Live-line work risk assessment method and system based on multi-source data

Abstract

The invention relates to the technical field of hot-line work assessment, and discloses a hot-line work risk assessment method and system based on multi-source data, and the method comprises the steps: collecting equipment data, environment data and geological data, and carrying out the preprocessing; the method comprises the following steps: constructing a knowledge graph structure based on connection of electrified equipment, calculating link cost, analyzing node priority change to carry out dynamic shortest path calculation, optimizing risk propagation of abnormal points, calculating risk change of each grid point, and carrying out anomaly detection. According to the method, the physical and logic relations between the nodes are determined through the knowledge graph, multi-dimensional risk modeling and analysis are facilitated, the method can adapt to field data changes and reflect the newest risk distribution state through real-time monitoring and updating of node priority ranking, negative weight edges are supported through the Bellman-Ford algorithm, and the method has the advantages of being high in reliability and high in reliability. Possible negative influence factors in risk propagation can be effectively processed, the risk diffusion range and intensity of nodes can be accurately described through a Gaussian diffusion model, and the influence of the nodes on surrounding grid points is quantified.

Description

Technical Field

[0001] The present invention relates to the technical field of live working evaluation, and particularly to a method and system for live working risk assessment based on multi-source data. Background Art

[0002] With the rapid development of the power industry, live working technology, as an important means to ensure the stable operation of the power system, has been widely applied. Traditional live working mainly relies on the experience of operators and a single data source for risk assessment and decision-making. This method can meet the requirements in simple scenarios, and gradually simulates common situations to train the experience of live working personnel.

[0003] However, in the face of the increasingly complex power grid structure and changing working environment, its limitations gradually emerge. There is a lack of effective optimization means for anomaly detection and correction, and abnormal data often causes large deviations in the overall assessment results. In addition, existing path planning methods are mostly based on fixed weights or static parameter calculations, and cannot be optimized and adjusted according to the dynamic changes of real-time risks, resulting in insufficient efficiency of risk avoidance. Summary of the Invention

[0004] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0005] In view of the above existing problems, the present invention is proposed. Therefore, the present invention provides a method for live working risk assessment based on multi-source data to solve the problems of lack of effective optimization means for anomaly detection and correction, where abnormal data often causes large deviations in the overall assessment result, and in addition, existing path planning methods are mostly based on fixed weights or static parameter calculations and cannot be optimized and adjusted according to the dynamic changes of real-time risks, resulting in insufficient efficiency of risk avoidance.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, the present invention provides a method for live working risk assessment based on multi-source data, including: collecting equipment data, environmental data, and geological data, and performing preprocessing;

[0008] Based on the equipment data, environmental data, and geological data, constructing a knowledge graph structure, calculating link costs, analyzing node priorities, and dynamically calculating the shortest path;

[0009] Based on the knowledge graph structure, determine the three-dimensional grid coordinates of the nodes, calculate the risk value information of the grid, optimize the risk propagation of the abnormal points, correct the observed values, and calculate the risk change of each grid point;

[0010] Based on the risk change, calculate the main frequency period value and perform anomaly detection;

[0011] Visualize the detected abnormal links and generate a data report including operation data.

[0012] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: the collected device data, environmental data, and geological data include:

[0013] Scan the live working area to generate point cloud data;

[0014] Convert the point cloud data into a three-dimensional model to generate a simulation environment;

[0015] Collect the operation data, environmental data, and geological data of the live equipment in the simulation environment in real time. The operation data includes device current and voltage, the environmental data includes temperature, humidity, and wind speed data, and the geological data includes soil density data and groundwater level data;

[0016] Update the three-dimensional model in real time and perform time series alignment on data with different sampling frequencies.

[0017] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: constructing the knowledge graph structure includes:

[0018] Use the live equipment and data items as nodes to generate a node set. The attribute of each node is the risk index value of the node. The risk index values of the node include humidity data anomaly, temperature data anomaly, and voltage data anomaly. The connection cost of the node includes geographical distance, resistance value, and the processing time value of the node;

[0019] The normalized deviation value of the real-time value of the operation data of the node from the safety standard threshold is used as the operation anomaly value. The normalized deviation value of the real-time value of the environmental data of the node from the safety standard threshold is used as the environmental anomaly value. The normalized deviation value of the real-time value of the address data of the node from the safety standard threshold is used as the geological anomaly value. The operation anomaly value, environmental anomaly value, and geological anomaly value are combined as the comprehensive risk value of the node;

[0020] Based on the ratio of the comprehensive risk value of the node to the processing time value of the node, perform priority sorting of node tasks;

[0021] Construct node pairs into edges according to the electrical connection relationships of the nodes, and form an edge set;

[0022] Calculate the link cost according to the cost and pipe connection of the physical connection of the nodes;

[0023] According to the node set, edge set and link cost, construct a directed graph G for all nodes and edges according to the link cost, and use the initial link cost as the weight of the nodes.

