Electric power operation site safety intelligent management and control system
Through the intelligent power operation site safety control system, the environment and personnel behavior of the power operation site are monitored and analyzed in real time, potential risks are identified and early warning are solved, and the problem of inability to identify and predict safety hazards in the existing technology is solved, and the operation is improved Security and management efficiency.
Patent Information
- Application Number
- CN202411466814.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-21
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing power operation site safety control system cannot identify and predict safety hazards in advance, resulting in slow response speed and incomplete monitoring, and cannot effectively ensure the safety of operators.
A smart management and control system for power operations on site safety is designed, using intelligent terminal management units, edge service platforms and safety production risk management and control platforms, and monitoring the environment and personnel behavior in real time through sensors, high-definition cameras and data analysis modules to identify potential risks, and Early warning and emergency plan matching is carried out based on risk assessment and early warning models.
It realizes timely discovery and handling of potential safety hazards, improves operational safety, reduces manual intervention and data processing time, improves management and early warning efficiency, ensures the rational allocation and efficient utilization of resources, and provides safe production management Strong data support and decision-making basis.
Smart Images

Figure CN118966809A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power operation safety management and control, and in particular to an intelligent power operation site safety management and control system. Background Art
[0002] There are many safety hazards at the current power operation site, such as improper operation of personnel, equipment failure, environmental changes, etc., all of which may lead to serious accidents. Regarding the safety management of power operation sites, the patent application with publication number: CN118627911A discloses an intelligent safety management method for power site operators, including the following steps: S1, setting up a data center and an information synchronization terminal, and equipping each operator with an information synchronization terminal. S2, constructing a three-dimensional model of the power operation site in the data center, and marking each power facility in the three-dimensional model of the power operation site. S3, determining the electrical influence relationship between the power facilities, and marking the working status of each power facility. S4, when a leakage fault occurs, the data center determines the leakage risk area according to the three-dimensional model of the power operation site and the electrical influence relationship, and sends the leakage risk area to the information synchronization terminal. It can realize the intelligent management and control of the safety management of operators at the power site, effectively reduce the blind spots of supervision, eliminate safety hazards, improve the safety of operators, and contribute to the continuous safe operation of the power site.
[0003] However, although the above patent can mark the working status of power facilities and identify and eliminate safety hazards, it is often unable to identify and predict safety hazards in advance, the safety of operators cannot be effectively guaranteed, and there are problems such as slow response speed and incomplete monitoring. Summary of the invention
[0004] The purpose of the present invention is to provide an intelligent safety management and control system for power operation sites, which can effectively identify potential risks at the front end, improve management and early warning efficiency, and solve the problems raised in the above-mentioned background technology.
[0005] To achieve the above object, the present invention provides the following technical solutions: A safety intelligent management and control system for power operation sites, comprising: The intelligent terminal management unit is used to deploy various intelligent terminal devices at the work site, including sensor components and high-definition cameras. The high-definition cameras monitor the work site in real time, perform image recognition and analysis, and monitor the environmental parameters, personnel behavior and equipment status of the work site in real time; The edge service platform is used to manage various intelligent terminal devices, ensure the normal operation of the devices, receive monitoring data uploaded by various intelligent terminal devices, perform preliminary analysis and preprocessing, and trigger front-end warnings based on the preliminary analysis and processing results, including: The data acquisition module is used to receive the monitoring data uploaded from each intelligent terminal device, including environmental parameters, equipment status, personnel behavior, etc., to ensure the integrity and real-time nature of the data and provide a basis for subsequent data analysis and processing; The data analysis module is used to perform preliminary analysis and preprocessing on the received monitoring data, identify anomalies and potential risks in the monitoring data, and send abnormal warning data to the intelligent terminal management unit according to the identification results to provide a decision-making basis for triggering the warning; A data transmission module is used to transmit the data after preliminary analysis and processing and the received monitoring data to the production safety risk management and control platform, and at the same time, receive control instructions and scheduling information from the production safety risk management and control platform; The safety production risk management and control platform is used to receive actively uploaded operation information and send it to each mobile terminal, receive and analyze the monitoring data transmitted by the edge service platform and the processed data, and conduct risk assessment and early warning. The early warning methods include sound and light alarm, SMS notification, platform push, etc. The safety production risk management and control platform is also used to establish a safety production database, store historical data and risk assessment reports, match corresponding emergency plans based on risk assessment results, and perform remote monitoring and scheduling based on received monitoring data, so that managers can grasp the situation of the operation site in real time and adjust the operation plan and safety measures in a timely manner; The production safety risk management and control platform includes: The risk assessment and early warning module is used to receive the monitoring data and analyzed data transmitted by the edge service platform, perform data analysis based on the pre-generated risk assessment and early warning model, and perform risk assessment and early warning according to the analysis results; The judgment module is used to judge the analysis accuracy level of the risk assessment and early warning model according to the previous analysis results before the risk assessment and early warning module performs data analysis based on the pre-generated risk assessment and early warning model, and generate an update instruction for the risk assessment and early warning model when it is determined that the analysis accuracy level is less than the preset level.
[0006] Furthermore, the intelligent terminal management unit further includes: Positioning module, used to track the location information of operators and equipment in the power line field in real time; The early warning module is used to remind operators to pay attention to safety in real time, broadcast on-site environmental parameters based on the data collected by each intelligent terminal device, and broadcast corresponding early warning information based on the early warning data received from the edge service platform and the production safety risk management and control platform.
