Dam safety monitoring method and system based on knowledge association structure

Abstract

The invention discloses a dam safety monitoring method and system based on a knowledge association structure, and the method comprises the steps: collecting the multi-source heterogeneous data of a dam, and obtaining a dam safety knowledge graph; time sequence data in the dam safety knowledge graph are preprocessed, a deep learning model is adopted for modeling and analysis, anomaly detection is carried out, and potential safety hazards of the dam are recognized; the method comprises the following steps: constructing an integrated learning model in combination with multiple machine learning algorithms, performing fault diagnosis on potential safety hazards of a dam, generating dam early warning information, feeding back the dam early warning information to a dam safety knowledge graph, dynamically updating the dam safety knowledge graph in combination with historical data, and performing dam safety monitoring. Based on comprehensive integration of dam safety monitoring data, through modeling and analysis of the time series data by the deep learning model and data-based fault diagnosis of the model algorithm, the accuracy and robustness of dam fault diagnosis are improved, and safety monitoring and early warning of the dam are optimized.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydropower monitoring, and particularly relates to a dam safety monitoring method and system based on a knowledge association structure. Background Art

[0002] As an important water conservancy project facility, the safety of a dam is directly related to the safety of the lives and property of downstream users. Traditional dam safety monitoring methods mainly rely on manual inspections and regular detections, which have many limitations, such as limited monitoring scope, incomplete data collection, low monitoring efficiency, etc. With the continuous development of information technology, especially the rapid progress in fields such as big data and artificial intelligence, dam safety monitoring technologies are also constantly being updated.

[0003] For the safety monitoring of dams, as an improvement, currently a sensor network is used to collect various monitoring data of dams, such as displacement, deformation, seepage flow, stress, etc., and data analysis techniques are used to process and analyze the data to evaluate the safety of the dam. However, although various monitoring data of the dam can be collected currently, these data often come from different data sources with different formats and standards, resulting in great difficulties in data integration and utilization. In addition, current safety monitoring usually only focuses on the changes of single or multiple monitoring indicators, lacking in-depth analysis and utilization of the correlations between these indicators. And the current data collection and analysis for dam safety monitoring are not thorough enough, lacking in-depth diagnosis and early warning capabilities for potential safety hazards and fault types of the dam. For example, when abnormal data is detected, it is impossible to accurately judge the fault cause behind it and the possible influence range, resulting in insufficient accuracy of early warning information. To sum up, in the current dam safety monitoring, the degree of analysis of the correlations between various monitoring data is not deep, and the diagnosis and early warning capabilities for potential safety hazards and fault types need to be further improved. Summary of the Invention

[0004] The present invention provides a dam safety monitoring method and system based on a knowledge association structure, aiming to solve the problems in the current dam safety monitoring that the degree of analysis of the correlations between various monitoring data is not deep, and the diagnosis and early warning capabilities for potential safety hazards and fault types need to be further improved.

[0005] To achieve the above object, the present invention adopts the following technical solutions: The present invention provides a dam safety monitoring method based on a knowledge association structure, specifically including the following steps: S1. Based on the monitoring instruments in the dam, multi-source heterogeneous data of the dam is collected, and knowledge graph technology is used to structurally represent the multi-source heterogeneous data, obtaining a dam safety knowledge graph including nodes of dam components, monitoring indicators, fault types and their mutual relationships; S2. Preprocess the time-series data in the dam safety knowledge graph, use a deep learning model to model and analyze the preprocessed time-series data to obtain dam state characteristics; perform anomaly detection based on the dam state characteristics to identify potential safety hazards of the dam; S3. Construct an ensemble learning model by combining multiple machine learning algorithms. The ensemble learning model performs fault diagnosis based on the potential safety hazards of the dam, generates dam warning information according to the fault diagnosis results and the dam safety knowledge graph, and feeds the dam warning information back to the dam safety knowledge graph to dynamically update the dam safety knowledge graph in combination with historical data for dam safety monitoring.

[0006] In some embodiments, in S1, the multi-source heterogeneous data includes deformation monitoring data, seepage monitoring data, water regime monitoring data, environmental quantity monitoring data, and stress monitoring data.