[0024] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: the dynamic calculation of the shortest path includes:

[0025] Monitor the priority ranking of the nodes in real time, and update the node status according to the change value of the node priority ranking value, expressed as:

[0026] Wij = W'ij + ΔPi·dij

[0027] Wherein, Wij represents the new weight of node i and node j, W'ij represents the state weight of node i and node j at the previous time, ΔPi represents the change value of the priority ranking of node i, and d ij represents the distance between node i and node j;

[0028] Perform dynamic shortest path calculation through the Bellman-Ford algorithm, and find the path with the lowest propagation risk in the directed graph.

[0029] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: determining the three-dimensional grid coordinates of the nodes, calculating the risk value information of the grid, optimizing the risk propagation of the abnormal points, and correcting the observed values include:

[0030] Based on the simulation environment and the calculated directed graph, determine the three-dimensional grid coordinates of each node, and update and calculate the comprehensive risk value of the grid according to the optimal path of the current risk propagation;

[0031] Define the influence function of the node on the grid point through the Gaussian diffusion model of the distance between the node and the grid point;

[0032] Calculate the mean deviation between the risk value of the grid point and the neighboring points, and use the sum of the mean and standard deviation of the historical deviation values as the deviation threshold. If the mean deviation is greater than or equal to the deviation threshold, it is determined as an abnormal grid point;

[0033] Construct a wave equation through wavelength reversal simulation, expressed as:

[0034]

[0035] Among them, R(x, y, z, t) represents the distribution value of the risk value over time and space, v represents the risk propagation speed, represents the symbol of partial derivative, ▽ 2 represents the Laplace operator;

[0036] Optimize the propagation of abnormal points through time reversal and perform correction according to the observed values, which is expressed as:

[0037]

[0038] Among them, R J (x, y, z, t) represents the risk value of the corrected abnormal point, Δt represents the time change value, P(x, y, z, t) represents the comprehensive risk value of the grid point (x, y, z), and dt represents the integration variable;

[0039] Based on the risk value of the corrected abnormal point, calculate the risk change of each grid point, which is expressed as:

[0040] P(x, y, z, t + Δt) = P(x, y, z, t) + Δt·[-▽·(F) + S(x, y, z, t)]

[0041] Among them, P(x, y, z, t + Δt) represents the comprehensive risk value at the next time, ▽·F represents the divergence calculation of the risk propagation flux F, and S(x, y, z, t) represents the risk source term of the grid point.

[0042] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: the anomaly detection includes:

[0043] Calculate the change rate of the comprehensive risk value of the grid point between two consecutive time steps, and calculate the average risk value change rate of all grid points in the scene;

[0044] Perform a fast Fourier transform on the time series data of the risk value change to obtain a frequency domain signal, find the frequency point with the largest amplitude in the spectrum, and take the reciprocal of the spectrum amplitude of the largest frequency point as the main frequency period value;

[0045] Based on the main frequency period value, calculate the periodic anomaly ratio of the main frequency period and perform anomaly detection;

[0046] Take the sum of the mean and standard deviation of the historical periodic anomaly ratio as the anomaly threshold. If the periodic anomaly ratio is greater than or equal to the anomaly threshold, it is judged as a periodic anomaly.

[0047] As a preferred solution of the live working risk assessment method based on multi-source data according to the present invention, wherein: visually display the detected abnormal link, and generate a data report including operation data, including:

[0048] According to the optimized comprehensive risk value, upload a three-dimensional heat map for visual display, and highlight the key nodes and their risk diffusion links;

[0049] Based on the periodic anomaly, give an audible and visual alarm to warn the testers of live working;

[0050] Statistically analyze and store the operation data of the testers of live working, and transmit the data to a remote terminal and the cloud through wireless transmission for storage and backup;

[0051] Align the operation data of the testers with the operation data, environmental data, and geological data collected by the sensors in terms of time stamps to generate a data report.