[0007] Furthermore, the data analysis module includes: An adjustment module is used to pre-train a language model and use the labeled data in the power industry model library to adjust the pre-trained language model to obtain a power field language model; Matching modules for: Obtain the metadata of the monitoring data, input it into the power domain language model, extract the semantic features of the metadata, and obtain the semantic representation vector of the metadata; wherein the metadata includes the table name and the field name; The semantic representation vector of the metadata is matched with a preset semantic representation vector, and the target metadata is determined according to the matching result.
[0008] Furthermore, the data analysis module further includes: Projection module for: Determine the attribute information of the target metadata; the attribute information includes the predicted category, the starting position identifier and the ending position identifier of the corresponding data; Project the attribute information of the target metadata onto a two-dimensional scatter plot, input the name of the attribute information according to the predicted category, so as to expand the data feature space; the expanded feature is denoted as R, and the dimension of R is n*m, where n is the number of data items in the target metadata determined according to the starting position identifier and the ending position identifier, and m is the number of dimensions added; place the target metadata on the axis corresponding to the attribute information, and the placement position corresponds to the value of the target metadata on the attribute; The comparison module is used to compare the projected two-dimensional scatter plot with the preset two-dimensional scatter plot, identify anomalies and potential risks in the monitoring data according to the comparison results, and send abnormal warning data to the intelligent terminal management unit according to the identification results.
[0009] Furthermore, the data analysis module is specifically: The image analysis submodule is used to extract image data from the monitoring data, collect the power operation site environment in the monitoring area, and identify dangerous areas in the power operation site environment, such as high-altitude operation areas, live areas, flammable and explosive areas, etc.; The safety judgment submodule is used to judge whether there are any abnormalities in the identified dangerous areas, mark the abnormalities, collect and analyze the image data with abnormal marks, and send the first warning data to the warning module when there are abnormalities in more than two power operation sites; A human body detection submodule, used to detect whether there is a human body intrusion in the dangerous area with abnormal marks, and transmit the detection information to the judgment module; The safety judgment submodule combines the information transmitted by the human body detection module and sends second warning data to the warning module when it is determined that there is a human body intrusion in the dangerous area with the abnormal mark.
[0010] Furthermore, the safety production risk management and control platform also includes: The operation information management module is used to create, edit and maintain the operation plan, including operation time, location, personnel, equipment and other information, and send the operation plan and instructions to the personnel's mobile terminal; A data receiving module is used to receive the monitoring data actively uploaded by each smart terminal device and the sub-data carried by it, wherein the sub-data includes the operation data and IP data of each smart terminal; The risk assessment module is used to conduct real-time assessment of the safety risks of the work site based on the acquired monitoring data, and send assessment and warning data to the warning module based on the assessment results; The emergency plan management module is used to establish a database containing various emergency plans and match the corresponding emergency plans according to the risk assessment results; The human-computer interaction module is used to display the monitoring data and risk assessment results of the work site in real time, and adjust the work plan and safety measures according to the risk assessment results.
[0011] Furthermore, the risk assessment module specifically includes: The operation behavior management submodule is used to formulate operation behavior specifications, extract image data from monitoring data, identify human body data in image data, analyze human behavior based on the identification results, and identify whether the behavior of operators is abnormal and violates regulations in combination with the operation behavior specifications; The equipment management submodule is used to extract environmental data and operating data of each smart terminal from the monitoring data, including equipment performance, fault conditions, etc., and formulate equipment maintenance plans based on the IP data corresponding to the abnormal operating data of each smart device.
[0012] Furthermore, the operation behavior management submodule is specifically: Formulate operating behavior specifications based on the actual conditions and safety requirements of the power operation site, including the operating procedures, behavioral codes and safety precautions of the operators; Performing human body detection in the image data of the work site acquired in real time according to human body features, identifying the position and outline of the human body, and extracting key features of the human body based on the recognition results to determine the behavior data of the human body; Extracting working environment features from image data, classifying the working environment according to the extracted working environment features, and determining the specific area and location of the workers; Learning the behavior patterns of operators in different working environments, and predicting the operator's behavior operations based on the current working environment and the human body's behavior data; Compare the predicted behavior operation with the established operation behavior specification to determine whether the behavior operation of the operator has potential safety risks; If the operator's behavior violates the operating behavior norms, identify and record the operator's abnormal behavior and violations, and classify and grade the abnormalities and violations according to the severity of the violations and the possible consequences, and issue corresponding warning data.
[0013] Furthermore, the emergency plan management module specifically includes: The emergency plan formulation submodule is used to collect and organize various emergency plans, establish an emergency plan database, formulate corresponding emergency plans according to the operation type, risk level and possible accident type, and review and update the emergency plan according to the feedback results; The risk matching submodule is used to analyze the risk assessment results provided by the risk assessment module, determine potential safety hazards and possible accident types, match the corresponding emergency plan from the emergency plan library according to the risk assessment results, and activate the corresponding emergency plan when the risk assessment results reach the preset threshold.
