update content, and assess the feasibility of the update based on the skill level of the supervisor.

[0032] In this embodiment, the update instructions for cable laying are transmitted in real time via wireless communication. The updated content is semantically understood, and the specified information on the grounding device material and specifications is extracted. Combined with the skill level and communication ability of the supervisor, the feasibility of the update is evaluated. Specifically, the system receives an update instruction, encrypts it using AES-256, and generates an integrity verification value using the SHA-256 algorithm. Natural language processing is then performed on the encrypted update instruction, including word segmentation, part-of-speech tagging, and named entity recognition, to extract specified information from the design change instruction. Based on this specified information, a supervisor's competency assessment model is constructed. This model employs a multilayer perceptron algorithm, with a preset number of nodes in the input layer, two preset number of nodes in each of the two hidden layers, and a preset number of nodes in the output layer, yielding the supervisor's skill level score and communication ability score. Based on the specified information from the update instruction, the supervisor's skill level score, and their communication ability score, a decision tree model is constructed. The decision nodes of the decision tree model include change complexity, material availability, and supervisor competency matching degree. The optimal splitting feature is selected using the information gain ratio, and a pessimistic pruning method is used to prevent overfitting, resulting in a change feasibility assessment result.

[0033] Specifically, a low-power wide-area network (LPWAN) is deployed, employing the LoRaWAN protocol for real-time transmission of design change instructions. Data is uploaded to a cloud server via a gateway, and message queuing technology ensures reliable transmission and sequential processing of instructions. Received instructions are encrypted using AES-256 and verified for integrity using SHA-256. A local caching mechanism is implemented to automatically save instructions when communication is interrupted and retransmit them upon network recovery. Natural language processing (NLP) techniques are used to preprocess the change instructions, including word segmentation, part-of-speech tagging, and named entity recognition. Domain-specific dictionaries and rules are used to identify keywords and phrases related to the grounding device. Dependency parsing and semantic role labeling are used to extract specific information on materials and specifications, and relation extraction techniques are used to identify relationships between entities. The extracted information is matched with predefined supervision skill requirements to generate a preliminary assessment of the required supervision capabilities. A capability assessment model for construction supervisors was constructed, using their professional qualifications, work experience, and historical performance as input features. A multilayer perceptron algorithm was used to train the regression model. The network structure consisted of an input layer with 10 nodes, two hidden layers with 20 nodes each, and an output layer with 2 nodes. The ReLU activation function and Adam optimizer were used to output quantitative scores of skill level and communication ability, establishing a database of supervisor capability profiles. Based on extracted change information and supervisor capability scores, a C4.5 decision tree model was constructed to assess change feasibility. Change complexity, material availability, and supervisor capability matching were used as decision nodes. The optimal splitting feature was selected using information gain ratio, and pessimistic pruning was used to prevent overfitting, generating feasibility assessment results and corresponding decision-making basis. A feasibility threshold of 0.75 was set. When the assessment result exceeded this value, a change suggestion report was automatically generated, including specific implementation steps and a list of required resources. In the cable laying project, 50 LoRaWAN nodes were deployed, covering a 10 square kilometer construction area, with a data transmission rate of 5.5 kbps and communication latency controlled within 100 ms. The cloud server uses a Kafka message queue, processing 1000 instructions per second. The AES-256 encryption key is 256 bits long, and the SHA-256 hash value is 32 bytes long. The local cache has a capacity of 1GB, capable of storing instructions from the past 24 hours. The natural language processing module uses the jieba word segmenter, achieving an accuracy of 98%, and a named entity recognition F1 score of 0.92. The domain dictionary contains 5000 specialized terms. Dependency parsing uses the Stanford parser, achieving an accuracy of 85%. Semantic role labeling uses the BERT model, achieving an F1 score of 0.88. Relation extraction accuracy reaches 80%. The supervision capability assessment model is based on the PyTorch framework. The input layer contains 10 features, such as education level, years of work experience, and project experience. Each of the two hidden layers has 20 neurons, and the output layer has two nodes representing skill level and communication ability, respectively.The ReLU activation function was used, with a learning rate of 0.001 and a batch size of 64. After 5000 training epochs, convergence was achieved, and the average error decreased to 0.05. The maximum depth of the C4.5 decision tree was set to 8, and the minimum number of splits was 10. Cross-validation was used to select the optimal parameters. The information gain ratio threshold was set to 0.1, and pessimistic error rate estimation was used during pruning. The feasibility assessment accuracy reached 85%. For changes with an assessment result higher than 0.75, a recommendation report containing 10 implementation steps and 20 resource requirements was automatically generated.