[0052] In a second aspect, the present invention provides a live working risk assessment system based on multi-source data, including: a collection module for collecting equipment data, environmental data, and geological data and performing preprocessing;

[0053] A construction module for constructing a knowledge graph structure based on the equipment data, environmental data, and geological data, calculating link costs, analyzing node priorities, and dynamically calculating the shortest path;

[0054] A calculation module for determining the three-dimensional grid coordinates of nodes based on the knowledge graph structure, calculating the risk value information of the grid, optimizing the risk propagation of abnormal points, correcting observation values, and calculating the risk change of each grid point;

[0055] A detection module for calculating the main frequency period value based on the risk change and performing anomaly detection;

[0056] An output module for visually displaying the detected abnormal links and generating a data report including operation data.

[0057] In a third aspect, the present invention provides an electronic device, including:

[0058] A memory and a processor;

[0059] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the live working risk assessment method based on multi-source data are implemented.

[0060] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the live working risk assessment method based on multi-source data are implemented.

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows: By means of the knowledge graph, the physical and logical relationships between nodes are clarified, facilitating multi-dimensional risk modeling and analysis. Through real-time monitoring and updating of node priority sorting, the system can adapt to on-site data changes and reflect the latest risk distribution status. By supporting negative-weight edges through the Bellman-Ford algorithm, it can effectively handle the possible negative influencing factors in risk propagation. Through the Gaussian diffusion model, the risk diffusion range and intensity of nodes can be accurately described, quantifying the influence of nodes on surrounding grid points. By calculating the influence value of nodes on grid points, the risk level of local high-risk areas can be evaluated, facilitating priority processing. By combining the wave equation with time-reversal technology, the risk value of abnormal points is optimized according to historical propagation laws, improving the reliability of anomaly correction. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description 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.

[0063] Figure 1 It is a schematic diagram of the overall process of the live working risk assessment method based on multi-source data according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the accompanying drawings of the specification. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention.

[0065] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0066] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments.

[0067] The present invention will be described in detail with reference to the schematic diagrams. When describing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally not in accordance with the general scale, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0068] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0069] Unless otherwise clearly defined and limited in the present invention, the terms "installed, connected, and coupled" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can also be a mechanical connection, an electrical connection, or a direct connection, and can also be indirectly connected through an intermediate medium, or can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0070] Referring to Figure 1 , an embodiment of the present invention provides a method for risk assessment of live working based on multi-source data, including:

[0071] S101, collecting device data, environmental data, and geological data, and performing preprocessing;

[0072] S102, based on the device data, environmental data, and geological data, constructing a knowledge graph structure, calculating the link cost, analyzing the node priority, and dynamically calculating the shortest path;

[0073] S103, based on the knowledge graph structure, determining the three-dimensional grid coordinates of the nodes, calculating the risk value information of the grids, optimizing the risk propagation of the abnormal points, correcting the observed values, and calculating the risk change of each grid point;

[0074] S104, based on the risk change, calculating the main frequency period value and performing anomaly detection;

[0075] S105, visually displaying the detected abnormal links and generating a data report including operation data.

[0076] In a preferred embodiment, collecting the device data, environmental data, and geological data includes:

[0077] Scanning the live working area to generate point cloud data;

[0078] Convert the point cloud data into a 3D model to generate a simulation environment;

[0079] Collect the operation data, environmental data, and geological data of the energized equipment in the simulation environment in real time. The operation data includes equipment current and voltage, the environmental data includes temperature, humidity, and wind speed data, and the geological data includes soil density data and groundwater level data;

[0080] Update the 3D model in real time and perform time series alignment on data with different sampling frequencies.

[0081] Specifically, a 3D laser scanner can be used to scan the live working area, the operation data is obtained through sensors, the geological data is collected through a ground penetrating radar, and through a two-way communication architecture, the 3D model is updated in real time using the synchronization engine Unity Reflect, and the dynamic time warping DTW is used to perform time series alignment on data with different sampling frequencies.

[0082] It should be noted that high-precision spatial geometric information is provided through the point cloud data, which can truly restore the physical structure of the live working area, including the shape, size, and relative position of the equipment. The simulation environment generated by the 3D model can cover complex spatial layouts, facilitating subsequent risk assessment and operation planning. The visual 3D simulation environment provides an intuitive operation platform for risk assessment, operation simulation, and personnel training; through high-frequency sampling, subtle fluctuations in equipment current and voltage can be captured, which helps to detect potential faults such as short circuits and arc discharges in a timely manner. By integrating environmental data into the simulation model, the actual operation conditions can be simulated more realistically, improving the reliability of risk assessment. By combining geological data with environmental and equipment data, it provides support for comprehensive risk assessment in all aspects and with multiple factors; through DTW to align data with different sampling frequencies, the problem of inconsistent sampling frequencies of equipment data, environmental data, and geological data is solved, ensuring the consistency of data fusion.

