Power communication network field operation and maintenance work order scheduling method and device

Abstract

The invention relates to the technical field of digital data processing, in particular to an electric power communication network field operation and maintenance work order scheduling method and device, and the method comprises the steps: receiving and summarizing electric power communication fault information from a plurality of regions; classifying the fault information by adopting a support vector machine and generating an operation and maintenance work order; matching maintenance personnel with different maintenance skill levels and in a maintenance state based on personnel constraint conditions, and sending a scheduling notification; matching a solution corresponding to the fault information by using a historical database, and pushing the solution to a dispatched maintainer; and evaluating the scheduling process based on the fault processing time and the fault processing completion degree of the maintenance personnel. Therefore, the response speed and the processing efficiency of power communication network maintenance can be improved, the scheduling logic can be continuously optimized through continuously accumulated empirical data, and the overall service quality and customer satisfaction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of digital data processing, and particularly to a method and device for dispatching on-site operation and maintenance work orders in a power communication network. Background Art

[0002] The dispatching of on-site operation and maintenance work orders in a power communication network refers to the process of, when a fault occurs or regular maintenance is required in the power communication network, allocating specific repair tasks to corresponding technical personnel through a professional work order management system. This process starts from fault detection, collects fault information through automated tools or manual means, and forms a repair work order. Subsequently, according to the urgency and type of the fault, the dispatching system will reasonably arrange repair personnel and resources to ensure that problems are solved in a timely and effective manner. The entire process also includes steps such as work order status tracking, feedback on repair results, and subsequent analysis and optimization to improve operation and maintenance efficiency and service quality, and ensure the stable operation of the power communication network.

[0003] In the existing operation and maintenance work order dispatching methods, less attention is paid to human factors, and the differences in skills and capabilities of human resources are ignored. This problem severely limits the application of the dispatching algorithm model, resulting in the inability to reasonably and effectively utilize the existing operation and maintenance resources, thereby reducing the efficiency of operation and maintenance operations. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and device for dispatching on-site operation and maintenance work orders in a power communication network, aiming to improve the response speed and processing efficiency of power communication network maintenance, and continuously optimize the dispatching logic through the continuously accumulated experience data, thereby enhancing the overall service quality and customer satisfaction.

[0005] To achieve the above purpose, in a first aspect, the present invention provides a method for dispatching on-site operation and maintenance work orders in a power communication network, including receiving and summarizing power communication fault information from multiple regions; classifying the fault information using a support vector machine and generating operation and maintenance work orders; matching repair personnel with different repair skill levels and in a repair state based on personnel constraint conditions, and sending a dispatching notice; matching a solution corresponding to the fault information using a historical database and pushing it to the dispatched repair personnel; evaluating the dispatching process based on the fault handling time and fault handling completion degree of the repair personnel.

[0006] Among them, the specific steps of receiving and summarizing power communication fault information from multiple regions include: real-time collecting network status data through monitoring devices installed in the power communication network, and automatically uploading fault information when an abnormal situation is detected; sorting the collected fault information into a fault table and confirming it through the user. Store the collected fault information in the central database.

[0007] Among them, the fault information includes the fault location, fault type, discovery time, and fault description.

[0008] Among them, the specific steps of using the support vector machine to classify the fault information and generate an operation and maintenance work order include: Classify the fault information into different categories according to the type, severity, and impact scope of the fault; Collect historical fault information from historical records; Extract fault features from the historical fault information, and the fault features include fault location codes, fault type codes, and impact scope codes; Set classification labels for the fault information and label the historical fault information to assign corresponding labels to each fault information; Set the support vector machine as the classification model; Convert the extracted fault features into numerical vectors and generate a training set; Use the training set data to train the support vector machine model; Collect current fault information and extract fault features to convert them into real-time numerical vectors; Use the trained support vector machine model to classify the real-time numerical vectors, output the type, severity, and impact scope of the fault, and assign a unique work order number to generate an operation and maintenance work order.

[0009] Among them, the fault types in the classification labels include communication interruption, equipment failure, and software error, the severity levels include level one, level two, level three, and level four, and the impact scopes include widespread impact, local impact, and individual impact.

[0010] Among them, the specific steps of matching maintenance personnel with different maintenance skill levels and in the maintenance state based on personnel constraint conditions and sending a dispatch notice include: Set the skill levels of the maintenance personnel and obtain the current maintenance status of the personnel; Use the dispatch algorithm to match maintenance personnel according to the operation and maintenance work order and personnel constraint conditions; Notify the matched maintenance personnel to go to the fault location through electronic communication means.

