Hydraulic engineering safety monitoring method and device, electronic equipment and storage medium

Abstract

The invention discloses a hydraulic engineering safety monitoring method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring monitoring data of a target water conservancy project and a corresponding evaluation index tree; the method comprises the following steps: pre-constructing a large safety monitoring model comprising a data association sub-model, a special collaboration sub-model and a dynamic evaluation sub-model to realize fusion processing on monitoring data associated with evaluation indexes, correction processing on a simulation result of a special model and targeted scoring processing on the evaluation indexes; and finally, according to the score value of each evaluation index, multi-dimensional risk visualization information of the target water conservancy project is generated, and early warning is carried out according to the multi-dimensional risk visualization information. According to the technical scheme, the structured evaluation knowledge base fused with the industry standard is constructed, and the large model is utilized to realize autonomous perception, intelligent association and deep analysis of the multi-source data, so that real-time dynamic full-dimensional risk early warning of the operation state of the water conservancy project is realized.

Description

Technical Field

[0001] This application relates to the field of monitoring technology for water conservancy projects, and in particular to a method, device, electronic equipment storage and medium for safety monitoring of water conservancy projects. Background Technology

[0002] Water conservancy projects are a collective term for projects related to flood control, drainage, irrigation, power generation, water supply, land reclamation, soil and water conservation, resettlement, and water resource protection, as well as their supporting and ancillary works. Examples include reservoirs, dams, spillways, power stations, and ship locks. The safe and stable operation of water conservancy projects has beneficial impacts on the ecological environment, natural landscape, regional climate, and socio-economic development. However, water conservancy projects are subject to water pressure, buoyancy, seepage, erosion, scouring, and freezing during operation. Furthermore, the varying hydrological, topographical, and geological conditions of the regions where these projects are located make their operation and management quite complex.

[0003] The current safety monitoring and operation management of water conservancy projects have the following problems: 1) Data value is not fully explored: multi-source heterogeneous data such as seepage, deformation, video, and inspection records are isolated and lack intelligent correlation analysis and fusion reasoning based on business rules; 2) Evaluation is static and one-sided: traditional evaluations rely on periodic manual inspections or fixed models, which cannot achieve dynamic and comprehensive scoring based on real-time data, and often focus on physical safety prediction while ignoring the standardization and compliance evaluation of operation management.

[0004] Therefore, how to provide a technical solution that can provide real-time, dynamic, and multi-dimensional risk warning for the operation status of water conservancy projects is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This application provides a method, device, electronic equipment, and storage medium for safety monitoring of water conservancy projects. By constructing a structured evaluation knowledge base that integrates industry standards and utilizing a large model to achieve autonomous perception, intelligent correlation, and in-depth analysis of multi-source data, it enables real-time, dynamic, and multi-dimensional risk warning of the operational status of water conservancy projects.

[0006] According to one aspect of this application, a method for safety monitoring of water conservancy projects is provided, the method comprising: Acquire monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project; wherein, the monitoring data includes structured data, unstructured data and time series data, the evaluation index tree is constructed based on preset evaluation rules in a knowledge base, the leaf nodes of the evaluation index tree are evaluation indicators, and the evaluation indicators include a first indicator and a second indicator, the first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator. Based on the data association sub-model in the pre-built security monitoring big model, the monitoring data corresponding to each evaluation index is associated according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, and a fusion feature vector is generated; wherein, the preset data source mapping rules are used to represent the mapping relationship between the evaluation index and the monitoring data, and the preset association reasoning rules are used to represent the association logic of the monitoring data. Based on the specialized collaborative sub-model in the aforementioned safety monitoring big model, simulation data of the specialized model is obtained, and the simulation data is corrected according to the monitoring data to obtain early warning data; wherein, the specialized model is used to perform simulation analysis on at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain characteristic data for safety evaluation; Based on the dynamic evaluation sub-model in the aforementioned safety monitoring big model, according to the preset evaluation rules, the rule engine is invoked to make logical judgments on the early warning data and the fusion feature vector of the first indicator, and the logical reasoning module is invoked to reason and evaluate the fusion feature vector of the second indicator to determine the score value of each evaluation indicator. Based on the scores of each evaluation indicator, multi-dimensional risk visualization information of the target water conservancy project is generated, so as to provide early warning based on the multi-dimensional risk visualization information.

[0007] According to another aspect of this application, a water conservancy project safety monitoring device is provided, characterized in that the device comprises: The data acquisition module is used to acquire monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project; wherein, the monitoring data includes structured data, unstructured data and time series data, the evaluation index tree is constructed based on preset evaluation rules in the knowledge base, the leaf nodes of the evaluation index tree are evaluation indicators, and the evaluation indicators include a first indicator and a second indicator, the first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator; The data association module is used to associate the monitoring data corresponding to each evaluation indicator based on the data association sub-model in the pre-built security monitoring big model, according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, and generate a fusion feature vector; wherein, the preset data source mapping rules are used to represent the mapping relationship between the evaluation indicators and the monitoring data, and the preset association reasoning rules are used to represent the association logic of the monitoring data. The data correction module is used to obtain simulation data of a special model based on a special collaborative sub-model in the large safety monitoring model, and to correct the simulation data according to the monitoring data to obtain early warning data; wherein, the special model is used to perform simulation analysis on at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain feature data for safety evaluation; The data scoring module is used to determine the score value of each evaluation indicator by calling the rule engine to make logical judgments on the early warning data and the fusion feature vector of the first indicator based on the dynamic evaluation sub-model in the large safety monitoring model and according to the preset evaluation rules, and by calling the logical reasoning module to reason and evaluate the fusion feature vector of the second indicator. The risk warning module is used to generate multi-dimensional risk visualization information of the target water conservancy project based on the score values ​​of each evaluation indicator, so as to provide early warning based on the multi-dimensional risk visualization information.