[0083] In a preferred embodiment, constructing a knowledge graph structure includes:

[0084] Taking the energized equipment and data items as nodes to generate a node set. The attribute of each node is the risk index value of the node. The risk index value of the node includes abnormal humidity data, abnormal temperature data, and abnormal voltage data (the abnormal standard is determined based on the safety standard of the corresponding node equipment), and the connection cost of the node includes geographical distance, resistance value, and the processing time value of the node;

[0085] The normalized deviation value of the real-time value of the operating data of the node from the safety standard threshold is used as the operating outlier. The normalized deviation value of the real-time value of the environmental data of the node from the safety standard threshold is used as the environmental outlier. The normalized deviation value of the real-time value of the address data of the node from the safety standard threshold is used as the geological outlier. The combined operating outlier, environmental outlier, and geological outlier are used as the comprehensive risk value of the node;

[0086] Based on the ratio of the comprehensive risk value of the node to the processing time value of the node, the priority of the node tasks is sorted. Among them, the processing time value of the node refers to the time required from risk identification to completion of disposal when a risk occurs or intervention is required for a certain node (i.e., a live working device or data item) in the live working risk assessment system;

[0087] According to the electrical connection relationship of the nodes, the node pairs are constructed as edges and an edge set is formed;

[0088] Calculate the link cost based on the cost and pipe connection of the physical connection of the nodes;

[0089] According to the node set, edge set, and link cost, all nodes and edges are constructed into a directed graph G according to the link cost, and the initial link cost is used as the weight of the node.

[0090] Exemplarily, based on the cost and pipe connection of the physical connection of the nodes, the link cost is defined as:

[0091]

[0092] Among them, Cij represents the distance between node i and node j, Pi represents the priority sorting value of node i, ρ represents the wire resistivity, A represents the wire cross-sectional area, d ij represents the distance between node i and node j.

[0093] In a preferred embodiment, the dynamic calculation of the shortest path includes:

[0094] Real-time monitor the priority sorting of the nodes, and update the node status according to the change value of the node priority sorting value, which is expressed as:

[0095] Wij = W'ij + ΔPi·dij

[0096] Among them, Wij represents the new weight of node i and node j, W'ij represents the state weight of node i and node j at the previous time, ΔPi represents the change value of the priority sorting of node i, d ij represents the distance between node i and node j;

[0097] Through the Bellman-Ford algorithm, perform dynamic shortest path calculation to find the path with the lowest propagation risk in the directed graph, which is expressed as:

[0098]

[0099] Among them, Py represents the optimal path of current risk propagation, and E represents the set of edges.

[0100] It should be noted that by organically integrating operation, environment, and geological data, clarifying the physical and logical relationships between nodes through a knowledge graph, facilitating multi-dimensional risk modeling and analysis, facilitating the discovery of potential logical links of risk propagation through the structured data organization method of the knowledge graph, improving the comprehensiveness and accuracy of risk analysis, generating the comprehensive risk value of nodes through normalizing and calculating operation outliers, environment outliers, and geological outliers, realizing risk quantification and standardization processing, supporting dynamic adjustment of monitoring and operation strategies through task priority sorting, timely responding to risk changes of high-priority nodes, enabling the system to adapt to on-site data changes and reflect the latest risk distribution status through real-time monitoring and updating of node priority sorting, and supporting negative weight edges through the Bellman-Ford algorithm, which can effectively handle possible negative influencing factors in risk propagation.

[0101] In a preferred embodiment, determining the three-dimensional grid coordinates of nodes, calculating the risk value information of the grid, and optimizing the risk propagation of abnormal points, the correction of the observed value includes:

[0102] Based on the simulation environment and the calculated directed graph, determine the three-dimensional grid coordinates of each node, and update and calculate the comprehensive risk value of the grid according to the optimal path of current risk propagation, expressed as:

[0103]

[0104] Among them, P(x, y, z, t) represents the comprehensive risk value of the grid point (x, y, z), Pi represents the risk value of node i, Dij represents the Euclidean distance between the node and the grid point, σ represents the risk standard deviation value, and N is the total number of nodes;

[0105] Define the influence function of the node on the grid point through the Gaussian diffusion model of the distance between the node and the grid point, expressed as:

[0106]

[0107] Among them, f(x, y, z, Ni) represents the influence value of the node on the grid point;