[0011] Among them, the specific steps of using the dispatch algorithm to match maintenance personnel according to the operation and maintenance work order and personnel constraint conditions include: According to the skill levels and professional fields required by the work order, screen out the target group of maintenance personnel that meet the conditions; Calculate the distance of the maintenance personnel from the fault location in the target group of maintenance personnel and obtain the distance score; Calculate the current workload of each maintenance staff and obtain a workload score; Confirm the remaining working time of the maintenance staff and obtain a time score; Based on the historical maintenance staff matching results, use a convolutional neural network model to calculate the weights of each scoring item; Calculate a comprehensive score based on the distance score, workload score, waiting time score, and the weights of each scoring item, and select the maintenance staff with the highest comprehensive score as the best matching object.

[0012] In a second aspect, the present invention also provides a field operation and maintenance work order scheduling device for a power communication network, including an information acquisition module, a work order generation module, a personnel matching module, a solution push module, and an evaluation module; The information acquisition module is used to receive and summarize power communication fault information from multiple regions; The work order generation module is used to classify the fault information using a support vector machine and generate an operation and maintenance work order; The personnel matching module is used to match maintenance staff with different maintenance skill levels and in a maintenance state based on personnel constraint conditions and issue a scheduling notice; The solution push module is used to match the solution corresponding to the fault information using a historical database and push it to the scheduled maintenance staff; The evaluation module is used to evaluate the scheduling process based on the fault handling time and fault handling completion degree of the maintenance staff.

[0013] A field operation and maintenance work order scheduling method and device of the present invention enable the system to receive fault reports of power communication equipment from multiple different regions. These reports may come from various channels, such as an automatic detection system, user feedback, etc. The system organizes and summarizes these scattered information to form a comprehensive fault information list. Then, the support vector machine (SVM) algorithm in machine learning is used to classify the collected fault information. SVM is a supervised learning model that can effectively find the best decision boundary in a high-dimensional space to achieve accurate data classification. According to the classification results, the system automatically generates corresponding operation and maintenance work orders, and each work order details information such as the specific situation of the fault, the occurrence location, and the initially judged fault type.

[0014] The system will intelligently select the most suitable maintenance personnel to handle specific faults based on factors such as the current list of available maintenance personnel, their skill levels, and work status. This step not only considers the professional skill levels of the maintenance personnel but also takes into account their current workload and personal work preferences to ensure that each work order can be handled by the most appropriate personnel. Once the maintenance personnel are determined, the system will immediately send a scheduling notice to the relevant personnel. To speed up the fault handling process, the system will also query the internal historical database to find historical cases similar to the current fault and their solutions. After finding them, the system will directly push these solutions to the dispatched maintenance personnel to help them locate the problem faster and take effective measures to solve it. The system will evaluate the entire scheduling process, mainly based on two indicators: one is the time required for the maintenance personnel to complete the task from receiving the task, and the other is the completion degree of the fault handling. By analyzing these two indicators, the effectiveness of the scheduling plan can be effectively evaluated, providing data support for optimizing the scheduling strategy in the future.

[0015] This method not only improves the response speed and processing efficiency of power communication network maintenance but also can continuously optimize the scheduling logic through the accumulated experience data, thereby enhancing the overall service quality and customer satisfaction. Brief Description of the Drawings

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of a method for dispatching on-site operation and maintenance work orders in a power communication network of the present invention.

[0018] Figure 2 It is a flowchart of receiving and summarizing power communication fault information from multiple regions of the present invention.

[0019] Figure 3 It is a flowchart of classifying the fault information using a support vector machine and generating operation and maintenance work orders of the present invention.

[0020] Figure 4 It is a flowchart of matching maintenance personnel with different maintenance skill levels and in a maintenance state based on personnel constraint conditions and sending a scheduling notice of the present invention.

[0021] Figure 5 It is a flowchart of using a scheduling algorithm to match maintenance personnel according to operation and maintenance work orders and personnel constraint conditions of the present invention. Detailed Embodiment

[0022] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where like or similar reference numerals denote like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0023] First Embodiment Please refer to Figures 1 to 5 , the present invention provides a method for scheduling on-site operation and maintenance work orders in a power communication network, including: S101 Receive and aggregate power communication fault information from multiple regions; The specific steps include: S201 Real-time collect network status data through monitoring devices installed in the power communication network, and automatically upload fault information when an abnormal situation is detected; Install various monitoring devices, such as sensors, data collectors, and monitoring cameras, on key nodes and devices in the power communication network. These devices can real-time collect network status data, including but not limited to signal strength, transmission rate, device temperature, current and voltage, etc. The monitoring devices are connected to the central monitoring system through wired or wireless means to transmit the collected data in real-time. The data collection frequency can be adjusted according to actual needs, usually once per second or once per minute.