[0008] According to another aspect of this application, an electronic device is provided, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the water conservancy project safety monitoring method according to any embodiment of this application.

[0009] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the water conservancy project safety monitoring method according to any embodiment of this application.

[0010] According to another aspect of this application, a computer program product is provided, the computer program product including a computer program that, when executed by a processor, implements the water conservancy project safety monitoring method described in any embodiment of this application.

[0011] The technical solution provided in this application acquires monitoring data of the target water conservancy project and its corresponding evaluation index tree; pre-constructs a large-scale safety monitoring model including a data association sub-model, a specialized collaborative sub-model, and a dynamic evaluation sub-model to achieve fusion processing of monitoring data associated with the evaluation indicators, correction processing of simulation results from specialized models, and targeted scoring processing of the evaluation indicators; finally, based on the score values ​​of each evaluation indicator, it generates multi-dimensional risk visualization information of the target water conservancy project for early warning. This technical solution, by constructing a structured evaluation knowledge base integrating industry standards and utilizing a large-scale model to achieve autonomous perception, intelligent association, and in-depth analysis of multi-source data, enables real-time, dynamic, and multi-dimensional risk early warning of the water conservancy project's operational status.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of a water conservancy project safety monitoring method provided in Embodiment 1 of this application.

[0015] Figure 2 This is a flowchart of a water conservancy project safety monitoring method provided in Embodiment 2 of this application.

[0016] Figure 3 This is a flowchart of a water conservancy project safety monitoring method provided in Embodiment 3 of this application.

[0017] Figure 4 This is a schematic diagram of the structure of a water conservancy project safety monitoring device provided in Embodiment 4 of this application.

[0018] Figure 5 This is a schematic diagram of the structure of a device for implementing a water conservancy project safety monitoring method according to an embodiment of this application. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0020] It should be noted that the terms "first," "second," "original," "intermediate," "monitoring," and "to be reviewed," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] Example 1 Figure 1 This is a flowchart illustrating a water conservancy project safety monitoring method provided in Embodiment 1 of this application. This embodiment is applicable to situations involving safety monitoring of water conservancy projects. The method can be executed by a water conservancy project safety monitoring device, which can be implemented in hardware and / or software. This device can be configured in equipment with data processing capabilities. Figure 1 As shown, the method includes the following steps.

[0022] S110. Obtain monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project. The monitoring data includes structured data, unstructured data, and time-series data. The evaluation index tree is constructed based on preset evaluation rules in a knowledge base. The leaf nodes of the evaluation index tree are evaluation indicators. The evaluation indicators include a first indicator and a second indicator. The first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator.

[0023] Among them, monitoring data is used to monitor the safety status and operation and management level of the target water conservancy project. It can be obtained through various sensors deployed in the target water conservancy project and its surrounding environment, manual inspection systems, water conservancy-related business systems, etc.

[0024] Based on their source, monitoring data can be categorized into structured data, unstructured data, and time-series data. Unstructured data can include visual and audio data, document data, etc., such as visual and audio data acquired from fixed cameras, drone inspection video streams, and sound recordings of important equipment operation; and document data such as safety assessment reports, emergency plans, management system documents, and acceptance reports uploaded manually. Structured data can be business process data, such as daily inspection records, operation logs, and maintenance work orders acquired by the inspection system. Time-series data can be IoT sensing data, such as data monitored in real time by GNSS displacement monitors, piezometers, stress gauges, and vibration sensors.

[0025] The evaluation index tree is a hierarchical, tree-structured evaluation index system, consisting of a root node, multiple intermediate nodes, and leaf nodes. Each leaf node can contain the standard score, deduction threshold, evaluation period, etc.

[0026] For example, the root node of the evaluation index tree is the spillway gate, and the intermediate nodes can be "Engineering Status," "Safety Management," "Operation and Maintenance," "Management Support," and "Information Technology Construction." The next level nodes under "Engineering Status" can be "Engineering Appearance and Environment," "Gate Chamber," "Gate," "Hinges and Electromechanical Equipment," "Upstream and Downstream Rivers and Embankments," "Management Facilities," and "Signage." The leaf nodes under "Engineering Appearance and Environment" are "Is the Engineering Appearance Poor?", "Is the Engineering Management Scope Disorganized?", "Is the Greening Rate Low in Suitable Greening Areas within the Engineering Management Scope?", and "Does the Engineering Management Scope Experience Soil Erosion or Poor Aquatic Ecological Environment?". The standard score for the node "Engineering Appearance and Environment" is 25 points. The evaluation period for the evaluation index "Is the Engineering Appearance Poor?" is daily. If the evaluation result corresponding to the evaluation index is "Poor Engineering Appearance," 10 points are deducted.

[0027] To facilitate rapid and accurate scoring of the evaluation indicators, this application divides them into two categories: a first category of deterministic evaluation indicators and a second category of complex logical evaluation indicators. The first category refers to indicators that can be directly measured by instruments and quantified and precisely calculated using specific numerical values, such as "whether the document has been uploaded" or "whether the sensor value exceeds the threshold." The second category refers to indicators that are difficult to measure precisely with continuous numerical values ​​and require expert observation, inspection, analysis, and evaluation based on standards or experience, such as "whether the project's appearance is poor" or "the degree of equipment corrosion."

[0028] The knowledge base stores preset evaluation rules, which can be design specification clauses, safety assessment standards, industry guidelines, expert experience rules, etc. For example, preset evaluation rules can be industry standards such as the "Standardized Scoring Table for Spillway Gates." Specifically, evaluation index trees can be constructed by dragging and dropping through a graphical interface, or by importing predefined index tree templates and adaptively modifying them according to the preset evaluation rules.

[0029] S120. Based on the data association sub-model in the pre-constructed safety monitoring large model, and according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, the monitoring data corresponding to each evaluation indicator is associated to generate a fusion feature vector. The preset data source mapping rules represent the mapping relationship between the evaluation indicators and the monitoring data, and the preset association reasoning rules represent the association logic of the monitoring data.