[0108] Calculate the mean deviation of the grid point risk value from its neighborhood points, and take the sum of the mean and standard deviation of the historical deviation values as the deviation threshold. If the mean deviation is greater than or equal to the deviation threshold, it is determined as an abnormal grid point;

[0109] Construct a wave equation through wavelength reversal simulation, expressed as:

[0110]

[0111] Among them, R(x, y, z, t) represents the distribution value of the risk value over time and space, v represents the risk propagation speed, and the propagation rate is fitted through historical risk event data. Represents the symbol of partial derivative, ▽ 2 Represents the Laplace operator, ▽ 2 R(x, y, z, t) represents the second-order partial derivative of R(x, y, z, t) with respect to space, indicating the speed at which the risk value changes in space. Represents the second-order partial derivative of R(x, y, z, t) with respect to time, indicating the acceleration of the risk value over time.

[0112] Optimize the propagation of outliers through time reversal and perform correction based on the observed values, expressed as:

[0113]

[0114] Among them, R J (x, y, z, t) represents the risk value of the corrected outlier, Δt represents the time change value, P(x, y, z, t) represents the comprehensive risk value of the grid point (x, y, z), and dt represents the integration variable.

[0115] Using hydrodynamic simulation technology, assuming that the risk propagation is an unsteady fluid, based on the risk value of the corrected outlier, calculate the risk change of each grid point, expressed as:

[0116] P(x, y, z, t + Δt) = P(x, y, z, t) + Δt·[-▽·(F) + S(x, y, z, t)]

[0117]

[0118] F = -▽P(x, y, z, t)

[0119]

[0120] Among them, P(x, y, z, t + Δt) represents the comprehensive risk value at the next time, ▽·F represents the divergence calculation of the risk propagation flux F, S(x, y, z, t) represents the risk source term of the grid point, and Fx, Fy, and Fz respectively represent the classification of the comprehensive risk value in the x, y, and z directions (obtained through the gradient calculation of the comprehensive risk value).

[0121] It should be noted that through the three-dimensional simulation environment, the spatial structure of the live working area can be comprehensively restored, providing a real physical background for risk assessment. By calculating the three-dimensional grid coordinates and grid risk values of nodes, the refined division of the scene is realized, facilitating zonal management and local risk control. Through the Gaussian diffusion model, the risk diffusion range and intensity of nodes can be accurately described, quantifying the impact of nodes on surrounding grid points. By calculating the influence value of nodes on grid points, the risk level of local high-risk areas can be evaluated, facilitating priority treatment. By calculating the mean deviation between the grid point risk value and its neighborhood points and setting the historical deviation threshold, abnormal grid points can be dynamically identified. Through the wave equation combined with time reversal technology, the risk value of abnormal points is optimized according to the historical propagation law, improving the reliability of abnormal correction. Through time reversal technology, the cumulative deviation caused by noise or abnormal data is eliminated by backtracking the propagation history, ensuring the stability of the risk propagation model. Through the dynamic update of the risk value of abnormal points after correction, the input for subsequent grid risk calculation becomes more accurate.

[0122] By using the assumptions of hydrodynamic simulation and based on the calculation formulas of flux and gradient, the dynamic diffusion of risk in the grid space is accurately simulated. By combining flux calculation with the risk source term S(x, y, z, t), the influence of external risk sources on grid points is dynamically simulated, making the risk model closer to the real scenario. Through dynamic optimal path, grid risk value update, and time reversal correction, the real-time dynamic adaptation of risk propagation in complex scenarios is achieved. Through grid-based risk value calculation, Gaussian diffusion model, and anomaly detection technology, a multi-dimensional fine risk assessment from nodes to grid points is provided. Combining the content of calculating the dynamic shortest path through the Bellman-Ford algorithm in the previous text, the optimal risk propagation path after updating the grid risk value is adjusted in real time. This dynamic adjustment enables the system to quickly respond to newly emerging high-risk areas, ensuring that the risk propagation path is always optimal. Combining the result of the shortest path calculation with the node priority ranking makes resource scheduling and allocation more scientific, avoiding excessive monitoring or intervention in low-risk areas. The node priority ranking provides a basis for global optimization, while the shortest path further clarifies the local optimization strategy, realizing the collaborative management of global and local risks.