[0014] Furthermore, the judgment module includes: The first determination module is used to: Obtain the analysis results of risk assessment and early warning models in multiple analysis cycles; Input the analysis results of multiple analysis cycles into the judgment model, and output the analysis scores of the risk assessment and early warning model in each analysis cycle; Determine the calculation function of the judgment model: ; in, To judge the analysis score of the model output for the sample risk assessment and early warning model in the i-th analysis cycle; The total number of analysis cycles for calculating analysis scores for the sample risk assessment and early warning models selected for training the judgment model; To determine the connection weight from the input layer to the output layer in the model, the value range is (0,1); To judge the fitting coefficient of the model, the value range is (0.2, 0.5); is the analysis result of the sample risk assessment and early warning model in the i-th analysis cycle; Based on Determined cluster centers; is the maximum value among the selected cluster centers; To determine the sensitivity threshold of the model, the value is (0,1);
[0015] The analysis results of the risk assessment and early warning model in m analysis cycles are input into the judgment model, and the analysis scores of the risk assessment and early warning model in each analysis cycle are output to obtain the current analysis cycle. and the first m analysis cycles , , …, ; Calculate the overall analysis accuracy of the risk assessment and early warning model for the tth analysis period : ; in, is the correction coefficient, the value range is (-0.05, 0.05); For risk assessment and early warning models The analysis score of the analysis cycle, that is, the analysis score of the current analysis cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; The second determination module is used to evaluate the overall analysis accuracy of the tth analysis cycle according to the risk assessment and early warning model. The analysis accuracy level is determined, and when it is determined that the analysis accuracy level is less than a preset level, an update instruction for the risk assessment and early warning model is generated.
[0016] Compared with the prior art, the present invention has the following beneficial effects: Through real-time monitoring and intelligent early warning of the intelligent terminal management unit, potential safety hazards can be discovered and dealt with in a timely manner, accidents can be prevented, and operational safety can be effectively improved. The edge service platform realizes rapid data processing and transmission, effectively identifies potential risks at the front end, reduces manual intervention and data processing time, and improves management and early warning efficiency. The safe production risk management and control platform matches emergency plans based on risk assessment results, realizes reasonable allocation and efficient use of resources, and provides strong data support and decision-making basis for safe production management. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 This is a module diagram of the power operation site safety intelligent management and control system of the present invention. DETAILED DESCRIPTION
[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] In order to solve the technical problems of being unable to identify and predict safety hazards in advance, the safety of operators cannot be effectively guaranteed, and there are slow response speeds and incomplete monitoring, please refer to Figure 1 , this embodiment provides the following technical solutions: A safety intelligent management and control system for power operation sites, comprising: The intelligent terminal management unit is used to deploy various intelligent terminal devices at the work site, including sensor components and high-definition cameras. The sensor components include sensors such as temperature, humidity, current, voltage, and gas concentration to monitor environmental parameters in real time. The high-definition camera monitors the work site in real time and performs image recognition and analysis to monitor the environmental parameters, personnel behavior, and equipment status of the work site in real time. It also includes: Positioning module, used to track the location information of operators and equipment in the power line field in real time; The early warning module is used to remind operators to pay attention to safety in real time, broadcast the on-site environmental parameters based on the data collected by each intelligent terminal device, and broadcast the corresponding early warning information based on the early warning data received from the edge service platform and the safety production risk management and control platform; The edge service platform is used to manage various intelligent terminal devices, ensure the normal operation of the devices, receive the monitoring data uploaded by various intelligent terminal devices, and perform preliminary analysis and preprocessing to reduce data transmission delays, realize rapid data processing and transmission, and trigger front-end warnings based on preliminary analysis and processing results; The safety production risk management and control platform is used to receive actively uploaded operation information and send it to each mobile terminal. It receives and analyzes the monitoring data and processed data transmitted by the edge service platform, conducts risk assessment and early warning, and evaluates the safety risks of the operation site through big data analysis and artificial intelligence algorithms. The early warning methods include sound and light alarms, SMS notifications, platform push, etc. The production safety risk management and control platform is also used to establish a production safety database, store historical data and risk assessment reports, provide data support for production safety management, match corresponding emergency plans based on risk assessment results, and perform remote monitoring and scheduling based on the received monitoring data, so that managers can grasp the situation on the work site in real time and adjust work plans and safety measures in a timely manner.
[0020] In this embodiment, through real-time monitoring and intelligent early warning of the intelligent terminal management unit, potential safety hazards are discovered and dealt with in a timely manner, accidents are prevented, and operational safety is effectively improved. The edge service platform realizes rapid data processing and transmission, effectively identifies potential risks at the front end, reduces manual intervention and data processing time, and improves management and early warning efficiency. The production safety risk management and control platform matches emergency plans based on risk assessment results, realizes reasonable allocation and efficient utilization of resources, and provides strong data support and decision-making basis for production safety management.
[0021] In this embodiment, the edge service platform includes: The data acquisition module is used to receive the monitoring data uploaded from each intelligent terminal device, including environmental parameters (such as temperature, humidity, gas concentration), equipment status (such as current, voltage, running time), personnel behavior (obtained through image recognition technology), etc., to ensure the integrity and real-time nature of the data and provide a basis for subsequent data analysis and processing; The data analysis module is used to perform preliminary analysis and preprocessing on the received monitoring data, identify anomalies and potential risks in the monitoring data, and send abnormal warning data to the intelligent terminal management unit based on the identification results to provide a decision-making basis for triggering the warning. At the same time, the preprocessed data is easier to store and transmit, which improves the efficiency of the overall system. Specifically: The image analysis submodule is used to extract image data from the monitoring data, collect the power operation site environment in the monitoring area, and identify dangerous areas in the power operation site environment, such as high-altitude operation areas, live areas, flammable and explosive areas, etc.; The safety judgment submodule is used to judge whether there are any abnormalities in the identified dangerous areas, mark the abnormalities, collect and analyze the image data with abnormal marks, and send the first warning data to the warning module when there are abnormalities in more than two power operation sites; A human body detection submodule, used to detect whether there is a human body intrusion in the dangerous area with abnormal marks, and transmit the detection information to the judgment module; The safety judgment submodule combines the information transmitted by the human body detection module and sends second warning data to the warning module when judging that there is a human body intrusion in the dangerous area with the abnormal mark; The data transmission module is used to transmit the data after preliminary analysis and processing and the received monitoring data to the production safety risk management and control platform. At the same time, it receives control instructions and scheduling information from the production safety risk management and control platform to realize remote management and control of smart terminal devices.