[0034] S106. Digitize and store data related to cable laying construction, record the review opinions and operation logs of supervisors, include the supervision records of cable laying construction changes in the storage, and continuously evaluate and dynamically adjust the allocation of supervisors to form a closed-loop optimization.

[0035] In this embodiment, cable laying construction data is digitally stored, recording the review opinions and operation logs of each supervisor to achieve accountability. Supervision records of cable laying construction changes are also included in the digital storage scope. Simultaneously, the supervisor configuration plan is continuously evaluated to ensure it meets preset requirements, and the supervisor configuration is dynamically adjusted based on the evaluation results, forming a closed-loop optimization. Specifically, cable laying construction data is acquired and stored in a distributed file storage system. The hash value and timestamp of each document are calculated based on the construction data in the distributed file storage system. The hash value and timestamp are recorded to obtain an immutable digital storage chain. Supervisor review operations and system interactions are received; click paths, dwell times, and operation frequencies are recorded based on these operations. Change requests, review processes, and execution information are acquired and included in the digital storage scope. The content of each change is recorded. Association rules between change items and supervisors are established using the FP-Growth algorithm. Historical performance data of supervisors is acquired and processed in large-scale parallel processing. Based on the parallel processing results, the ARIMA model is used to evaluate whether the supervisor configuration plan meets preset requirements. If not, the personnel scheduling strategy is optimized through a deep Q-learning network. The state space of the deep Q-learning network includes the current personnel distribution and workload, while the action space contains personnel allocation options.

[0036] Specifically, a distributed file storage system is used to digitize cable laying construction data. Image recognition, OCR, and audio / video transcription technologies are used to extract key information from drawings, text, and multimedia files respectively. Blockchain technology is used to record the hash value and timestamp of each document, constructing an immutable digital evidence chain. A supervision operation record system is used, leveraging Google Analytics for Firebase to capture the review operations and system interactions of supervisors, recording click paths, dwell times, and operation frequencies. Natural language processing technology is used to perform semantic analysis on review opinions, extracting key viewpoints and decision-making basis. Through a construction change management platform, change requests, review processes, and implementation status are included in the scope of digital evidence storage. Version control technology records the detailed content of each change, and the FP-Growth algorithm is used to establish association rules between change items and supervisors, generating a change impact assessment report. Apache Spark was used for large-scale parallel processing of historical performance data of supervisors. The ARIMA model was employed to evaluate the rationality of the supervisor configuration scheme. A Deep Q-Learning Network (DQN) was used to optimize the personnel scheduling strategy. The state space was defined to include the current personnel distribution and workload, the action space to represent personnel allocation options, and the reward function was designed based on project progress and quality indicators. Monthly evaluations were conducted, and feedback from the construction team and the owner was collected. The scheduling suggestions, combined with those from the DQN output, were reviewed by the project management committee before adjustments were implemented, forming a closed-loop optimization process. In the cable laying project, a distributed storage system with a capacity of 100TB was used, supporting 1000 concurrent read / write operations per second. Image recognition used a ResNet50 network with an accuracy of 95%; OCR used the Tesseract engine with a recognition rate of 98%; audio and video transcription used the DeepSpeech model, with a word error rate controlled within 6%. The blockchain adopted the Hyperledger Fabric framework, processing 500 transactions per second. The supervision operation record system, based on the Firebase platform, collects approximately 100,000 user behavior data points daily, including 5,000 review operations. Natural language processing uses the BERT model, achieving an F1 score of 0.85 for key idea extraction. The construction change management platform employs Git version control, supporting 100 concurrent branches. The FP-Growth algorithm has a minimum support of 0.05 and a confidence level of 0.7, generating 300 association rules. The Apache Spark cluster comprises 20 nodes, achieving a processing speed of 1TB / hour. The ARIMA model has parameters (2,1,2) and a prediction accuracy of 85%. The DQN network consists of three fully connected layers, each with 128 neurons. The ε-greedy strategy has an initial ε value of 0.9, decaying by 0.995 per round. The state space has a dimension of 50, representing five resource allocation scenarios across ten regions; the action space has a dimension of 100, representing different personnel deployment schemes.The reward function comprehensively considers both progress achievement rate and quality pass rate, with weights of 0.6 and 0.4 respectively. The model can be updated weekly, generating a scheduling optimization report monthly, containing 20 specific recommendations. The project management committee reviews the recommendations through a visual interface, with an average adoption rate of 80%.