[0123] In a preferred embodiment, the anomaly detection includes:

[0124] Calculating the change rate of the comprehensive risk value of grid points between two consecutive time steps, and calculating the average risk value change rate of all grid points in the scene;

[0125] Performing a fast Fourier transform on the time series data of the risk value change to obtain a frequency domain signal, finding the frequency point with the largest amplitude in the spectrum, and taking the reciprocal of the spectrum amplitude of the largest frequency point as the main frequency period value;

[0126] Based on the main frequency period value, the periodic anomaly ratio of the main frequency period is calculated to perform anomaly detection, which is expressed as:

[0127]

[0128] Where Rz represents the periodic anomaly ratio of the main frequency period, F(fz) represents the energy of the main frequency component of the main frequency period value, T represents the total number of frequencies, and F(fk) represents the amplitude of the frequency component of frequency fk;

[0129] The sum of the mean and standard deviation of the historical periodic anomaly ratio is taken as the anomaly threshold. If the periodic anomaly ratio is greater than or equal to the anomaly threshold, it is judged as a periodic anomaly.

[0130] It should be noted that by calculating the rate of change of the risk value of the grid point, the risk change of each grid point can be dynamically captured, especially the response to sudden events is more agile. The average risk value change rate of all grid points in the scene quantifies the speed of overall risk change, which helps to judge the risk evolution trend of the entire scene and provide a quantitative basis for global decision-making. The calculation of the risk value change rate can quickly identify areas with drastic risk changes, support priority resource allocation and intervention, and extract the frequency domain signal through FFT to find the main frequency and period of the risk change of the grid point, revealing the periodic law in the risk fluctuation. After finding the main frequency period value, the periodic abnormal characteristics of the risk fluctuation can be detected, and regular risk events (such as the period of equipment failure) can be effectively identified. By calculating the energy percentage of the main frequency component, the risk anomalies caused by periodic events (such as periodic environmental disturbances or equipment failures) can be accurately identified. The periodic anomaly percentage directly reflects the structural characteristics of risk changes and provides an explainable indicator for data-based intelligent analysis. Combined with the risk value change rate calculation and frequency domain signal extraction, multi-dimensional dynamic risk monitoring of grid points, local areas and global scenarios is realized. The calculation of the periodic anomaly percentage is combined with the historical threshold to support the accurate detection of periodic anomalies, which is particularly suitable for monitoring regular risk events. Through periodic anomaly detection, the periodic operation failures of equipment can be identified in time, reducing safety accidents caused by equipment failures and realizing risk assessment in live working tests.

[0131] In a preferred embodiment, visually displaying the detected abnormal link and generating a data report including operation data include:

[0132] According to the optimized comprehensive risk value, upload the three-dimensional heat map through the image upload tool for visual display, highlighting the key nodes and their risk diffusion links;

[0133] Sound and light alarms are issued based on periodic anomalies to warn testers working with live wires;

[0134] Statistically analyze and store the operation data of the testers for live working, and transmit the data to a remote terminal and the cloud via wireless transmission for storage and backup.

[0135] Align the operation data of the testers with the operation data, environmental data, and geological data collected by the sensors in terms of time stamps, and generate a data report through a report generation tool.

[0136] It should be noted that by using the optimized comprehensive risk value, visualizing the risk distribution through a three-dimensional heat map, showing the risk level at each position in the grid, the three-dimensional heat map clearly displays the overall picture of the risk distribution, enabling testers and decision-makers to intuitively understand the location and severity of high-risk areas. By highlighting the key nodes and their risk diffusion links, it is convenient to quickly identify the core nodes of risk propagation and the areas that may be affected, thus providing accurate reference for adjusting the operation strategy. The visual display supports remote collaboration. Through an image upload tool, other departments or teams can quickly obtain the risk status, enhancing the efficiency of team collaboration.

[0137] Through cloud backup, it is ensured that the operation data is not affected by local device failures, improving the security and reliability of data storage. The collected operation data can be used for statistical analysis to help identify high-risk or inefficient links in the operation, thereby optimizing the test process and improving the operation efficiency. The data report clearly displays the key data in the form of charts, facilitating testers and managers to quickly grasp the device status, environmental conditions, and operation records.

[0138] The present invention clarifies the physical and logical relationships between nodes through a knowledge graph, facilitating multi-dimensional risk modeling and analysis. Through real-time monitoring and updating of node priority sorting, the system can adapt to on-site data changes and reflect the latest risk distribution status. By supporting negative weight edges through the Bellman-Ford algorithm, it can effectively handle the possible negative influencing factors in risk propagation. Through the Gaussian diffusion model, it can accurately describe the risk diffusion range and intensity of nodes, quantifying the influence of nodes on surrounding grid points. By calculating the influence value of nodes on grid points, the risk level of local high-risk areas can be evaluated, facilitating priority processing. By combining the wave equation with time reversal technology, the risk value of abnormal points is optimized according to historical propagation laws, improving the reliability of anomaly correction.