[0022] In this embodiment, by integrating multiple intelligent functional modules, comprehensive monitoring and intelligent analysis of the power operation site is achieved, the intelligence level and automation degree of the system are improved, real-time data analysis and preliminary decision-making are carried out, potential safety risks are discovered and warned in a timely manner, the time for data transmission from the site to the production safety risk management and control platform is reduced, the processing speed is improved, accidents are quickly identified and emergency plans are initiated, the spread and impact of accidents are reduced, and strong support is provided for rapid response and disposal.
[0023] In this embodiment, the production safety risk management and control platform includes: The operation information management module is used to create, edit and maintain operation plans, including operation time, location, personnel, equipment and other information, and send operation plans and instructions to personnel's mobile terminals to ensure that operators receive accurate information, improve the transparency and execution efficiency of operation plans, and reduce safety risks caused by poor information or misunderstandings; The data receiving module is used to receive the monitoring data actively uploaded by each intelligent terminal device and the sub-data it carries. The sub-data includes the operation data and IP data of each intelligent terminal, which provides important data support for subsequent risk assessment and emergency response, realizes real-time and comprehensive data collection, and provides a solid data foundation for risk assessment and emergency response; The risk assessment module is used to conduct real-time assessment of the safety risks of the work site based on the acquired monitoring data, and send assessment and warning data to the warning module based on the assessment results, including: The operation behavior management submodule is used to formulate operation behavior specifications, extract image data from monitoring data, identify human body data in image data, analyze human behavior based on the identification results, and identify whether the behavior of operators is abnormal and violates regulations in combination with the operation behavior specifications; The device management submodule is used to extract environmental data from the monitoring data and the operating data of each smart terminal, including device performance, fault conditions, etc. Based on the IP data corresponding to the abnormal operating data of each smart device, an equipment maintenance plan is formulated to ensure that the equipment is in good operating condition; In this embodiment, through real-time risk assessment, potential safety hazards can be discovered and warned in a timely manner, providing a scientific basis for taking effective safety measures; The emergency plan management module is used to establish a database containing various emergency plans and match the corresponding emergency plans according to the risk assessment results, including: The emergency plan formulation submodule is used to collect and organize various emergency plans, establish an emergency plan database, formulate corresponding emergency plans according to the operation type, risk level and possible accident type, review and update the emergency plan according to the feedback results, and ensure the timeliness and effectiveness of the plan; The risk matching submodule is used to analyze the risk assessment results provided by the risk assessment module, determine potential safety hazards and possible accident types, match the corresponding emergency plan from the emergency plan library according to the risk assessment results, and activate the corresponding emergency plan when the risk assessment results reach the preset threshold; In this embodiment, through the precise matching and rapid activation of the emergency plan, effective response measures can be taken quickly when an accident occurs to reduce the loss and impact of the accident; The human-computer interaction module is used to display the monitoring data and risk assessment results of the work site in real time, and adjust the work plan and safety measures according to the risk assessment results. It provides an intuitive and convenient information exchange platform for managers and operators, which helps managers to grasp the situation on site in a timely manner and make scientific decisions; it also helps operators understand the safety status of their own behavior and equipment status and enhance safety awareness.
[0024] In this embodiment, the production safety risk management and control platform realizes comprehensive monitoring, real-time evaluation and rapid response to the work site, which not only improves the execution efficiency and transparency of the work plan, but also enhances the accuracy of risk assessment and the pertinence of emergency plans; at the same time, through the information exchange function of the human-computer interaction module, the safety awareness and response capabilities of managers and operators are further enhanced.
[0025] In this embodiment, the operation behavior management submodule is specifically: Formulate operating behavior specifications based on the actual conditions and safety requirements of the power operation site, including the operating procedures, behavioral codes and safety precautions of the operators, providing clear guidance for the management of operating behaviors; Performing human body detection in the image data of the work site acquired in real time according to human body features, identifying the position and outline of the human body, and extracting key features of the human body based on the recognition results, including key features of posture and action, to determine the behavior data of the human body; Extracting working environment features from image data, classifying the working environment according to the extracted working environment features, and determining the specific area and location of the workers; Learn the behavior patterns of operators in different working environments, predict the operator's behavior operations based on the current working environment and the human body's behavior data, and identify potential risk points in advance, providing the possibility for timely preventive measures; Compare the predicted behavior operation with the established operation behavior specification to determine whether the behavior operation of the operator has potential safety risks; If the operator's behavior violates the operating behavior norms, identify and record the operator's abnormal behavior and violations, and classify and grade the abnormalities and violations according to the severity of the violations and the possible consequences, and issue corresponding warning data.