[0037] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to specific implementations. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for resource allocation and optimized scheduling of supervision resources, characterized in that, The methods include: Acquire cable laying path data, construct a 3D visualization model, match the attributes of supervisors with the supervision areas in the model, and dynamically adjust the allocation of supervisors based on the correlation analysis between historical electrical engineering quality problems and supervisor attributes. In concealed locations along the cable laying path, real-time data on environmental parameters, equipment status, and personnel location at the cable laying site are collected to visualize and monitor the construction process. In case of anomalies, warnings are sent to the appropriate supervisors, and the abnormal locations are highlighted in the 3D model. Images and videos of cable fireproofing and insulation resistance testing at the construction site are analyzed to extract key information on the fireproofing materials and resistance testing instruments. This key information is then compared with the cable attributes at the corresponding locations in the 3D visualization model to determine whether the construction quality meets the standards. If the fireproofing materials are found to be inconsistent or the insulation resistance test is abnormal, appropriate supervisors are assigned to handle the issue based on its severity. The verification results are then fed back to the 3D visualization model, and the cable status is updated. Receive update instructions for cable laying construction, perform semantic understanding of the update content, and assess the feasibility of the update based on the skill level of the supervisors; digitize and store the data of cable laying construction, record the review opinions and operation logs of the supervisors, include the supervision records of cable laying construction changes in the storage, and continuously evaluate and dynamically adjust the configuration of supervisors to form a closed-loop optimization. The process of acquiring cable laying path data, constructing a three-dimensional visualization model, matching the attributes of supervisors with the supervision areas in the model, and dynamically adjusting the allocation of supervisors based on the correlation analysis between historical electrical engineering quality problems and supervisor attributes includes: Based on the electrical engineering design drawings, obtain the cable laying path data and use AutoCAD software to build a three-dimensional visualization model; Risk level assessment is performed on the 3D visualization model. The weights of engineering difficulty coefficient and equipment complexity factors are calculated using the analytic hierarchy process to obtain the risk score for each supervision area. For risk scoring, K-means clustering algorithm is used to classify historical electrical engineering quality problems and establish a correlation model between quality problems and the attributes of supervisors. After the correlation model is established, the optimization strategy for the configuration of supervisors is obtained by training a BP neural network. The input layer of the BP neural network includes the features of the supervisors' attributes, and the output layer is the probability of quality problems occurring. Match the attributes of the supervisors with the supervisory areas in the 3D visualization model to generate an initial supervisor configuration plan; Based on the initial personnel configuration plan, the frequency of quality problems in each supervised area was statistically analyzed; If the frequency of quality problems exceeds the preset threshold, the dynamic adjustment process for supervisors will be initiated. The system collects real-time data on environmental parameters, equipment status, and personnel location at concealed locations along the cable laying path. This allows for visualized monitoring of the construction process, sending alerts to appropriate supervisors in case of anomalies, and highlighting abnormal locations in the 3D model. This includes: The system acquires environmental parameters and equipment operating status data collected by a wireless sensor network, which is deployed in concealed locations, including cable trenches and cable wells. Generate sensor data containing sensor ID, timestamp, measurement value, and battery level based on the collected data; The mean, standard deviation, and rate of change of sensor data are calculated using a sliding time window, and abnormal states are identified through preset threshold rules and the isolated forest algorithm. Based on the identified abnormal status, query the supervisor's information from the pre-set diagram database; The system identifies matching supervisors and sends them alerts containing information about the anomaly type, location, severity, and time of occurrence. The 3D scene is rendered after associating abnormal state data with spatial location information in the 3D visualization model. Calculate the world coordinates of the anomaly location and adjust the red highlight intensity based on the distance from the anomaly point.