[0139] The above is a schematic solution of a live working risk assessment method based on multi-source data in this embodiment. It should be noted that the technical solution of the live working risk assessment system based on multi-source data belongs to the same concept as the technical solution of the above-mentioned live working risk assessment method based on multi-source data. For the details not described in detail in the technical solution of the live working risk assessment system based on multi-source data in this embodiment, reference can be made to the description of the technical solution of the above-mentioned live working risk assessment method based on multi-source data.

[0140] The live working risk assessment system based on multi-source data in this embodiment includes:

[0141] A collection module, configured to collect equipment data, environmental data, and geological data, and perform preprocessing;

[0142] A construction module, configured to construct a knowledge graph structure based on the equipment data, environmental data, and geological data, calculate the link cost, analyze the node priority, and dynamically calculate the shortest path;

[0143] A calculation module, configured to determine the three-dimensional grid coordinates of the nodes based on the knowledge graph structure, calculate the risk value information of the grid, optimize the risk propagation of the abnormal points, correct the observed values, and calculate the risk change of each grid point;

[0144] A detection module, configured to calculate the main frequency period value based on the risk change and perform anomaly detection;

[0145] An output module, configured to visually display the detected abnormal links and generate a data report including operation data.

[0146] This embodiment also provides an electronic device applicable to the situation of live working risk assessment based on multi-source data, including:

[0147] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the live working risk assessment method based on multi-source data as proposed in the above embodiment.

[0148] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the live working risk assessment method based on multi-source data as proposed in the above embodiment.

[0149] The storage medium proposed in this embodiment and the live working risk assessment method based on multi-source data proposed in the above embodiment belong to the same inventive concept. For the technical details not described in detail in this embodiment, reference can be made to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0150] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general hardware. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk or optical disc of a computer, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0151] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for risk assessment of live working based on multi-source data, characterized in that: include: Collect equipment data, environmental data and geological data, and perform pre-processing; Based on the equipment data, environmental data and geological data, a knowledge graph structure is constructed, link costs are calculated, node priorities are analyzed, and the shortest path is dynamically calculated; Based on the knowledge graph structure, determine the three-dimensional grid coordinates of the nodes, calculate the risk value information of the grid, optimize the risk propagation of the outliers, correct the observed values, and calculate the risk change of each grid point; Based on the risk change, the main frequency period value is calculated to perform anomaly detection; Visualize the detected abnormal links and generate data reports including operation data.

2. The method for risk assessment of live working based on multi-source data according to claim 1, characterized in that: The acquisition equipment data, environmental data and geological data include: Scan the live working area and generate point cloud data; Convert the point cloud data into a three-dimensional model to generate a simulation environment; Collecting in real time the operation data, environmental data and geological data of the live equipment in the simulation environment, wherein the operation data includes equipment current and voltage, the environmental data includes temperature, humidity and wind speed data, and the geological data includes soil density data and groundwater level data; The three-dimensional model is updated in real time, and the data with different sampling frequencies are time-series aligned.

3. The method for risk assessment of live working based on multi-source data according to claim 1 or 2, characterized in that: Building a knowledge graph structure includes: The powered device and the data item are taken as nodes to generate a node set, each node attribute is a risk index value of the node, the risk index value of the node includes humidity data anomaly, temperature data anomaly and voltage data anomaly, and the connection cost of the node includes geographical distance, resistance value and processing time value of the node; The normalized value of the deviation between the real-time value of the operation data of the node and the safety standard threshold is used as the operation abnormality value, the normalized value of the deviation between the real-time value of the environmental data of the node and the safety standard threshold is used as the environmental abnormality value, the normalized value of the deviation between the real-time value of the address data of the node and the safety standard threshold is used as the geological abnormality value, and the operation abnormality value, environmental abnormality value and geological abnormality value are combined to serve as the comprehensive risk value of the node; Prioritize the node tasks based on the ratio of the comprehensive risk value of the node to the processing time value of the node; According to the electrical connection relationship of the nodes, node pairs are constructed as edges and formed into edge sets; Calculating link costs based on the cost of physical connections of the nodes and pipe connections; According to the node set, edge set and link cost, all nodes and edges are constructed into a directed graph G according to the link cost, and the initial link cost is used as the weight of the node.