[0007] Further, in S1, the collection of multi-source heterogeneous data also includes using an association rule mining algorithm to verify and filter the collected data; the nodes in the dam safety knowledge graph include dam components, monitoring indicators, and fault types, and the edges are the relationships between the nodes.

[0008] In some embodiments, in S2, the time-series data in the dam safety knowledge graph is formed by classifying the multi-source heterogeneous data and entering it according to the time sequence to form time-series measurement points or curves of the multi-source heterogeneous data, and aggregating the time-series measurement points or curves of the multi-source heterogeneous data.

[0009] Further, in S2, preprocessing the time-series data in the dam safety knowledge graph includes: cleaning, denoising, normalizing, and interpolating the time-series data.

[0010] In some embodiments, in S2, the deep learning models used include long short-term memory network models and gated recurrent unit models.

[0011] Further, in S2, based on the dam state characteristics extracted by the deep learning model, combined with the methods of reconstruction error and statistics, anomaly detection of time-series data is realized, and a dam anomaly threshold is set. When the real-time data of the dam exceeds the anomaly threshold, an alarm mechanism is triggered.

[0012] In some embodiments, in S3, the multiple machine learning algorithms combined in the ensemble learning model include random forest, gradient boosting tree, and support vector machine.

[0013] In some embodiments, in S3, establish communication between the dam warning information and the corresponding terminal device, and visually display the data corresponding to the dam safety monitoring.

[0014] The present invention also provides a dam safety monitoring system based on a knowledge association structure, including a data acquisition and atlas construction module, a data analysis module, and a fault diagnosis and early warning module, wherein: The data acquisition and atlas construction module is used to collect multi-source heterogeneous data of the dam, and use knowledge graph technology to structurally represent the multi-source heterogeneous data to obtain a dam safety knowledge graph; The data analysis module is used to preprocess the time-series data in the dam safety knowledge graph, use a deep learning model to model and analyze the preprocessed time-series data to obtain dam state characteristics, perform anomaly detection, and identify potential safety hazards of the dam; The fault diagnosis and early warning module is used to perform fault diagnosis based on the potential safety hazards of the dam using an integrated learning model, generate dam early warning information according to the fault diagnosis results and the dam safety knowledge graph, and feedback the dam early warning information to the dam safety knowledge graph, dynamically update the dam safety knowledge graph in combination with historical data, and perform dam safety monitoring.

[0015] Compared with the prior art, the dam safety monitoring method and system based on a knowledge association structure of the present invention have the following beneficial effects: A dam safety monitoring method based on a knowledge association structure according to the present invention specifically includes the following steps: S1. Collect multi-source heterogeneous data of the dam based on monitoring instruments in the dam, and use knowledge graph technology to structurally represent the multi-source heterogeneous data to obtain a dam safety knowledge graph including nodes such as dam components, monitoring indicators, and failure types and their mutual relationships; S2. Preprocess the time-series data in the dam safety knowledge graph, use a deep learning model to model and analyze the preprocessed time-series data to obtain dam state characteristics; perform anomaly detection based on the dam state characteristics to identify potential safety hazards of the dam; S3. Combine multiple machine learning algorithms to construct an ensemble learning model. The ensemble learning model performs fault diagnosis based on the potential safety hazards of the dam, generates dam warning information according to the fault diagnosis results and the dam safety knowledge graph, and feeds back the dam warning information to the dam safety knowledge graph to dynamically update the dam safety knowledge graph in combination with historical data for dam safety monitoring. The present invention collects various monitoring data of the dam based on monitoring instruments in the dam, uses knowledge graph technology to structurally represent the multi-source heterogeneous data to form a dam safety knowledge graph including nodes such as dam components, monitoring indicators, and failure types and their mutual relationships, extracts dam state characteristics through modeling and analysis using a deep learning model, and performs anomaly detection to identify potential safety hazards. The present invention combines multiple machine learning algorithms to construct an ensemble learning model, performs fault diagnosis based on the anomaly detection results, generates warning information, and feeds it back to the knowledge graph for dynamic update. The present invention realizes the comprehensive integration and structural representation of dam safety monitoring data, improves the extraction accuracy of dam state characteristics and the accuracy of anomaly detection through the modeling and analysis of time-series data by a deep learning model. Through the application of the ensemble learning model, the accuracy and robustness of fault diagnosis are improved, providing technical support for dam safety monitoring and warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings in the specification are used to provide a further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0017] Figure 1 It is a schematic diagram of the architecture of the system in a dam safety monitoring method and system based on a knowledge association structure according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown in the accompanying drawings here can be arranged and designed in various different configurations.