[0024] The central monitoring system performs real-time analysis on the received data, and uses preset thresholds to detect abnormal situations. Once an abnormality is detected, the system will automatically trigger a fault alarm. When an abnormal situation is detected, the system will automatically generate fault information and upload it to the central database. The fault information includes but not limited to the following: Fault location: The specific location where the fault occurs, such as a substation, a transmission line, a distribution device, etc.

[0025] Fault type: The type of the fault, such as communication interruption, equipment failure, software error, etc.

[0026] Discovery time: The time when the fault is detected.

[0027] Fault description: The specific manifestations and related phenomena of the fault, such as signal loss, equipment shutdown, data transmission failure, etc.

[0028] S202 Organize the collected fault information into a fault form and confirm it through the user; Format the fault information uploaded to the central database to ensure data consistency and readability. Generate a fault form, and the form contains the following: Work order number: Generate a unique work order number for each fault information.

[0029] Fault location: The specific location where the fault occurs.

[0030] Fault type: The type of the fault.

[0031] Discovery time: The time when the fault is detected.

[0032] Fault description: The specific manifestations and related phenomena of the fault.

[0033] Status: The processing status of the fault, such as pending, in progress, completed, etc.

[0034] Push the generated fault form to relevant users or technicians through electronic communication means (such as email, SMS, mobile application), and request the users to confirm the accuracy of the fault information. The users can confirm or modify the fault information by replying. According to the user feedback, verify and correct the fault information to ensure the accuracy and integrity of the information.

[0035] S203 stores the collected fault information in the central database.

[0036] Store the fault information confirmed by the users in the central database to ensure centralized management and backup of the data. The central database should have high availability and high reliability to ensure the security and integrity of the data. Before storing the fault information, perform data verification to ensure the integrity and consistency of the data. The verification content includes whether fields are missing, whether the data format is correct, etc. Regularly back up the central database to prevent data loss. The backup data should be stored in a secure physical or cloud storage environment. Set data access permissions to ensure that only authorized personnel can access and modify the fault information. Use authentication and access control mechanisms to ensure the security of the data.

[0037] Through the above steps, it can be ensured that the power communication fault information from multiple regions is effectively received, aggregated and stored, providing reliable data support for subsequent work order generation and fault handling.

[0038] S102 uses a support vector machine to classify the fault information and generate an operation and maintenance work order; The specific steps include: S301 classifies the fault information into different categories according to the type, severity and impact scope of the fault; Fault type classification: Classify the fault information into different categories according to the type, such as communication interruption, equipment failure, software error, etc.

[0039] Severity classification: Classify it into different levels according to the severity of the fault, such as level 1 (urgent), level 2 (high), level 3 (medium), level 4 (low).

[0040] Impact scope classification: Classify faults into different categories according to their impact scope, such as widespread impact, local impact, and individual impact.

[0041] S302 Collect historical fault information from historical records; Collect a large amount of fault information from historical records, including fault location, fault type, discovery time, fault description, handling method, handling result, etc. Remove invalid or incomplete data to ensure the quality and consistency of the data.

[0042] S303 Extract fault features from the historical fault information, where the fault features include fault location code, fault type code, and impact scope code; Encode the fault location, such as encoding Substation A as 1, Substation B as 2, etc. Encode the fault type, such as encoding communication interruption as 1, equipment failure as 2, software error as 3, etc. Encode the impact scope, such as encoding widespread impact as 1, local impact as 2, individual impact as 3, etc.

[0043] S304 Set classification labels for the fault information and label the historical fault information to assign corresponding labels to each fault information; Fault types include communication interruption, equipment failure, software error; Severity levels include level 1 (urgent), level 2 (high), level 3 (medium), level 4 (low); Impact scopes include widespread impact, local impact, individual impact Manually label the historical fault information and assign corresponding labels to each fault information to ensure the accuracy and consistency of the labels.

[0044] S305 Set the support vector machine as the classification model; Select the support vector machine (SVM) as the classification model. SVM is an effective supervised learning algorithm. Select a suitable kernel function (such as linear kernel, polynomial kernel, RBF kernel, etc.) and set hyperparameters such as the regularization parameter C.

[0045] S306 Convert the extracted fault features into numerical vectors and generate a training set; Convert the extracted fault features (fault location code, fault type code, impact scope code, etc.) into numerical vectors for input into the SVM model.

[0046] Combine the labeled historical fault information and the corresponding numerical vectors into a training set, and the training set is used to train the SVM model.

[0047] S307 Use the training set data to train the support vector machine model; Train the SVM model using the training set data, and adjust the model parameters through iterative optimization algorithms (such as the SMO algorithm) to minimize the classification error of the model on the training set. Evaluate the performance of the model using methods such as cross-validation, and calculate metrics such as accuracy, recall rate, and F1 score to ensure the generalization ability of the model.