2. The method according to claim 1, characterized in that, The method further includes: Online learning platforms are used to conduct skills training for supervisors, and virtual reality simulations are used for practical exercises. Supervisors' operations are scored, and the training results are linked to their attributes to optimize personnel scheduling strategies. Based on the specific circumstances of the supervision area, the matching degree between the attributes of the supervisors and the supervision tasks is analyzed. Combining the importance weights of different professional fields, as well as the work experience and skill level of the supervisors, an objective function and constraints are constructed to solve for the optimal supervisor configuration scheme.

3. The method according to claim 2, characterized in that, The aforementioned method involves conducting skills training for supervisory personnel through an online learning platform, performing practical exercises via virtual reality simulation, evaluating the supervisors' performance, and linking training results with personnel attributes to optimize personnel scheduling strategies. This includes: The process involves acquiring training course content for construction supervisors, setting up learning modules with multiple difficulty levels based on the course content, and constructing the relationships between learning content using a course content association matrix. 3D modeling software is used to convert engineering design drawings into 3D visualization models. Based on these models, construction scenarios for cable laying and grounding devices are built, transforming construction specifications into executable virtual operation procedures. The operational behaviors of construction supervisors within these virtual operation procedures are recorded, and the similarity between these behaviors and preset standard operation procedures is calculated. Training results are obtained based on the operational similarity, and these results are correlated with the supervisors' attribute data to construct a skills evaluation index system. A genetic algorithm is used to calculate the optimal personnel scheduling plan. If project progress and quality feedback change, personnel allocation is dynamically adjusted accordingly to achieve continuous optimization of the construction supervisor configuration.

4. The method according to claim 2, characterized in that, The method analyzes the matching degree between the professional qualifications of the supervisors and the supervisory tasks, taking into account the specific circumstances of the supervised area. It constructs an objective function and constraints, considering the importance weights of different professional fields, as well as the supervisors' work experience and skill levels, to solve for the optimal supervisory personnel configuration scheme, including: The process involves obtaining expert questionnaire data, calculating the importance weights of each professional field, and obtaining a professional demand score vector for each supervision area. Based on this score vector, data is retrieved from a database of supervisors' professional qualifications, work experience, and skill levels. A multi-index evaluation method is used to quantify the comprehensive capabilities of each supervisor, resulting in a supervisor capability score vector. A weighted cosine similarity algorithm is employed to calculate the similarity between the supervisor capability score vector and the supervision area professional demand score vector, yielding a matching score matrix for each supervision area. After obtaining the matching score matrix, an integer programming model is established with the objective function of maximizing the overall matching degree. The decision variable of the integer programming model is the allocation relationship between supervisors and supervision areas. Based on the integer programming model, a branch and bound algorithm is used to solve for the optimal supervisor configuration scheme, outputting the personnel allocation results for each supervision area.