[0019] Accordingly, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0020] It should be noted that like reference numerals and letters denote like items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0021] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper", "lower", "horizontal", "inner", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the inventive product is customarily placed during use, it 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 should not be construed as a limitation of the present invention. In addition, terms such as "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0022] In addition, if the term "horizontal" appears, it does not mean that the component is required to be absolutely horizontal, but it can be slightly inclined. For example, "horizontal" only means that its direction is more horizontal relative to "vertical", and does not mean that the structure must be completely horizontal, but it can be slightly inclined.

[0023] In the description of the embodiments of the present invention, it should also be noted that unless otherwise clearly specified and limited, if terms such as "set", "installed", "connected", "connected" are understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and it 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.

[0024] How to effectively improve the extraction accuracy of dam state characteristics and the accuracy of anomaly detection, improve the accuracy and robustness of fault diagnosis, and provide technical support for the safety monitoring and early warning of dams.

[0025] As Figure 1 shown, a dam safety monitoring method based on a knowledge association structure of the present invention includes the following steps: S1. Collect multi-source heterogeneous data of the dam based on the monitoring instruments in the dam, and use knowledge graph technology to structurally represent the multi-source heterogeneous data, obtaining a dam safety knowledge graph including nodes such as dam components, monitoring indicators, and failure types and their mutual relationships. S2. Preprocess the time-series data in the dam safety knowledge graph, use a deep learning model to model and analyze the preprocessed time-series data to obtain dam state characteristics; perform anomaly detection based on the dam state characteristics to identify potential safety hazards of the dam. S3. Construct an ensemble learning model by combining multiple machine learning algorithms. The ensemble learning model performs fault diagnosis based on the potential safety hazards of the dam. According to the fault diagnosis results and the dam safety knowledge graph, generate dam warning information, and feedback the dam warning information to the dam safety knowledge graph, and dynamically update the dam safety knowledge graph in combination with historical data to perform safety monitoring of the dam.

[0026] The dam safety monitoring method based on the knowledge association structure of the present invention improves the effectiveness of dam state feature extraction by constructing a dam safety knowledge graph, integrating and structurally representing multi-source heterogeneous data, and using a deep learning model to model and analyze time-series data. The present invention constructs an ensemble learning model by combining multiple machine learning algorithms, improves the accuracy of fault diagnosis and warning information, and thus realizes comprehensive and accurate monitoring of the dam state. The present invention has better monitoring efficiency, diagnostic accuracy and warning ability, and has certain practical significance.

[0027] In some embodiments, the dam safety monitoring method based on the knowledge association structure of the present invention verifies and filters the collected data to ensure the accuracy and integrity of the data. At the same time, the meanings of nodes and edges in the knowledge graph are clarified, providing a basis for subsequent analysis and warning, and improving the accuracy and efficiency of dam safety monitoring.

[0028] Furthermore, the dam safety monitoring method based on the knowledge association structure of the present invention classifies multi-source heterogeneous data and enters them in chronological order to form time-series measurement points or curves. Combining these time-series measurement points or curves forms the time-series data in the dam safety knowledge graph, making the processing of time-series data more orderly and efficient.

[0029] Even further, the dam safety monitoring method based on the knowledge association structure of the present invention uses deep learning models such as long short-term memory network models and gated recurrent unit models to model and analyze the preprocessed time-series data, improving the extraction accuracy of dam state characteristics and the accuracy of anomaly detection.