[0048] S308 Collect the current fault information and extract the fault features and convert them into a real-time numerical vector; Collect new fault information in real time, including the fault location, fault type, discovery time, fault description, etc. Extract the same features from the real-time fault information, such as fault location encoding, fault type encoding, impact scope encoding, etc. Convert the extracted features into a numerical vector to generate a real-time numerical vector.

[0049] S309 Use the trained support vector machine model to classify the real-time numerical vector, output the type, severity, and impact scope of the fault, and assign a unique work order number to generate an operation and maintenance work order.

[0050] Use the trained SVM model to classify the real-time numerical vector, and output the type, severity, and impact scope of the fault.

[0051] According to the prediction results of the model, generate an operation and maintenance work order, and the work order content includes: Work order number: Generate a unique work order number for each fault information.

[0052] Fault location: Record the specific location where the fault occurred.

[0053] Fault type: Fill in according to the prediction results of the SVM model.

[0054] Severity: Fill in according to the prediction results of the SVM model.

[0055] Impact scope: Fill in according to the prediction results of the SVM model.

[0056] Discovery time: Record the time when the fault was discovered.

[0057] Fault description: Describe in detail the specific manifestations and related phenomena of the fault.

[0058] Processing requirements: According to the type and severity of the fault, put forward specific processing requirements and time limits.

[0059] Dispatch time: Record the time when the work order is generated and dispatched.

[0060] S103 Match maintenance personnel with different maintenance skill levels and in the maintenance state based on personnel constraint conditions, and issue a dispatch notice; The specific steps include: S401 Set the skill level of maintenance personnel and obtain their current maintenance status; Define different skill levels according to the experience and training of maintenance personnel, such as junior, intermediate, senior, etc. Evaluate the skill levels of maintenance personnel through examinations, assessments, or historical performance, etc., and record them in the personnel files. Regularly conduct skill assessments and training for maintenance personnel, and adjust their skill levels according to the assessment results.

[0061] Record the current status of each maintenance personnel in the central database, such as idle, busy, on leave, etc. Through mobile applications or electronic communication means, update the status of maintenance personnel in real time to ensure the accuracy and timeliness of information.

[0062] Idle means there is no current task and can immediately accept new tasks.

[0063] Busy means is currently handling other tasks and needs to wait until the current task is completed before accepting new tasks.

[0064] On leave means not in a working state and unable to accept new tasks.

[0065] S402 Use a scheduling algorithm to match maintenance personnel according to operation and maintenance work orders and personnel constraints; The specific steps include: S501 Screen out a group of target maintenance personnel who meet the requirements according to the skill level and professional field required by the work order.

[0066] Determine the minimum skill level required for maintenance personnel according to the type of fault involved in the work order. For example, for complex or high-level faults, senior engineers may be required to handle them.

[0067] Identify the professional field to which the fault belongs, such as fiber optic communication, switch configuration, network security, etc., to ensure that maintenance personnel have professional knowledge in related fields.

[0068] Maintain a database containing the basic information of all maintenance personnel, such as name, contact information, skill level, professional field, current working status (idle / busy), geographical location, etc. Ensure that the working status of maintenance personnel is updated in real time so that the system can accurately understand which personnel are currently available for scheduling.

[0069] According to the requirements of the work order, screen out the personnel who meet the skill level and professional field requirements from the maintenance personnel database. On the basis of the preliminary screening, exclude those maintenance personnel who are currently performing other tasks or are not within the service scope. For geographical location requirements, it can be determined according to the distance between the location where the fault occurred and the current location of the maintenance personnel.

[0070] S502 calculates the distance of the maintenance personnel from the fault location in the target maintenance personnel group and obtains a distance score; Obtain the current location of the maintenance personnel and the location of the fault location from the central database. Use a Geographic Information System (GIS) or a map API to calculate the distance between the current location of the maintenance personnel and the fault location.

[0071] Scoring formula: Distance score = (Maximum distance - Actual distance) / Maximum distance × 100 Example: Assume the maximum distance is 50 kilometers and the actual distance is 10 kilometers, then the distance score is: (50 - 10) / 50 × 100 = 80; S503 calculates the current workload of each maintenance personnel and obtains a workload score; Obtain the current workload of the maintenance personnel from the central database, such as the current number of tasks, estimated completion time, etc.

[0072] Based on the workload data, calculate the workload percentage of each maintenance personnel.