[0030] Among them, the safety monitoring big model refers to an artificial intelligence system that integrates multimodal data processing, professional knowledge-driven approaches, physical mechanism fusion, and dynamic intelligent decision-making for the field of water conservancy project safety monitoring. Safety detection, as the intelligent analysis hub, consists of three collaborative sub-models: a data association sub-model, a specialized collaborative sub-model, and a dynamic evaluation sub-model.

[0031] The preset data source mapping rules exist in the knowledge base in the form of configuration tables or knowledge graphs, which are used to clarify which monitoring data are used to judge the evaluation indicators. For example, the evaluation indicator "whether the project appearance is clean" is mapped to "daily photos captured by designated cameras"; the evaluation indicator "whether the safety assessment report is submitted on time" is mapped to "the upload time and content of 'safety assessment' type documents in the document library"; and the evaluation indicator "whether there is water seepage in the gate chamber" is mapped to "the inspection results of the 'wall seepage' item in the inspection system".

[0032] Among them, the pre-defined association reasoning rules are used to define the judgment logic of complex evaluation indicators.

[0033] The fusion feature vector is a structured vector with fixed or variable dimensions. Each element in the fusion feature vector represents a feature highly correlated with the evaluation indicator, extracted from monitoring data associated with the evaluation indicator. It should be noted that the fusion feature vector can have one or more dimensions.

[0034] In this application, a graph neural network can be used to automatically associate structured data, unstructured data, and time-series data based on preset data source mapping rules and preset association inference rules. Specifically, monitoring data associated with a certain evaluation index can be characterized and then concatenated or arranged into a multimodal sequence according to the preset data source mapping rules; in the self-attention calculation, the attention bias matrix is ​​determined according to the preset association inference rules, and an improved Transformer encoder is used to process the multimodal sequence to obtain a fused feature vector.

[0035] S130. Based on the specialized collaborative sub-model in the aforementioned safety monitoring large model, simulation data of the specialized model is obtained, and the simulation data is corrected according to the monitoring data to obtain early warning data. The specialized model is used to simulate and analyze at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain characteristic data for safety evaluation.

[0036] Among them, the special collaborative sub-model is used to call or access evaluation indicators that rely on professional simulation results, and compare, correct and integrate them with monitoring data to provide deeper physical mechanism support for evaluation.

[0037] Among them, specialized models are mathematical or physical simulation models constructed for a specific function or problem of a water conservancy project. For example, specialized models for simulating and analyzing the mechanical response of a target water conservancy project include: finite element stress analysis models; specialized models for simulating and analyzing the physical field of a target water conservancy project include: seepage field simulation models and flood simulation models.

[0038] Taking the seepage field simulation model as an example, using dam piezometer data, reservoir water level data, and geological parameters, and solving Darcy's law governing equations, the spatial distribution cloud map, equipotential lines, streamlines, and escape point locations of pore water pressure in the dam body and foundation at a specific water level are simulated and calculated. Furthermore, characteristic data such as seepage gradient and unit width seepage flow rate on key sections are extracted from the simulation results as simulation data. Since the predicted data obtained from the specialized model simulation has certain deviations, the simulation data obtained from the specialized model can be corrected by combining monitoring data to obtain early warning data.

[0039] S140. Based on the dynamic evaluation sub-model in the large safety monitoring model, according to the preset evaluation rules, the rule engine is called to make logical judgments on the early warning data and the fusion feature vector of the first indicator, and the logical reasoning module is called to reason and evaluate the fusion feature vector of the second indicator to determine the score value of each evaluation indicator.

[0040] The dynamic evaluation sub-model acts as the executor of the scoring, employing a hybrid approach of a rule engine and a logical reasoning module to perform scoring. This enables differentiated and automated evaluation of both deterministic and complex logical evaluation indicators. The rule engine is a configurable component that executes corresponding business rules based on explicit thresholds, numerical ranges, rates of change, or simple statistical relationships. The logical reasoning module can be a component for comprehensive judgment, experience comparison, and uncertain reasoning when rigid mathematical formulas are lacking. The score can be either a deduction or a gain for the evaluation indicator.

[0041] Specifically, for evaluation indicators with clearly defined evaluation rules, such as "whether the document has been uploaded" or "whether the sensor value exceeds the threshold", as well as early warning data simulated by specialized models, the rule engine can be directly called to perform logical judgments such as Boolean operations or numerical comparisons, and the score value of the evaluation indicator can be determined based on the logical judgment result.

[0042] For indicators requiring complex understanding, such as "poor engineering appearance" or "equipment corrosion level," the logic reasoning module can be invoked to process the fused feature vectors output by the data association sub-model, or the monitoring data can be processed directly to output a score. For example, the computer vision analysis model built into or integrated into the logic reasoning module can be invoked to perform multi-label recognition and segmentation on camera images, determine the cleanliness of the environment, and output a score.

[0043] S150. Based on the scores of each evaluation indicator, generate multi-dimensional risk visualization information of the target water conservancy project, and issue early warnings based on the multi-dimensional risk visualization information.

[0044] Multidimensional risk visualization information refers to a digital interface that integrates multiple dimensions such as space, time, logic, and risk level, presented using various visual elements such as graphics, charts, maps, and animations. Specifically, it can automatically summarize the scores of all evaluation indicators according to a preset scoring cycle or triggered event to obtain the real-time safety score of the current target water conservancy project. At the same time, it can visualize the scores and risk levels of each category and part in the form of heat maps, dashboards, etc., to intuitively locate weak links.

[0045] An alert can be triggered when the score of an evaluation indicator or the total score falls below a threshold, causing a sudden increase in the risk level. Based on the risk level, the alert will automatically notify the relevant management personnel through platform messages, SMS, and app push notifications.