[0026] In this embodiment, by real-time monitoring and analysis of the behavioral data of operators, potential safety risks are discovered and corrected in a timely manner, the possibility of accidents is effectively reduced, the behavioral data of operators is automatically recorded and analyzed, the supervisory burden of managers is reduced, the operating processes and specifications are further optimized, and the supervision efficiency is improved. By formulating and implementing operating behavior specifications, operators are guided to form correct safety awareness and behavioral habits. According to the behavioral data and analysis results of the operators, human and material resources can be reasonably allocated to ensure that key operating links are fully supported and guaranteed, which helps to improve resource utilization efficiency and reduce operating costs.
[0027] Furthermore, the production safety risk management and control platform includes: The risk assessment and early warning module is used to receive the monitoring data and analyzed data transmitted by the edge service platform, perform data analysis based on the pre-generated risk assessment and early warning model, and perform risk assessment and early warning according to the analysis results; A judgment module is used to judge the analysis accuracy level of the risk assessment and early warning model according to the previous analysis results before the risk assessment and early warning module performs data analysis based on the pre-generated risk assessment and early warning model, and generate an update instruction for the risk assessment and early warning model when it is determined that the analysis accuracy level is less than the preset level; The judging module comprises: The first determination module is used to: Obtain the analysis results of risk assessment and early warning models in multiple analysis cycles; Input the analysis results of multiple analysis cycles into the judgment model, and output the analysis scores of the risk assessment and early warning model in each analysis cycle; Determine the calculation function of the judgment model: ;
[0028] in, To judge the analysis score of the model output for the sample risk assessment and early warning model in the i-th analysis cycle; The total number of analysis cycles for calculating analysis scores for the sample risk assessment and early warning models selected for training the judgment model; To determine the connection weight from the input layer to the output layer in the model, the value range is (0,1); To judge the fitting coefficient of the model, the value range is (0.2, 0.5); is the analysis result of the sample risk assessment and early warning model in the i-th analysis cycle; Based on Determined cluster centers; is the maximum value among the selected cluster centers; To determine the sensitivity threshold of the model, the value is (0,1); The analysis results of the risk assessment and early warning model in m analysis cycles are input into the judgment model, and the analysis scores of the risk assessment and early warning model in each analysis cycle are output to obtain the current analysis cycle. and the first m analysis cycles , , …, ; Calculate the overall analysis accuracy of the risk assessment and early warning model for the tth analysis period : ; in, is the correction coefficient, the value range is (-0.05, 0.05); For risk assessment and early warning models The analysis score of the analysis cycle, that is, the analysis score of the current analysis cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; The second determination module is used to evaluate the overall analysis accuracy of the tth analysis cycle according to the risk assessment and early warning model. The analysis accuracy level is determined, and when it is determined that the analysis accuracy level is less than a preset level, an update instruction for the risk assessment and early warning model is generated.
[0029] The working principle and beneficial effects of the above technical solution: The risk assessment and early warning module includes a data interface, which is connected to the edge service platform to ensure that the monitoring data and analysis results can be received stably and in real time. The risk assessment and early warning model includes machine learning algorithms, such as decision trees, random forests, neural networks, etc. The risk assessment and early warning model is used to receive the monitoring data and the processed data for data analysis, generate analysis results, and perform risk assessment and early warning according to the analysis results. The judgment module is used to judge the analysis accuracy level of the risk assessment and early warning model according to the previous analysis results, and also characterize the accuracy of the analysis results, and then also characterize the accuracy of risk assessment and early warning. The analysis cycle can be divided according to the working cycle of the power operation task, or according to the execution steps of the power operation task. Based on the first determination module included in the judgment module, the analysis results of the risk assessment and early warning model in multiple analysis cycles are obtained, the analysis results of multiple analysis cycles are input into the judgment model, and the analysis scores of the risk assessment and early warning model in each analysis cycle are output; when determining the calculation function of the judgment model, firstly, the analysis scores of the sample risk assessment and early warning model in each analysis cycle are trained to obtain the calculation function. Among them, the connection weight from the input layer to the output layer in the judgment model defines the connection strength between the input layer and the output layer. The value range of the connection weight is (0,1). It means that the weight is normalized or standardized to this interval, which is achieved by some form of activation function (such as sigmoid) or the weight is directly scaled. When training the model, these weights are adjusted to minimize the loss function to optimize the predictive performance of the model. The fit coefficient of the judgment model is used to measure the fit of the model to the data. The range of 0.2 to 0.5 represents an acceptable range of the model's fit, where lower values indicate underfitting and higher values indicate overfitting. During the model selection or tuning process, you can try to keep the fit coefficient within this range by adjusting model parameters (such as regularization strength) to achieve the best generalization ability. The "center of mass" of the cluster center data point represents the center position of a cluster. The maximum value between cluster centers is used to evaluate the dispersion or coverage of clusters in order to more accurately reflect the risk pattern in the data. The sensitivity threshold of the judgment model is used to determine whether the model output is sensitive enough to trigger an early warning. The analysis results of the risk assessment and early warning model in m analysis cycles are input into the judgment model, and the analysis scores of the risk assessment and early warning model in each analysis cycle are output, and then the overall analysis accuracy evaluation results of the risk assessment and early warning model in the tth analysis cycle are calculated, and compared with the preset level division parameters to determine the analysis accuracy level. When it is determined that the analysis accuracy level is less than the preset level, an update instruction for the risk assessment and early warning model is generated. Accurate monitoring of the risk assessment and early warning performance of the risk assessment and early warning model is achieved, and the accuracy of risk assessment and early warning is improved.