[0073] Scoring formula: Workload score = (Maximum workload - Actual workload) / Maximum workload × 100 Example: Assume the maximum workload is 100% and the actual workload is 30%, then the workload score is: (100 - 30) / 100 × 100 = 70 S504 confirms the remaining working time of the maintenance personnel and obtains a time score; Obtain the tasks that the maintenance personnel are currently handling and their estimated completion times from the central database.

[0074] Calculate the remaining time: Calculate the remaining working time of each maintenance personnel's current task.

[0075] Scoring formula: Time score = (Earliest available time - Current time) / (Earliest available time - Latest available time) × 100 Example: Assume the current time is 10:00, the earliest available time is 10:30, and the latest available time is 12:00, then the time score is: (10:30 - 10:00) / (10:30 - 12:00) × 100 = 66.67 S505 calculates the weights of each scoring item using a convolutional neural network model based on historical maintenance personnel matching results; The specific steps include: S601 collects historical recommendation records, including the recommended maintenance personnel, work order information, and recommendation results (whether the fault was successfully handled); S602 records the values of each scoring item in each recommendation, including distance score, workload score, waiting time score; Record the values of each scoring item in each match, including distance score, load score, time score, etc.

[0076] S603 Extract features from historical recommendation data; Extract features from historical matching data, including work order information (fault location, fault type, severity, impact scope, etc.), maintenance personnel information (skill level, current location, workload, available time, etc.) and scoring item values.

[0077] S604 Generate labels according to the recommendation results, such as 1 for successfully handling the fault and 0 for unsuccessfully handling the fault; S605 Construct and train a convolutional neural network model; The input layer receives the extracted features, including work order information, maintenance personnel information and scoring item values.

[0078] Convolutional layer: Use one-dimensional or two-dimensional convolutional layers to extract local patterns of features.

[0079] Pooling layer: Use the pooling layer to reduce the feature dimension and retain important information.

[0080] Fully connected layer: Use the fully connected layer to map the extracted features to the output layer.

[0081] Output layer: The output layer uses the sigmoid activation function to output the probability of successful matching.

[0082] Loss function: Use the binary cross-entropy loss function to measure the difference between the model prediction value and the true label.

[0083] Optimizer: Use optimizers such as Adam to adjust the model parameters to minimize the loss function.

[0084] Training process: Divide the preprocessed dataset into a training set and a validation set, use the training set data to train the model, and use the validation set data to evaluate the model performance.

[0085] Hyperparameter tuning: Adjust hyperparameters such as the number of convolutional layers, the size of the convolutional kernel, the size of the pooling layer, and the number of nodes in the fully connected layer to obtain the best model performance.

[0086] S606 Use the trained CNN model to calculate the gradient of each scoring item on the model output through the gradient algorithm to obtain the corresponding weights.

[0087] Use the trained CNN model to calculate the weights of each scoring item through gradient or feature importance analysis methods. Calculate the gradient of each scoring item on the model output. The larger the gradient, the greater the impact of the scoring item on the matching result.

[0088] S506 calculates a comprehensive score based on the distance score, load score, waiting time score, and the weights of each scoring item, and selects the maintenance personnel with the highest comprehensive score as the best matching object.

[0089] Comprehensive score = w1 × distance score + w2 × load score + w3 × waiting time score Example: Suppose there are three maintenance personnel A, B, and C: Maintenance personnel A: Distance score: 80 Load score: 70 Waiting time score: 66.67 Comprehensive score: 0.45×80 + 0.35×70 + 0.20×66.67 = 36 + 24.5 + 13.334 = 73.834 Maintenance personnel B: Distance score: 60 Load score: 50 Waiting time score: 60 Comprehensive score: 0.45×60 + 0.35×50 + 0.20×60 = 27 + 17.5 + 12 = 56.5 Maintenance personnel C: Distance score: 70 Load score: 80 Waiting time score: 68.75 Comprehensive score: 0.45×70 + 0.35×80 + 0.20×68.75 = 31.5 + 28 + 13.75 = 73.25 S403 notifies the matched maintenance personnel to go to the fault location through electronic communication means.

[0090] Generate a notification template, which includes: Work order number: Ensure that the unique number of the work order is included in the notification for the maintenance personnel to quickly search and confirm.

[0091] Fault location: Clearly define the specific location where the fault occurred, such as the name and address of the substation.

[0092] Fault type: Describe in detail the type of the fault, such as communication interruption, equipment failure, software error, etc.

[0093] Severity level: Indicate the severity level of the fault, such as level 1 (urgent), level 2 (high), level 3 (medium), level 4 (low).

[0094] Influence scope: Describe the influence scope of the fault, such as wide influence, local influence, individual influence.

[0095] Discovery Time: Record the time when the fault was discovered.