4. The method for risk assessment of live working based on multi-source data according to claim 3, characterized in that: The dynamic calculation of the shortest path includes: The priority ranking of the node is monitored in real time, and the node status is updated according to the change value of the node priority ranking value, which is expressed as: Wij=W'ij+ΔPi·dij Where Wij represents the new weights of nodes i and j, W'ij represents the state weights of nodes i and j at the previous time, ΔPi represents the priority ranking change value of node i, and d ij Represents the distance between node i and node j; The Bellman-Ford algorithm is used to perform dynamic shortest path calculation to find the path with the lowest propagation risk in the directed graph.

5. The method for risk assessment of live working based on multi-source data according to claim 4, characterized in that: Determine the three-dimensional grid coordinates of the nodes, calculate the risk value information of the grid, optimize the risk propagation of outliers, and correct the observed values ​​including: Based on the simulation environment and the calculated directed graph, determine the three-dimensional grid coordinates of each node, and update the comprehensive risk value of the calculation grid according to the optimal path of current risk propagation; The influence function of nodes on grid points is defined through the Gaussian diffusion model of the distance between nodes and grid points; Calculate the mean deviation between the risk value of the grid point and the neighboring points, and use the sum of the mean and standard deviation of the historical deviation values ​​as the deviation threshold. If the mean deviation is greater than or equal to the deviation threshold, it is judged as an abnormal grid point; Through the wavelength reversal simulation, the wave equation is constructed and expressed as: Among them, R(x,y,z,t) represents the distribution of risk value over time and space, v represents the speed of risk propagation, represents the symbol of partial derivative, represents the Laplace operator; The propagation of outliers is optimized by time reversal and corrected according to the observed values, which can be expressed as: Among them, R J (x, y, z, t) represents the corrected outlier risk value, Δt represents the time change value, P(x, y, z, t) represents the comprehensive risk value of the grid point (x, y, z), and dt represents the integral variable; Based on the corrected outlier risk value, the risk change of each grid point is calculated, which is expressed as: P(x,y,z,t+Δt)=P(x,y,z,t)+Δt·[-▽·(F)+S(x,y,z,t)] Among them, P(x,y,z,t+Δt) represents the comprehensive risk value at the next time, ▽·F represents the divergence calculation of the risk propagation flux F, and S(x,y,z,t) represents the risk source term of the grid point.

6. The method for risk assessment of live working based on multi-source data according to claim 1, characterized in that: The anomaly detection includes: Calculating the rate of change of the comprehensive risk value of the grid point between two consecutive time steps, and calculating the average rate of change of the risk value of all grid points in the scene; Perform fast Fourier transform on the time series data of risk value changes to obtain frequency domain signals, and find the frequency point with the largest amplitude in the spectrum. The inverse of the spectrum amplitude of the largest frequency point is taken as the main frequency period value. Based on the main frequency period value, the periodic anomaly ratio of the main frequency period is calculated to perform anomaly detection; The sum of the mean and standard deviation of the historical periodic anomaly ratio is used as the anomaly threshold. If the periodic anomaly ratio is greater than or equal to the anomaly threshold, it is judged as a periodic anomaly.

7. The method for risk assessment of live working based on multi-source data according to claim 6, characterized in that: Visualize the detected abnormal links and generate data reports including operation data, including: According to the optimized comprehensive risk value, upload a three-dimensional heat map for visual display, highlighting the key nodes and their risk diffusion links; Based on the periodic anomaly, an audible and visual alarm is issued to warn the test personnel who are working with live wires; The operation data of testers performing live operations are counted and stored, and the data is transmitted to remote terminals and the cloud via wireless transmission for storage and backup; The operating data, environmental data and geological data collected by the sensors are time-stamped and aligned with the tester's operation data to generate data reports.

8. A live working risk assessment system based on multi-source data, characterized in that: include, The collection module is used to collect equipment data, environmental data and geological data and perform preprocessing; A construction module, used to construct a knowledge graph structure, calculate link costs, analyze node priorities, and dynamically calculate the shortest path based on the device data, environmental data, and geological data; A calculation module is used to determine the three-dimensional grid coordinates of the nodes based on the knowledge graph structure, calculate the risk value information of the grid, optimize the risk propagation of the outliers, correct the observed values, and calculate the risk change of each grid point; A detection module, used to calculate the main frequency period value based on the risk change and perform abnormality detection; The output module is used to visualize the detected abnormal links and generate data reports including operation data.

9. An electronic device, characterized in that: include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the live working risk assessment method based on multi-source data as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, implement the steps of the live working risk assessment method based on multi-source data as described in any one of claims 1 to 7.

Images