[0046] This invention provides a method for safety monitoring of water conservancy projects. The method acquires monitoring data of the target water conservancy project and its corresponding evaluation index tree; pre-constructs a large-scale safety monitoring model including a data association sub-model, a specialized collaborative sub-model, and a dynamic evaluation sub-model; this model enables the fusion processing of monitoring data associated with the evaluation indicators, the correction of simulation results from specialized models, and targeted scoring of the evaluation indicators; finally, based on the scores of each evaluation indicator, it generates multi-dimensional risk visualization information for the target water conservancy project, providing early warning based on this information. This technical solution, by constructing a structured evaluation knowledge base integrating industry standards and utilizing a large-scale model to achieve autonomous perception, intelligent association, and in-depth analysis of multi-source data, enables real-time, dynamic, and multi-dimensional risk early warning of the water conservancy project's operational status.

[0047] Example 2 Figure 2 This is a flowchart of a water conservancy project safety monitoring method provided in Embodiment 2 of this application. This embodiment is based on the above embodiment and optimized, specifically by optimizing the process of acquiring monitoring data. Figure 2 As shown, the method in this embodiment specifically includes the following steps.

[0048] S210. Obtain the raw data of the target water conservancy project and the evaluation index tree corresponding to the target water conservancy project. The evaluation index tree is constructed based on preset evaluation rules in a knowledge base. The leaf nodes of the evaluation index tree are evaluation indicators, including a first indicator and a second indicator. The first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator.

[0049] Raw data refers to the initial, unprocessed data set directly collected or recorded from various monitoring terminals, equipment, and systems at the water conservancy project site. Examples include: IoT sensing data (real-time monitoring data from GNSS displacement, piezometers, stress gauges, vibration sensors, etc.); visual and audio data (video streams from fixed cameras and drone inspections, and sound recordings of important equipment operation); operational process data (daily inspection records, operation logs, and maintenance work orders obtained from inspection systems such as Jin Siwei); and document data (manually uploaded safety assessment reports, emergency plans, management system documents, acceptance reports, etc.).

[0050] S220. Perform data cleaning, noise reduction and normalization on the original data to obtain intermediate data.

[0051] Data cleaning refers to removing errors, inconsistencies, and missing values ​​from raw data. For example, physically impossible outliers caused by sensor malfunctions, such as sudden deformation exceeding the measurement range, can be identified and removed; reasonable data gaps caused by communication interruptions can be filled using interpolation methods.

[0052] Noise reduction refers to filtering out random high-frequency fluctuations in data caused by environmental interference, electronic noise, etc., and extracting effective signals that reflect the true state of the project. For example, wavelet transform or Kalman filtering can be applied to vibration acceleration data to separate noise from structural vibration signals.

[0053] Normalization refers to scaling monitoring data of different dimensions and orders of magnitude to a unified, relatively standard numerical range (such as [0, 1] or [-1, 1]) through linear or nonlinear transformation. This helps eliminate the influence of different physical quantity units, enabling subsequent correlation analysis and feature fusion models to work more stably and efficiently. For example, stress (unit MPa, value between 0 and 10) and displacement (unit mm, value between 0 and 100) are both normalized to the 0-1 range.

[0054] S230. Add a unified spatiotemporal identifier to the intermediate data, and map and align the intermediate data with the spatial coordinates of the preset engineering information model to generate monitoring data.

[0055] The spatiotemporal identifiers include time identifiers and spatial identifiers. Time identifiers typically use a unified Coordinated Universal Time (UTC) or the standard time of the target water conservancy project's location to ensure all data is strictly aligned on the timeline. Spatial identifiers are unique location codes or coordinates associated with the data point or described object within a pre-defined engineering information model. For example, for the displacement data of "GPS point of section 3 on the downstream face of the dam," its corresponding model coordinates (X1234.5, Y5678.9, Z100.0) are added.

[0056] The preset engineering information model is a high-precision mapping of the target water conservancy project in virtual space, which includes, but is not limited to, building information model, geographic information system model, oblique photogrammetry model, or point cloud model. In this embodiment, a BIM model is preferably used to associate building component attributes, and a GIS model is used to associate geospatial coordinates.

[0057] Specifically, a mapping table can be pre-established to correspond the spatial location of each physical sensor and monitoring point to the preset engineering information model. Intermediate data with spatial identifiers can be mapped to the corresponding spatial location in the preset engineering information model, so that the monitoring data can be transformed from an abstract numerical sequence into locatable engineering status information with clear spatial ownership.

[0058] S240. Based on the data association sub-model in the pre-constructed safety monitoring large model, and according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, the monitoring data corresponding to each evaluation indicator is associated to generate a fused feature vector. The preset data source mapping rules represent the mapping relationship between the evaluation indicators and the monitoring data, and the preset association reasoning rules represent the association logic of the monitoring data.

[0059] Specifically, graph neural networks can be used to automatically associate monitoring data corresponding to the same evaluation indicator based on preset data source mapping rules and preset association inference rules.

[0060] In some embodiments, optionally, monitoring data corresponding to each evaluation indicator are associated according to preset data source mapping rules and preset association reasoning rules in the knowledge base to generate a fusion feature vector, including: for each evaluation indicator in the evaluation indicator tree, extracting associated data under the same spatiotemporal window from the monitoring data according to the preset data source mapping rules in the knowledge base; performing fusion calculation and logical reasoning on the associated data according to the preset association reasoning rules in the knowledge base to generate a fusion feature vector.

[0061] In this context, a spatiotemporal window refers to the geographical and temporal coverage of the window. For example, a spatiotemporal window could be the past 12 hours of a defined area.

[0062] Specifically, the time and space ranges of the spatiotemporal window can be predefined first; then, based on the preset data source mapping rules in the knowledge base, related data associated with the evaluation indicators under the same spatiotemporal window can be extracted from the monitoring data pool; finally, based on the preset association reasoning rules, the related data can be fused and logically reasoned to generate a fused feature vector.