[0096] Handling Requirements: Clearly define the requirements and time limits for fault handling, such as immediately dispatching senior technicians to the site for handling and ensuring communication restoration within 2 hours.

[0097] Dispatch Time: Record the time when the work order is dispatched.

[0098] Contact Information: Provide the contact information of the fault reporter for the maintenance personnel to communicate when necessary.

[0099] Maintenance Personnel Information includes the name, employee number, contact information, etc. of the maintenance personnel, the skill level and professional field of the maintenance personnel, and provide specific handling suggestions and precautions.

[0100] Send a text message through the SMS platform to ensure that the maintenance personnel can receive the notice promptly.

[0101] Example: Work order number 20241121-0001, fault location: a certain substation, fault type: communication interruption, handling requirements: immediately go to the site for handling and ensure communication restoration within 2 hours. Contact person: Zhang San, phone number: 1234567890.

[0102] In addition, an email can also be sent through the enterprise email system to ensure that the maintenance personnel can view and save the notice in a timely manner.

[0103] S104 uses the historical database to match the solution corresponding to the fault information and push it to the dispatched maintenance personnel; The specific steps include: S701 collects historical fault records, including fault location, fault type, fault description, handling methods, handling results, etc.; This step involves a comprehensive review of all past fault events, including but not limited to the specific location of the fault, the type of the fault (such as hardware fault, software fault, etc.), the detailed description of the fault, the handling methods taken, and the final handling results. These data form the basis for subsequent analysis and are crucial for building an efficient and accurate solution library.

[0104] S702 establishes a solution library to record the common solutions and handling steps for each fault type; Based on the data collected in S701, a knowledge base containing various fault types and their corresponding solutions needs to be created next. This library should detail the common handling steps and suggestions for each fault for quick lookup and application. In addition, as new fault cases are continuously added, the solution library should be updated regularly to maintain its timeliness and effectiveness.

[0105] S703 extracts information about the current fault from the generated operation and maintenance work order; When a new fault occurs, the system generates an operation and maintenance work order. From this work order, all relevant information about the current fault needs to be extracted, such as the device where the fault occurred, the fault manifestation, the preliminary diagnosis result, etc. This information will be used in the next processing and matching processes.

[0106] S704 extracts keywords and phrases from the fault description and performs word segmentation and part-of-speech tagging using natural language processing (NLP) techniques; Using natural language processing (NLP) techniques, the fault description extracted in S703 is deeply analyzed. This step includes performing word segmentation on the text, identifying and marking important words and phrases, and at the same time performing part-of-speech tagging to better understand the core content of the fault description. In this way, the essential characteristics of the fault can be captured more accurately.

[0107] S705 uses similarity calculation methods (such as cosine similarity, Jaccard similarity, etc.) to compare the current fault information with historical fault records and find the most similar historical fault records; Adopt a suitable similarity algorithm (such as cosine similarity, Jaccard similarity, etc.) to compare the current fault information after NLP processing with historical fault records. The purpose is to find those historical cases that are closest to the current situation. This step is the core of the entire process, which directly determines the relevance and effectiveness of the recommended solutions.

[0108] S706 extracts the corresponding solutions and processing steps from the matched historical fault records.

[0109] Once the most similar historical fault record is determined, the corresponding solutions and processing steps can be extracted from it. This information will then be organized into an easy-to-understand format and sent to the technical personnel responsible for maintenance through appropriate communication channels (such as email, text message, or a dedicated application). In this way, the maintenance personnel can quickly obtain the required guidance, thereby accelerating the speed of fault troubleshooting and reducing downtime.

[0110] S105 evaluates the scheduling process based on the fault handling time and fault handling completion rate of the maintenance personnel.

[0111] The fault handling time refers to the time period from when the fault is reported or detected until the fault is completely repaired and the normal operation is restored. This time period can be divided into the following parts: Response time: The time from when the fault report is submitted to when the maintenance personnel arrive at the scene or start remote processing.

[0112] Diagnosis time: The time it takes maintenance personnel to make a preliminary diagnosis of a fault and determine the type and cause of the fault.

[0113] Repair time: The time from the start of implementation of repair measures to the complete resolution of the fault.

[0114] Verification time: The time required to conduct functional testing and performance verification after a fault is repaired.

[0115] Recording method: The operation and maintenance management system automatically records the generation time of each fault work order, the arrival time of the maintenance personnel, the fault diagnosis time, the fault repair time and the verification time. These data will be used for subsequent analysis and evaluation.

[0116] Troubleshooting completion refers to the quality and thoroughness of troubleshooting, and is usually evaluated through the following aspects: Whether the fault is completely resolved: After the fault is repaired, whether the system resumes normal operation without any remaining issues.