[0030] Furthermore, the data analysis module includes: An adjustment module is used to pre-train a language model and use the labeled data in the power industry model library to adjust the pre-trained language model to obtain a power field language model; Matching modules for: Obtain the metadata of the monitoring data, input it into the power domain language model, extract the semantic features of the metadata, and obtain the semantic representation vector of the metadata; wherein the metadata includes the table name and the field name; Matching the semantic representation vector of the metadata with a preset semantic representation vector, and determining the target metadata according to the matching result; Projection module for: Determine the attribute information of the target metadata; the attribute information includes the predicted category, the starting position identifier and the ending position identifier of the corresponding data; Project the attribute information of the target metadata onto a two-dimensional scatter plot, input the name of the attribute information according to the predicted category, so as to expand the data feature space; the expanded feature is denoted as R, and the dimension of R is n*m, where n is the number of data items in the target metadata determined according to the starting position identifier and the ending position identifier, and m is the number of dimensions added; place the target metadata on the axis corresponding to the attribute information, and the placement position corresponds to the value of the target metadata on the attribute; The comparison module is used to compare the projected two-dimensional scatter plot with the preset two-dimensional scatter plot, identify anomalies and potential risks in the monitoring data according to the comparison results, and send abnormal warning data to the intelligent terminal management unit according to the identification results.
[0031] Working principle of the above technical solution: In this embodiment, pre-trained language models such as BERT, GPT series, etc. Load the selected pre-trained model into the training environment, divide the labeled power industry data into training set, validation set and test set, and use the power industry data to fine-tune the pre-trained model. Adjust hyperparameters such as learning rate, batch size, and training rounds to obtain optimal performance. Evaluate the performance of the model on the validation set and test set to ensure that the model performs well on the power field data. The pre-trained language model can be adjusted to a language model that adapts to the needs of the power field, thereby achieving more efficient and accurate natural language processing tasks in this field.
[0032] In this embodiment, the metadata is segmented so that the model can better understand the meaning of each word, ensure that all metadata follow the same naming conventions, such as unified capitalization, removal of unnecessary spaces or special characters, etc., and use word embeddings or character embeddings to convert the metadata into a numerical form that the model can process. The preprocessed metadata is encoded into an input format that the model can accept, and the semantic representation vector of each metadata is obtained through the forward propagation of the model. These vectors are located in the hidden layer or output layer of the model and can capture the semantic features of the metadata. According to business needs or specific goals, a preset semantic representation vector is defined. The preset semantic representation vector represents a specific power field concept, category or task. The similarity between the semantic representation vector of each metadata and the preset semantic representation vector is calculated using a similarity measurement method (such as cosine similarity, Euclidean distance, etc.). The metadata is sorted according to the similarity score, and the metadata with the highest score and greater than the similarity threshold is selected to determine the target metadata as valid data.
[0033] In this embodiment, based on the business rules of the target metadata, one or more category labels are assigned to the target metadata to determine the predicted category. The starting position identifier and the end position identifier are used to determine the specific position range of the target metadata in the original data set. In order to visualize and expand the data feature space, the attribute information is projected onto a two-dimensional scatter plot. However, since we have multiple attributes (including predicted categories and position information), direct projection into a two-dimensional space is not enough to fully display all information. Therefore, different colors or shapes are used to represent different predicted categories. The two-dimensional scatter plot has two axes, X and Y, which are used to display the starting position or the end position. Additional attribute information is displayed through tooltips, labels, or text annotations. The target metadata is regarded as a series of data points, and m is the number of dimensions added. These dimensions include certain statistical features of the original data (such as mean, standard deviation), converted features (such as Fourier transform coefficients), and encodings related to the predicted category (such as one-hot encoding). On the two-dimensional scatter plot, the data items of each target metadata are placed on the axis determined by its attribute information, and the placement position corresponds to the value of the target metadata on the attribute. The expanded feature space R is combined with a two-dimensional scatter plot to more comprehensively display the attribute information of the target metadata and the relationship between them.
[0034] In this embodiment, the preset two-dimensional scatter plot is constructed based on historical normal data, business rules or expert knowledge. The points on the preset scatter plot represent the data distribution and characteristics under normal conditions.
[0035] Based on density estimation, cluster analysis, outlier detection algorithms, etc., the projected two-dimensional scatter plot is compared with the preset two-dimensional scatter plot to quantify the difference between the two scatter plots. For example, the distribution difference, distance measurement (such as Euclidean distance, Manhattan distance) or statistical test results (such as t-test, chi-square test) of the two data sets can be calculated. Based on the comparison results of the scatter plots, the points in the projected scatter plot that are significantly different from the preset scatter plot are identified as abnormal points. These points represent outliers or abnormal behaviors in the monitoring data. According to the attribute information of the abnormal points (such as prediction category, location information, etc.) and business rules, the potential risks that these abnormalities may bring are evaluated. For example, some abnormalities may indicate equipment failure, data leakage or bottlenecks in business processes. The identified abnormal points and their related information (such as abnormality type, occurrence time, impact range, etc.) are sorted into early warning data. The early warning data is sent to the intelligent terminal management unit through an appropriate communication protocol (such as HTTP, MQTT, etc.). These data can trigger alarms, notifications or automatic response mechanisms on the intelligent terminal.