[0063] For example, the displacement data sequence, seepage data sequence and the content describing "seepage" in the inspection text of the same spatiotemporal window are associated to generate a fusion feature vector that comprehensively represents the seepage state of the area.

[0064] The advantage of the above technical solution is that by defining a spatiotemporal window and mapping time, space, and data, it can flexibly cope with complex evaluation scenarios.

[0065] S250. Based on the special collaborative sub-model in the large safety monitoring model, obtain the simulation data of the special model, and correct the simulation data according to the monitoring data to obtain the early warning data.

[0066] The specific model is used to simulate and analyze at least one mechanical response and / or at least one physical field of the target water conservancy project in order to obtain characteristic data for safety evaluation.

[0067] S260. Based on the dynamic evaluation sub-model in the large safety monitoring model, according to the preset evaluation rules, the rule engine is called to make logical judgments on the early warning data and the fusion feature vector of the first indicator, and the logical reasoning module is called to reason and evaluate the fusion feature vector of the second indicator to determine the score value of each evaluation indicator.

[0068] S270. Based on the scores of each evaluation indicator, generate multi-dimensional risk visualization information of the target water conservancy project, so as to provide early warning based on the multi-dimensional risk visualization information.

[0069] This invention provides a method for safety monitoring of water conservancy projects. This method, by cleaning, denoising, normalizing, and spatiotemporally aligning the monitoring data, facilitates the automatic and accurate correlation of monitoring data corresponding to evaluation indicators, thereby improving the accuracy of safety monitoring of water conservancy projects.

[0070] Example 3 Figure 3 This is a flowchart of a water conservancy project safety monitoring method provided in Embodiment 3 of this application. This embodiment is based on the above embodiment but with optimizations, specifically optimizing the processing of the fused feature vectors. Figure 3 As shown, the method in this embodiment specifically includes the following steps.

[0071] S310. Obtain monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project.

[0072] The monitoring data includes structured data, unstructured data, and time-series data. The evaluation index tree is constructed based on preset evaluation rules in the knowledge base. The leaf nodes of the evaluation index tree are evaluation indicators. The evaluation indicators include a first indicator and a second indicator. The first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator.

[0073] S320. Based on the data association sub-model in the pre-built safety monitoring big model, according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, the monitoring data corresponding to each of the evaluation indicators are associated to generate a fusion feature vector.

[0074] The preset data source mapping rule is used to represent the mapping relationship between the evaluation index and the monitoring data, and the preset association reasoning rule is used to represent the association logic of the monitoring data; the fusion feature vector is a vector obtained by fusing monitoring data from at least two data sources.

[0075] In this application, the fused feature vector can be a vector obtained by fusing monitoring data from at least two data sources. For example, the fused feature vector of the deformation characteristics of dam section #5 at 10:00 on January 1, 2025 can be represented as {measured displacement, theoretical predicted displacement, displacement residual, and displacement residual change rate}.

[0076] S330. Based on the special collaborative sub-model in the large safety monitoring model, obtain the simulation data of the special model, and correct the simulation data according to the monitoring data to obtain the early warning data.

[0077] The specific model is used to simulate and analyze at least one mechanical response and / or at least one physical field of the target water conservancy project in order to obtain characteristic data for safety evaluation.

[0078] S340. Based on the dynamic evaluation sub-model in the large safety monitoring model, according to the preset evaluation rules, the rule engine is called to perform logical judgment on the monitoring data of each data source in the fusion feature vector of the first indicator, and the logical reasoning module is called to perform reasoning evaluation on the monitoring data of each data source in the fusion feature vector of the second indicator, so as to determine the scoring result corresponding to the monitoring data of each data source in the evaluation indicator.

[0079] Specifically, logical judgments or inferences can be performed on the monitoring data from each data source in the fused feature vector. For example, for the evaluation indicator "whether there is water seepage in the gate chamber", the corresponding monitoring data include the seepage pressure parameters measured by the piezometer in the gate chamber and the infrared images collected by the infrared thermal imager. The scoring results for the seepage pressure parameters and the infrared images can be determined based on preset evaluation rules.

[0080] S350. If the scoring results for the same evaluation indicator are consistent, then the score value of the evaluation indicator shall be determined based on the scoring results.

[0081] When the scoring results for monitoring data from different data sources are consistent, such as when the seepage pressure parameter of the evaluation indicator "whether there is water seepage in the gate chamber" is normal and the infrared image shows no abnormalities, the evaluation indicator "whether there is water seepage in the gate chamber" can be determined to have no deductions based on the preset evaluation rules, and the score value is 10 points.

[0082] S360. If the scoring results for the same evaluation indicator are inconsistent, the scoring results corresponding to the monitoring data of each data source are weighted and decided based on the historical confidence level of each data source to determine the score value of the evaluation indicator.

[0083] When different data sources or judgment methods lead to conflicting conclusions, such as when the evaluation indicator "whether there is water seepage in the gate chamber" is abnormal and triggers an alarm but no abnormality is seen in the infrared image, the dynamic evaluation sub-model can integrate the historical confidence or historical accuracy of each data source, assign different weight coefficients to each data source, and perform a weighted sum based on the weight coefficients and the corresponding scoring results of the monitoring data of each data source to determine the score value of the evaluation indicator.

[0084] In some embodiments, the method may optionally further include: after determining the score value of the evaluation indicator by weighting the scoring results corresponding to the monitoring data of each data source based on the historical confidence level of each data source, the method further includes: marking the evaluation indicator as an indicator to be reviewed, and obtaining the manual review result of the indicator to be reviewed; and determining the score value of the evaluation indicator to be reviewed based on the manual review result.

[0085] In cases where the scoring results of monitoring data from different data sources are inconsistent, this application provides a method for weighted decision-making based on confidence level. However, this method still has certain errors. Therefore, this application also provides a method for re-determining the scoring value based on manual review.