5. The method according to claim 1, characterized in that, The analysis of images and videos of cable fireproofing and insulation resistance testing at the construction site, extracting key information on the sealing materials and resistance testing instruments, and comparing this key information with the cable attributes at corresponding locations in the 3D visualization model, determines whether the construction quality meets the specifications and standards. This includes: Acquire video data of the cable fireproofing and insulation resistance testing process; Keyframes are extracted from the video data to obtain a representative image of one frame per second; Representative images are processed using target detection algorithms to identify sealing materials and testing instruments, resulting in labeled images. Extract text information from the labeled images. The text information includes the type, specifications and quantity of the sealing material, as well as the model, range, accuracy and calibration date of the resistance testing instrument. Based on the spatiotemporal information in the annotated images, the text information is associated with the cable locations in the 3D visualization model to obtain associated data; Build a rule engine and input related data into the rule base. The rule base contains rule sets for material specification matching, process parameter qualification, and instrument accuracy requirements. Based on the output of the rules engine, the quality of construction is assessed to determine its compliance, and a quality assessment report is generated.

6. The method according to claim 1, characterized in that, If the fireproof sealing material is found to be substandard or the insulation resistance test is abnormal, appropriate supervisory personnel will be assigned to handle the issue according to its severity. The verification results will be fed back to the 3D visualization model and the cable status will be updated, including: Convolutional neural networks are used to perform image recognition of fireproof sealing materials and obtain material specification information from the design database. A similarity score is calculated based on the image recognition results and material specifications. If the similarity score is lower than a preset threshold, the fireproof sealing material is deemed to be non-compliant. Calculate the deviation between the test data and the standard value based on the test data collected by the insulation resistance tester; The degree of deviation is determined based on the deviation, resulting in a severity index; Construct a capability assessment model for supervisors. The input features of the capability assessment model include the professional qualifications, work experience, and historical handling results of supervisors. The professional competence score of the supervisors is output through the supervisor competence assessment model; Based on the severity index of the abnormal situation information, match the supervisor with the corresponding professional ability score; Information about any abnormal situations will be pushed to the matched supervisors. Real-time communication protocols are used to synchronize on-site verification data to a 3D visualization model and update cable status information.

7. The method according to claim 1, characterized in that, The received cable laying construction update instruction performs semantic understanding of the update content, and assesses the feasibility of the update based on the skill level of the supervisor, including: Receive update instructions, encrypt the update instructions, and generate an integrity verification value for the update instructions; Natural language processing is performed on the encrypted update instructions, including word segmentation, part-of-speech tagging, and named entity recognition, to extract specified information from the design change instructions; Based on the specified information, a supervisory personnel competency assessment model is constructed. The supervisory personnel competency assessment model adopts a multilayer perceptron algorithm. The input layer contains a preset number of nodes, the two hidden layers each contain a preset number of nodes, and the output layer contains a preset number of nodes to obtain the supervisory personnel's skill level score and communication ability score. Based on the specified information of the update instruction, the skill level score and communication ability score of the supervisor, a decision tree model is constructed. The decision nodes of the decision tree model include change complexity, material availability and supervisor ability matching degree. The optimal splitting feature is selected by information gain ratio, and a pessimistic pruning method is used to prevent overfitting, so as to obtain the change feasibility assessment result.

8. The method according to claim 1, characterized in that, The digital storage of cable laying construction data includes recording the supervisor's review opinions and operation logs, incorporating supervision records of cable laying construction changes into the data storage, and continuously evaluating and dynamically adjusting the allocation of supervisory personnel to form a closed-loop optimization, including: Obtain cable laying construction data and store the construction data in a distributed file storage system; Calculate the hash value and timestamp of each document based on the construction data in the distributed file storage system; Record the hash value and timestamp to obtain an immutable digital evidence chain; Receive the review operations and system interaction behavior of the supervisors, and record the click path, dwell time and operation frequency; Obtain information on change applications, review processes, and implementation status, and incorporate this information into the scope of digital evidence storage; Record the details of each change; Establish rules for linking changes to supervisory personnel; Obtain historical performance data of supervisors and perform large-scale parallel processing on this historical performance data; Based on the results of parallel processing, evaluate whether the supervision configuration plan meets the preset requirements; If not, the personnel scheduling strategy is optimized through a deep Q-learning network. The state space of the deep Q-learning network includes the current personnel distribution and workload, while the action space contains personnel allocation options.

Images