[0036] The beneficial effects of the above technical solution are as follows: determining the valid data in the monitoring data and projecting it onto a two-dimensional scatter plot to achieve the expansion of the data feature space; comparing the projected two-dimensional scatter plot with the preset two-dimensional scatter plot to achieve the accuracy of data analysis; identifying anomalies and potential risks in the monitoring data based on the comparison results; and sending abnormal warning data to the intelligent terminal management unit based on the identification results, so as to facilitate the accurate determination of abnormal warning data.
[0037] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A power operation site safety intelligent management and control system, characterized in that: include: The intelligent terminal management unit is used to deploy various intelligent terminal devices at the work site to monitor the environmental parameters, personnel behavior and equipment status of the work site in real time; The edge service platform is used to manage various intelligent terminal devices, receive monitoring data uploaded by various intelligent terminal devices, and perform preliminary analysis and preprocessing. Based on the preliminary analysis and processing results, it triggers front-end warnings, including: A data acquisition module is used to receive uploaded monitoring data from each smart terminal device; The data analysis module is used to perform preliminary analysis and preprocessing on the received monitoring data, identify anomalies and potential risks in the monitoring data, and send abnormal warning data to the intelligent terminal management unit according to the identification results; A data transmission module is used to transmit the data after preliminary analysis and processing and the received monitoring data to the production safety risk management and control platform, and at the same time, receive control instructions and scheduling information from the production safety risk management and control platform; The safety production risk management and control platform is used to receive actively uploaded operation information, receive and analyze the monitoring data transmitted by the edge service platform and the processed data, and conduct risk assessment and early warning; The safety production risk management and control platform is also used to establish a safety production database, store historical data and risk assessment reports, match corresponding emergency plans based on risk assessment results, and perform remote monitoring and scheduling based on received monitoring data; The production safety risk management and control platform includes: The risk assessment and early warning module is used to receive the monitoring data and analyzed data transmitted by the edge service platform, perform data analysis based on the pre-generated risk assessment and early warning model, and perform risk assessment and early warning according to the analysis results; The judgment module is used to judge the analysis accuracy level of the risk assessment and early warning model according to the previous analysis results before the risk assessment and early warning module performs data analysis based on the pre-generated risk assessment and early warning model, and generate an update instruction for the risk assessment and early warning model when it is determined that the analysis accuracy level is less than the preset level.
2. The intelligent safety management and control system for electric power operation site according to claim 1, characterized in that: The intelligent terminal management unit also includes: Positioning module, used to track the location information of operators and equipment in the power line field in real time; The early warning module is used to broadcast the on-site environmental parameters based on the data collected by each intelligent terminal device, and to broadcast the corresponding early warning information based on the early warning data received from the edge service platform and the production safety risk management and control platform.
3. The intelligent safety management and control system for power operation sites according to claim 2, characterized in that: Data analysis module, specifically: An image analysis submodule is used to extract image data from the monitoring data, collect the power operation site environment in the monitoring area, and identify dangerous areas in the power operation site environment; The safety judgment submodule is used to judge whether there are any abnormalities in the identified dangerous areas, mark the abnormalities, collect and analyze the image data with abnormal marks, and send the first warning data to the warning module when there are abnormalities in more than two power operation sites; A human body detection submodule, used to detect whether there is a human body intrusion in the dangerous area with abnormal marks, and transmit the detection information to the judgment module; The safety judgment submodule combines the information transmitted by the human body detection module and sends second warning data to the warning module when it is determined that there is a human body intrusion in the dangerous area with the abnormal mark.
4. The intelligent safety management and control system for power operation sites according to claim 3, characterized in that: The data analysis module comprises: An adjustment module is used to pre-train a language model and use the labeled data in the power industry model library to adjust the pre-trained language model to obtain a power field language model; Matching modules for: Obtain the metadata of the monitoring data, input it into the power domain language model, extract the semantic features of the metadata, and obtain the semantic representation vector of the metadata; wherein the metadata includes the table name and the field name; The semantic representation vector of the metadata is matched with a preset semantic representation vector, and the target metadata is determined according to the matching result.
5. The intelligent safety management and control system for electric power operation site according to claim 4, characterized in that: The data analysis module further includes: Projection module for: Determine the attribute information of the target metadata; the attribute information includes the predicted category, the starting position identifier and the ending position identifier of the corresponding data; Project the attribute information of the target metadata onto a two-dimensional scatter plot, input the name of the attribute information according to the predicted category, so as to expand the data feature space; the expanded feature is denoted as R, and the dimension of R is n*m, where n is the number of data items in the target metadata determined according to the starting position identifier and the ending position identifier, and m is the number of dimensions added; place the target metadata on the axis corresponding to the attribute information, and the placement position corresponds to the value of the target metadata on the attribute; The comparison module is used to compare the projected two-dimensional scatter plot with the preset two-dimensional scatter plot, identify anomalies and potential risks in the monitoring data according to the comparison results, and send abnormal warning data to the intelligent terminal management unit according to the identification results.