[0086] Specifically, evaluation indicators with evaluation conflicts can be marked as indicators to be reviewed. Relevant personnel can then use the interactive interface to score, confirm, modify, or reject these indicators. The results of manual review of these indicators by relevant personnel can then be obtained to determine the score value of the evaluation indicators to be reviewed.

[0087] In some embodiments, optionally, after determining the score value of the evaluation indicator to be reviewed based on the manual review result, the method further includes: pairing the evaluation indicator to be reviewed, the score value of the evaluation indicator to be reviewed, and the fusion feature vector corresponding to the evaluation indicator to be reviewed to determine the first training data; and fine-tuning the security monitoring big model according to the first training data.

[0088] To improve the accuracy of the dynamic evaluation sub-model in scoring evaluation indicators with evaluation conflicts, this application pairs the indicators to be reviewed with the corresponding review results as training data and periodically performs incremental fine-tuning on the safety monitoring big model. This allows the dynamic evaluation sub-model in the safety monitoring big model to learn the scoring standards of manual review, reduce false alarms and false negatives, and improve the accuracy of automatic scoring.

[0089] S370. Based on the scores of each evaluation indicator, generate multi-dimensional risk visualization information of the target water conservancy project, so as to provide early warning based on the multi-dimensional risk visualization information.

[0090] This invention provides a method for safety monitoring of water conservancy projects. This method optimizes the evaluation process of fused feature vectors with two or more dimensions. When the scoring results of monitoring data from different dimensions are inconsistent, a confidence-weighted decision method is used for scoring, which improves the scoring efficiency and accuracy of evaluation indicators, thereby realizing real-time dynamic and multi-dimensional risk warning of the operation status of water conservancy projects.

[0091] Based on the above embodiments, optionally, after generating multi-dimensional risk visualization information of the target water conservancy project according to the score values ​​of each evaluation indicator, and issuing an early warning based on the multi-dimensional risk visualization information, the method further includes: generating rectification suggestions based on the early warning trigger indicators, the monitoring data corresponding to the trigger indicators, historical case sets, and preset handling procedures in the knowledge base, and obtaining rectification results; pairing the trigger indicators, the early warning rectification suggestions, and the rectification results to determine the second training data; and fine-tuning the knowledge base based on the second training data.

[0092] Among them, the trigger indicator refers to the evaluation indicator that reaches the warning threshold, the monitoring data corresponding to the trigger indicator refers to the monitoring data at the time of triggering the warning, the historical case set refers to the handling records and results of similar historical warnings, and the preset handling procedure refers to the standard handling process predefined in the knowledge base.

[0093] Specifically, after an alert is triggered, the system can search and match the triggering indicators and corresponding monitoring data in the knowledge base to obtain the handling records and results of similar historical cases. Finally, based on the handling records and results of similar historical cases and the preset handling procedures, rectification suggestions can be automatically generated.

[0094] For example, suggestions for an early warning of "leaking water at the gate stop" could include: "1. Increasing the frequency of observations to 2 hours / time; 2. Focusing on inspecting the rubber waterstop; 3. Referring to the maintenance records of similar problems at gate X in 2023."

[0095] Furthermore, early warnings and rectification suggestions can be simultaneously pushed to the command center's large screen, the PC management backend, and the on-site inspection terminal, forming a collaborative handling workflow. On-site personnel receive tasks and provide feedback on the handling process and results through the on-site inspection terminal, such as uploading post-repair photos and filling out processing records, and record the rectification process data and effect verification data in a structured manner. The rectification effect refers to whether the relevant indicator data has returned to normal after the system tracks and issues the early warning.

[0096] This application pairs the warning trigger indicators, rectification suggestions, and rectification results as training data, and periodically performs incremental fine-tuning on the preset handling procedures in the knowledge base to improve the accuracy of rectification suggestions.

[0097] In addition, for scoring indicators that frequently require manual intervention and adjustment, the interactive interface can prompt administrators to review the rationality of existing rules or data mappings, and support manual optimization and updates of scoring thresholds, weights, and even rule logic in the knowledge base, so that the evaluation system is more in line with management practice.

[0098] Example 4 Figure 4 This is a schematic diagram of the structure of a water conservancy project safety monitoring device provided in Embodiment 4 of the present invention. Figure 4 As shown, the device includes: The data acquisition module 410 is used to acquire monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project; wherein, the monitoring data includes structured data, unstructured data and time series data, the evaluation index tree is constructed based on preset evaluation rules in the knowledge base, the leaf nodes of the evaluation index tree are evaluation indicators, and the evaluation indicators include a first indicator and a second indicator, the first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator; The data association module 420 is used to associate the monitoring data corresponding to each evaluation index based on the data association sub-model in the pre-built security monitoring big model and according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, and generate a fusion feature vector; wherein, the preset data source mapping rules are used to represent the mapping relationship between the evaluation index and the monitoring data, and the preset association reasoning rules are used to represent the association logic of the monitoring data. The data correction module 430 is used to obtain simulation data of a special model based on a special collaborative sub-model in the large safety monitoring model, and to correct the simulation data according to the monitoring data to obtain early warning data; wherein, the special model is used to perform simulation analysis on at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain feature data for safety evaluation; The data scoring module 440 is used to, based on the dynamic evaluation sub-model in the large safety monitoring model, and according to the preset evaluation rules, call the rule engine to make logical judgments on the early warning data and the fusion feature vector of the first indicator, and call the logic reasoning module to reason and evaluate the fusion feature vector of the second indicator, so as to determine the score value of each evaluation indicator. The risk warning module 450 is used to generate multi-dimensional risk visualization information of the target water conservancy project based on the score values ​​of each evaluation indicator, so as to provide early warning based on the multi-dimensional risk visualization information.