[0117] User satisfaction: Through user feedback or questionnaires, understand the user's satisfaction with the troubleshooting results.

[0118] System performance recovery: After the fault is repaired, whether the system performance returns to the level before the fault, or whether there is further optimization.

[0119] Completeness of documentation: Is there a detailed record of the troubleshooting process, including the fault phenomenon, handling steps, tools and materials used, etc.

[0120] Evaluation method: Through the operation and maintenance management system, collect the data of the above indicators and make a comprehensive score. The score can be based on a five-point system or a ten-point system. The specific scoring criteria can be adjusted according to the actual situation.

[0121] Second embodiment The present invention also provides an on-site operation and maintenance work order scheduling device for an electric power communication network, comprising an information acquisition module, a work order generation module, a personnel matching module, a solution push module and an evaluation module; the information acquisition module is used to receive and summarize electric power communication fault information from multiple regions; the work order generation module is used to classify the fault information using a support vector machine and generate an operation and maintenance work order; the personnel matching module is used to match maintenance personnel with different maintenance skill levels and in a maintenance state based on personnel constraints, and issue a scheduling notification; the solution push module is used to use a historical database to match the solution corresponding to the fault information, and push it to the scheduled maintenance personnel; the evaluation module is used to evaluate the scheduling process based on the maintenance personnel's fault handling time and fault handling completion.

[0122] In this embodiment, the information acquisition module receives and aggregates power communication fault information from multiple regions. This information can come from various channels, such as reports from on-site inspection personnel, alarms from automated monitoring systems, user feedback, etc. Specifically, fault information in the power communication network is collected through various sensors, monitoring devices, and user terminals. The collected fault information is transmitted to the central server via wired or wireless networks. The central server aggregates and preliminarily processes the received fault information to ensure the integrity and accuracy of the data. The information acquisition module can adopt Internet of Things technology to collect data in real time through various sensors and monitoring devices, and store and process the data through a cloud computing platform.

[0123] The work order generation module classifies the collected fault information and generates corresponding operation and maintenance work orders. This module uses the Support Vector Machine (SVM) algorithm to accurately classify the fault information to ensure the accuracy and effectiveness of the work orders. Specifically, the SVM algorithm is used to classify the fault information to identify the type, severity, and urgency of the faults. According to the classification results, detailed operation and maintenance work orders are generated, including fault descriptions, handling requirements, estimated handling time, etc. The generated work orders are distributed to the corresponding processing queues for the next step of scheduling.

[0124] The personnel matching module matches maintenance personnel with different repair skill levels and in a repair state according to personnel constraints and issues a scheduling notice. This module takes into account factors such as the skills, locations, and current task status of the maintenance personnel to ensure the rationality and efficiency of scheduling. Specifically, a database containing all maintenance personnel is maintained, recording information such as the skill levels, current locations, and current task status of each maintenance personnel. According to the requirements of the operation and maintenance work orders, a matching algorithm (such as a multi-skilled worker scheduling algorithm) is used to select the most suitable maintenance personnel. A scheduling notice is sent to the matched maintenance personnel via text message, phone call, or mobile application, informing them of the task details and requirements. The personnel matching module can adopt a multi-skilled worker scheduling algorithm, comprehensively considering the skills, locations, and task status of the maintenance personnel to ensure the rationality and efficiency of scheduling.

[0125] The solution push module uses the historical database to match the solutions corresponding to the fault information and pushes them to the dispatched maintenance personnel. Specifically, it maintains a database containing historical fault records, recording the common solutions and processing steps for each fault type. Extracts the information of the current fault from the generated operation and maintenance work orders, including the fault location, fault type, fault description, etc. Uses natural language processing (NLP) technology to segment and tag the parts of speech of the fault description, and extracts keywords and phrases. Adopts similarity algorithms (such as cosine similarity, Jaccard similarity, etc.) to compare the current fault information with the historical fault records and find the most similar historical fault record. Extracts the corresponding solutions and processing steps from the matched historical fault records and pushes them to the maintenance personnel through appropriate means.

[0126] The evaluation module is responsible for evaluating the dispatching process based on the fault handling time and the completion degree of fault handling by the maintenance personnel. This module evaluates the effectiveness of the dispatching process through data analysis and user feedback, providing a basis for continuous improvement.

[0127] The full-process management of fault handling is realized through the combination of the above modules. This device not only improves the efficiency and quality of fault handling, but also provides a scientific basis and a mechanism for continuous improvement for operation and maintenance management. Through intelligent and automated means, this device provides a strong guarantee for the stable operation of the power communication network.