6. The intelligent safety management and control system for power operation site according to claim 5, characterized in that: The production safety risk management and control platform also includes: The operation information management module is used to create, edit and maintain operation plans, and send operation plans and instructions to the mobile terminals where personnel are located; A data receiving module is used to receive monitoring data actively uploaded by each smart terminal device and the sub-data it carries, wherein the sub-data includes the operation data and IP data of each smart terminal; The risk assessment module is used to conduct real-time assessment of the safety risks of the work site based on the acquired monitoring data, and send assessment and warning data to the warning module based on the assessment results; The emergency plan management module is used to establish a database containing various emergency plans and match the corresponding emergency plans according to the risk assessment results; The human-computer interaction module is used to display the monitoring data and risk assessment results of the work site in real time, and adjust the work plan and safety measures according to the risk assessment results.
7. The intelligent safety management and control system for electric power operation site according to claim 6, characterized in that: The risk assessment module specifically includes: The operation behavior management submodule is used to formulate operation behavior specifications, extract image data from monitoring data, identify human body data in image data, analyze human behavior based on the identification results, and identify whether the behavior of operators is abnormal and violates regulations in combination with the operation behavior specifications; The device management submodule is used to extract environmental data and operating data of each smart terminal from the monitoring data, and formulate equipment maintenance plans based on the IP data corresponding to the abnormal operating data of each smart device.
8. The intelligent safety management and control system for electric power operation site according to claim 7, characterized in that: The operation behavior management submodule is as follows: Formulate operating behavior specifications based on the actual conditions and safety requirements of the power operation site, including the operating procedures, behavioral codes and safety precautions of the operators; Performing human body detection in the image data of the work site acquired in real time according to human body features, identifying the position and outline of the human body, and extracting key features of the human body based on the recognition results to determine the behavior data of the human body; Extracting working environment features from image data, classifying the working environment according to the extracted working environment features, and determining the specific area and location of the workers; Learning the behavior patterns of operators in different working environments, and predicting the operator's behavior operations based on the current working environment and the human body's behavior data; Compare the predicted behavior operation with the established operation behavior specification to determine whether the behavior operation of the operator has potential safety risks; If the operator's behavior violates the operating behavior norms, identify and record the operator's abnormal behavior and violations, and classify and grade the abnormalities and violations according to the severity of the violations and the possible consequences, and issue corresponding warning data.
9. The intelligent safety management and control system for electric power operation site according to claim 8, characterized in that: Emergency plan management module, including: The emergency plan formulation submodule is used to collect and organize various emergency plans, establish an emergency plan database, formulate corresponding emergency plans according to the operation type, risk level and possible accident type, and review and update the emergency plan according to the feedback results; The risk matching submodule is used to analyze the risk assessment results provided by the risk assessment module, determine potential safety hazards and possible accident types, match the corresponding emergency plan from the emergency plan library according to the risk assessment results, and activate the corresponding emergency plan when the risk assessment results reach the preset threshold.
10. The intelligent safety management and control system for electric power operation site according to claim 9, characterized in that: The judging module comprises: The first determination module is used to: Obtain the analysis results of risk assessment and early warning models in multiple analysis cycles; Input the analysis results of multiple analysis cycles into the judgment model, and output the analysis scores of the risk assessment and early warning model in each analysis cycle; Determine the calculation function of the judgment model: ; in, To judge the analysis score of the model output for the sample risk assessment and early warning model in the i-th analysis cycle; The total number of analysis cycles for calculating analysis scores for the sample risk assessment and early warning models selected for training the judgment model; To determine the connection weight from the input layer to the output layer in the model, the value range is (0,1); To judge the fitting coefficient of the model, the value range is (0.2, 0.5); is the analysis result of the sample risk assessment and early warning model in the i-th analysis cycle; Based on Determined cluster centers; is the maximum value among the selected cluster centers; To determine the sensitivity threshold of the model, the value is (0,1); The analysis results of the risk assessment and early warning model in m analysis cycles are input into the judgment model, and the analysis scores of the risk assessment and early warning model in each analysis cycle are output to obtain the current analysis cycle. and the first m analysis cycles , , …, ; Calculate the overall analysis accuracy of the risk assessment and early warning model for the tth analysis period : ; in, is the correction coefficient, the value range is (-0.05, 0.05); For risk assessment and early warning models The analysis score of the analysis cycle, that is, the analysis score of the current analysis cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; For risk assessment and early warning models Analytical scoring of the analytical cycle; The second determination module is used to evaluate the overall analysis accuracy of the tth analysis cycle according to the risk assessment and early warning model. The analysis accuracy level is determined, and when it is determined that the analysis accuracy level is less than a preset level, an update instruction for the risk assessment and early warning model is generated.
Citation Information
Patent Citations
SYSTEM FOR MONITORING CONTROLLED AREAS IN WORK ENVIRONMENTS
BR102018067804A2
Industrial internet cloud platform of power consumption data application scene
CN115034619A
Intelligent management system based on electric power safety risk
CN115879774A
Traditional Chinese medicine compliance training data management and analysis platform based on AI artificial intelligence
CN118608342A
Building construction site intelligent monitoring system based on artificial intelligence
CN118799125A
Cited By
Method for identifying hidden danger of field operation through AI image
CN119181062A
Safety management and control method and system for electric power operation and maintenance operation
CN119417180A
Electric power operation environment intelligent positioning and risk management and control method based on three-dimensional model
CN119558659A
Intelligent positioning and risk control method of power operation environment based on three-dimensional model
CN119558659B
Supervision method and device for data compliance risk, supervision equipment and storage medium
CN119622753A