[0099] The water conservancy project safety monitoring device provided in this invention acquires monitoring data of the target water conservancy project and its corresponding evaluation index tree; pre-constructs a large-scale safety monitoring model including a data association sub-model, a special collaborative sub-model, and a dynamic evaluation sub-model to achieve fusion processing of monitoring data associated with the evaluation indicators, correction processing of simulation results of special models, and targeted scoring processing of evaluation indicators; finally, based on the score values ​​of each evaluation indicator, it generates multi-dimensional risk visualization information of the target water conservancy project for early warning. This technical solution, by constructing a structured evaluation knowledge base integrating industry standards and utilizing a large-scale model to achieve autonomous perception, intelligent association, and in-depth analysis of multi-source data, enables real-time, dynamic, and multi-dimensional risk early warning of the water conservancy project's operational status.

[0100] Optionally, the data acquisition module 410 includes: The raw data acquisition unit is used to acquire the raw data of the target water conservancy project; The data preprocessing unit is used to perform data cleaning, noise reduction and normalization on the raw data to obtain intermediate data. The spatiotemporal alignment unit is used to add a unified spatiotemporal identifier to the intermediate data and map and align the intermediate data with the spatial coordinates of the preset engineering information model to generate monitoring data.

[0101] Optional, the data association module 420 includes: The data association unit is used to extract associated data under the same spatiotemporal window from the monitoring data for each evaluation indicator in the evaluation indicator tree, according to the preset data source mapping rules in the knowledge base. The data fusion unit is used to perform fusion calculations and logical reasoning on the associated data according to the preset association reasoning rules in the knowledge base, and generate a fusion feature vector.

[0102] Optionally, the fused feature vector is a vector obtained by fusing monitoring data from at least two data sources; Accordingly, the data scoring module 440 includes: The scoring result determination unit is used to call the rule engine to perform logical judgment on the monitoring data of each data source in the fusion feature vector of the first indicator according to the preset evaluation rules, and to call the logic reasoning module to perform reasoning evaluation on the monitoring data of each data source in the fusion feature vector of the second indicator, so as to determine the scoring result corresponding to the monitoring data of each data source in the evaluation indicator. The first scoring value determining unit is used to determine the scoring value of the evaluation indicator based on the scoring results if the scoring results for the same evaluation indicator are consistent. The second scoring value determining unit is used to determine the scoring value of the evaluation indicator by weighting the scoring results corresponding to the monitoring data of each data source based on the historical confidence level of each data source if the scoring results of each evaluation indicator are inconsistent.

[0103] Optionally, the data scoring module 440 further includes: The indicator review and marking unit is used to mark the evaluation indicator as an indicator to be reviewed if the scoring results for the same evaluation indicator are inconsistent. The third scoring unit is used to obtain the manual review results of the indicators to be reviewed, and to determine the scoring value of the evaluation indicators to be reviewed based on the manual review results.

[0104] Optionally, the device further includes: The first training data determination module is used to determine the first training data by pairing the evaluation index to be reviewed, the evaluation index to be reviewed, the evaluation index to be reviewed, and the fusion feature vector corresponding to the evaluation index to be reviewed after determining the score value of the evaluation index to be reviewed based on the manual review result. The large model fine-tuning module is used to fine-tune the security monitoring large model based on the first training data.

[0105] Optionally, the device further includes: The early warning and rectification module is used to generate multi-dimensional risk visualization information of the target water conservancy project based on the score values ​​of each evaluation indicator. After issuing an early warning based on the multi-dimensional risk visualization information, it generates rectification suggestions based on the trigger indicators of the early warning, the monitoring data corresponding to the trigger indicators, the historical case set, and the preset handling procedures in the knowledge base, and obtains the rectification results. The second training data determination module is used to pair the triggering indicator, the rectification suggestion of the warning, and the rectification result to determine the second training data. The knowledge base fine-tuning module is used to fine-tune the knowledge base based on the second training data.

[0106] The water conservancy project safety monitoring device provided in the embodiments of the present invention can execute the water conservancy project safety monitoring method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0107] Example 5 Figure 5 A schematic diagram of the structure of a device 10 that can be used to implement embodiments of this application is shown. The device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the application described and / or claimed herein.

[0108] like Figure 5 As shown, device 10 includes at least one processor 11 and a memory, such as read-only memory (ROM) 12, random access memory (RAM) 13, etc., communicatively connected to at least one processor 11. The memory stores computer programs executable by at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from storage unit 18. The RAM 13 may also store various programs and data required for the operation of device 10. The processor 11, ROM 12, and RAM 13 are interconnected via bus 14. Input / output (I / O) interface 15 is also connected to bus 14.

[0109] Multiple components in device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of monitors, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0110] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as methods for safety monitoring in hydraulic engineering projects.

[0111] In some embodiments, the water conservancy project safety monitoring method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the water conservancy project safety monitoring method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the water conservancy project safety monitoring method by any other suitable means (e.g., by means of firmware).

[0112] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0113] Computer programs used to implement the methods of this application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0114] In the context of this application, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0115] To provide interaction with a user, the systems and techniques described herein can be implemented on a device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0116] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0117] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0118] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from ROM 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of the embodiments of the present invention.

[0119] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.

[0120] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for safety monitoring of water conservancy projects, characterized in that, The method includes: Acquire monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project; wherein, the monitoring data includes structured data, unstructured data and time series data, the evaluation index tree is constructed based on preset evaluation rules in a knowledge base, the leaf nodes of the evaluation index tree are evaluation indicators, and the evaluation indicators include a first indicator and a second indicator, the first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator. Based on the data association sub-model in the pre-built security monitoring big model, the monitoring data corresponding to each evaluation index is associated according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, and a fusion feature vector is generated; wherein, the preset data source mapping rules are used to represent the mapping relationship between the evaluation index and the monitoring data, and the preset association reasoning rules are used to represent the association logic of the monitoring data. Based on the specialized collaborative sub-model in the aforementioned safety monitoring big model, simulation data of the specialized model is obtained, and the simulation data is corrected according to the monitoring data to obtain early warning data; wherein, the specialized model is used to perform simulation analysis on at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain characteristic data for safety evaluation; Based on the dynamic evaluation sub-model in the aforementioned safety monitoring big model, according to the preset evaluation rules, the rule engine is invoked to make logical judgments on the early warning data and the fusion feature vector of the first indicator, and the logical reasoning module is invoked to reason and evaluate the fusion feature vector of the second indicator to determine the score value of each evaluation indicator. Based on the scores of each evaluation indicator, multi-dimensional risk visualization information of the target water conservancy project is generated, so as to provide early warning based on the multi-dimensional risk visualization information.