[0128] The above-disclosed is only a preferred embodiment of the present invention, and of course, it cannot be used to limit the scope of the rights of the present invention. Those of ordinary skill in the art can understand the whole or part of the processes of implementing the above embodiments, and the equivalent changes made according to the claims of the present invention still fall within the scope covered by the invention.

Claims

1. A method for dispatching on-site operation and maintenance work orders in a power communication network, characterized in that: It includes: Receiving and summarizing power communication fault information from multiple regions; Using a support vector machine to classify the fault information and generate operation and maintenance work orders; Based on personnel constraint conditions, matching maintenance personnel with different maintenance skill levels and in a maintenance state, and sending a dispatch notice; Using the historical database to match the solution corresponding to the fault information and pushing it to the dispatched maintenance personnel; Evaluating the dispatch process based on the fault handling time and fault handling completion degree of the maintenance personnel.

2. The method for dispatching on-site operation and maintenance work orders in a power communication network according to claim 1, characterized in that: The specific steps of receiving and summarizing power communication fault information from multiple regions include: Real-time collecting network status data through monitoring devices installed in the power communication network, and automatically uploading fault information when abnormal conditions are detected; Sorting the collected fault information into a fault table and confirming it through the user; Storing the collected fault information in the central database.

3. The method for dispatching on-site operation and maintenance work orders in a power communication network according to claim 2, characterized in that: The fault information includes the fault location, fault type, discovery time, and fault description.

4. The method for dispatching on-site operation and maintenance work orders in a power communication network according to claim 3, characterized in that: The specific steps of using a support vector machine to classify the fault information and generate operation and maintenance work orders include: Classifying the fault information into different categories according to the type, severity, and impact range of the fault; Collecting historical fault information from historical records; Extracting fault features from the historical fault information, where the fault features include fault location codes, fault type codes, and impact range codes; Setting classification labels for the fault information and annotating the historical fault information to assign corresponding labels to each fault information; Setting a support vector machine as the classification model; Converting the extracted fault features into numerical vectors and generating a training set; Training the support vector machine model using the training set data; Collecting the current fault information and extracting fault features to convert them into real-time numerical vectors; Using the trained support vector machine model to classify the real-time numerical vectors, outputting the type, severity, and impact range of the fault, and assigning a unique work order number to generate an operation and maintenance work order.

5. The method for dispatching on-site operation and maintenance work orders in a power communication network according to claim 4, characterized in that: The fault types in the classification labels include communication interruption, equipment failure, and software error, the severity levels include level one, level two, level three, and level four, and the impact ranges include wide impact, local impact, and individual impact.

6. The method for dispatching on-site operation and maintenance work orders in a power communication network according to claim 5, characterized in that: The specific steps of matching maintenance personnel with different maintenance skill levels and in a maintenance state based on personnel constraint conditions and sending a dispatch notice include: Setting the skill levels of the maintenance personnel and obtaining the current maintenance status of the personnel; Using a dispatch algorithm to match maintenance personnel according to the operation and maintenance work orders and personnel constraint conditions; Notifying the matched maintenance personnel to go to the fault location through electronic communication means.

7. A method for scheduling on-site operation and maintenance work orders in a power communication network, as described in claim 6, characterized in that The specific steps of using the scheduling algorithm to match maintenance personnel according to the operation and maintenance work orders and personnel constraint conditions include: According to the skill level and professional field required by the work order, screen out the target group of maintenance personnel that meet the conditions; Calculate the distance of the maintenance personnel from the fault location in the target group of maintenance personnel and obtain the distance score; Calculate the current workload of each maintenance personnel and obtain the workload score; Confirm the remaining working time of the maintenance personnel and obtain the time score; Use a convolutional neural network model to calculate the weights of each scoring item based on the historical maintenance personnel matching results; Calculate the comprehensive score based on the distance score, workload score, waiting time score and the weights of each scoring item, and select the maintenance personnel with the highest comprehensive score as the best matching object.

8. A device for scheduling on-site operation and maintenance work orders in a power communication network, applied to the method for scheduling on-site operation and maintenance work orders in a power communication network according to any one of claims 1 to 7, characterized in that It includes an information acquisition module, a work order generation module, a personnel matching module, a solution push module and an evaluation module; The information acquisition module is used to receive and summarize power communication fault information from multiple regions; The work order generation module is used to classify the fault information by using a support vector machine and generate an operation and maintenance work order; The personnel matching module is used to match maintenance personnel with different maintenance skill levels and in a maintenance state based on personnel constraint conditions and issue a scheduling notice; The solution push module is used to match the solution corresponding to the fault information by using the historical database and push it to the scheduled maintenance personnel; The evaluation module is used to evaluate the scheduling process based on the fault handling time and fault handling completion degree of the maintenance personnel.

Images