2. The method according to claim 1, characterized in that, Obtain monitoring data for the target water conservancy project, including: Obtain the raw data of the target water conservancy project; The original data is cleaned, denoised, and normalized to obtain intermediate data; A unified spatiotemporal identifier is added to the intermediate data, and the intermediate data is mapped and aligned with the spatial coordinates of a preset engineering information model to generate monitoring data.

3. The method according to claim 2, characterized in that, Based on the preset data source mapping rules and preset association reasoning rules in the knowledge base, the monitoring data corresponding to each of the evaluation indicators are associated to generate a fused feature vector, including: For each evaluation indicator in the evaluation indicator tree, related data under the same spatiotemporal window are extracted from the monitoring data according to the preset data source mapping rules in the knowledge base. Based on the preset association reasoning rules in the knowledge base, the associated data is fused and logically reasoned to generate a fused feature vector.

4. The method according to claim 1, characterized in that, The fused feature vector is a vector obtained by fusing monitoring data from at least two data sources. Accordingly, based on the preset evaluation rules, the rule engine is invoked to perform logical judgment on the early warning data and the fused feature vector of the first indicator, and the logical reasoning module is invoked to perform reasoning evaluation on the fused feature vector of the second indicator, to determine the score value of each evaluation indicator, including: According to the preset evaluation rules, the rule engine is called to make logical judgments on the monitoring data of each data source in the fusion feature vector of the first indicator, and the logic reasoning module is called to reason and evaluate the monitoring data of each data source in the fusion feature vector of the second indicator, so as to determine the scoring results corresponding to the monitoring data of each data source in the evaluation indicator. If the scores for the same evaluation indicator are consistent, then the score value of the evaluation indicator is determined based on the scores for each indicator. If the scoring results for the same evaluation indicator are inconsistent, a weighted decision is made based on the historical confidence levels of each data source to determine the scoring value of the evaluation indicator corresponding to the monitoring data of each data source.

5. The method according to claim 4, characterized in that, After determining the score value of the evaluation indicator by weighting the scoring results corresponding to the monitoring data of each data source based on the historical confidence level of each data source, the method further includes: The evaluation indicators are marked as indicators to be reviewed, and the manual review results of the indicators to be reviewed are obtained. The score value of the evaluation indicator to be reviewed is determined based on the results of the manual review.

6. The method according to claim 5, characterized in that, After determining the score value of the evaluation indicator to be reviewed based on the results of the manual review, the method further includes: The evaluation index to be reviewed, the score value of the evaluation index to be reviewed, and the fusion feature vector corresponding to the evaluation index to be reviewed are paired to determine the first training data. The security monitoring model is fine-tuned based on the first training data.

7. The method according to claim 1, characterized in that, After generating multi-dimensional risk visualization information for the target water conservancy project based on the scores of each evaluation indicator, and issuing early warnings based on the multi-dimensional risk visualization information, the method further includes: Based on the warning trigger indicators, the monitoring data corresponding to the trigger indicators, the historical case set, and the preset handling procedures in the knowledge base, rectification suggestions are generated and rectification results are obtained. The triggering indicators, the rectification suggestions for the early warning, and the rectification results are paired to determine the second training data; The knowledge base is fine-tuned based on the second training data.

8. A safety monitoring device for water conservancy projects, characterized in that, The device includes: The data acquisition module is used to acquire monitoring data of the target water conservancy project and an evaluation index tree corresponding to the target water conservancy project; wherein, the monitoring data includes structured data, unstructured data and time series data, the evaluation index tree is constructed based on preset evaluation rules in the knowledge base, the leaf nodes of the evaluation index tree are evaluation indicators, and the evaluation indicators include a first indicator and a second indicator, the first indicator is a deterministic evaluation indicator, and the second indicator is a complex logic evaluation indicator; The data association module is used to associate the monitoring data corresponding to each evaluation indicator based on the data association sub-model in the pre-built security monitoring big model, according to the preset data source mapping rules and preset association reasoning rules in the knowledge base, and generate a fusion feature vector; wherein, the preset data source mapping rules are used to represent the mapping relationship between the evaluation indicators and the monitoring data, and the preset association reasoning rules are used to represent the association logic of the monitoring data. The data correction module is used to obtain simulation data of a special model based on a special collaborative sub-model in the large safety monitoring model, and to correct the simulation data according to the monitoring data to obtain early warning data; wherein, the special model is used to perform simulation analysis on at least one mechanical response and / or at least one physical field of the target water conservancy project to obtain feature data for safety evaluation; The data scoring module is used to determine the score value of each evaluation indicator by calling the rule engine to make logical judgments on the early warning data and the fusion feature vector of the first indicator based on the dynamic evaluation sub-model in the large safety monitoring model and according to the preset evaluation rules, and by calling the logical reasoning module to reason and evaluate the fusion feature vector of the second indicator. The risk warning module is used to generate multi-dimensional risk visualization information of the target water conservancy project based on the score values ​​of each evaluation indicator, so as to provide early warning based on the multi-dimensional risk visualization information.

9. An electronic device, characterized in that, The device includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the water conservancy project safety monitoring method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute the water conservancy project safety monitoring method according to any one of claims 1-7.

Images