[0030] Based on the dam state features extracted by the deep learning model, this invention realizes the anomaly detection of time series data by combining the methods of reconstruction error and statistic, and sets the dam anomaly threshold. When the real-time data of the dam exceeds the anomaly threshold, the alarm mechanism is triggered. Based on random forest, gradient boosting tree and support vector machine, this invention conducts fault diagnosis based on the potential safety hazards of the dam, thereby improving the accuracy of diagnosis.

[0031] This invention also provides a dam safety monitoring system based on the knowledge association structure, including a data acquisition and graph construction module, a data analysis module, and a fault diagnosis and early warning module, where: The data acquisition and graph construction module is used to collect multi-source heterogeneous data of the dam, and use the knowledge graph technology to structurally represent the multi-source heterogeneous data to obtain the dam safety knowledge graph; The data analysis module is used to preprocess the time series data in the dam safety knowledge graph, use the deep learning model to model and analyze the preprocessed time series data to obtain the dam state features, conduct anomaly detection, and identify the potential safety hazards of the dam; The fault diagnosis and early warning module is used to conduct fault diagnosis based on the potential safety hazards of the dam by using the integrated learning model, generate the dam early warning information according to the fault diagnosis result and the dam safety knowledge graph, and feedback the dam early warning information to the dam safety knowledge graph to dynamically update the dam safety knowledge graph in combination with historical data, and conduct dam safety monitoring. The system of this invention is used to implement the above-mentioned dam safety monitoring method, and effectively improves the efficiency and accuracy of monitoring through the modular carrier of the system.

[0032] The following further details a dam safety monitoring method and system based on the knowledge association structure of this invention through specific embodiments.

[0033] During the operation of the dam, various monitoring instruments are constantly acquiring multi-source heterogeneous data. These multi-source heterogeneous data reflect the basic situation of the dam operation. According to the data type, we can divide them into multiple categories. We list the following five basic categories: Deformation type, monitoring data reflecting the shape and structural deformation of the dam; seepage type, monitoring data reflecting the uplift pressure and seepage flow of the dam; water regime type, monitoring data reflecting the hydrological changes in the reservoir area and the basin; environmental quantity, monitoring data reflecting the natural environment of the reservoir area; stress type, monitoring data reflecting the structural stress, load, etc. of the dam.

[0034] Taking the above data as the standard, classify the dam monitoring data. After classification, each data is entered in chronological order, and the time-series measurement points or curves of various types of data can be formed. The time-series measurement points or curves of various types of data are combined to form the dam safety monitoring time-series database. After establishing the digital database, find the time-series data records of the required data and generate data tables or curve graphs.

[0035] Establish the time-series vector of the measurement point; There are often multiple parts or devices to be monitored in the dam, which are measurement points. Each measurement point will monitor multiple data simultaneously. Usually, when these measurement points fail, only a fault notification will be given, but no diagnostic analysis can be made. Generally, people need to conduct hidden danger investigation. The present invention provides a means for hidden danger investigation by using time-series vector analysis. The data types of each measurement point are nothing more than several major categories. According to the above example, the state of a measurement point is defined by five major categories of data, that is: State vector , through this five-dimensional state vector, according to geometric coordinate analysis, it is equivalent to the vector from the origin to the coordinate point in the five-dimensional coordinate system. When different data changes over time, the vector also changes accordingly. Record the changes, and by using the Apriori (association rule mining) algorithm, obtain the time-series vector change graph.

[0036] When a measurement point fails, one or more of the data monitored by the measurement point will show abnormalities. This kind of abnormality is not only the abnormality at the moment of failure, but it is possible that the data has been different from the normal value for some time before the failure occurs. Therefore, the failure is strongly correlated with the change of the time-series vector of the measurement point. By analyzing historical data, for each type of failure, there will be a corresponding change situation of the time-series vector. The change mode of a time-series vector before and after a failure is related to the time-series vectors of multiple adjacent measurement points.

[0037] Based on the above, when analyzing the failure of a specific measurement point, the present invention uses a kind of phenomenon analysis. First, according to the change of the time-series vector of measurement point A itself before and after the failure, analyze the type of the failure. Then use another kind of phenomenon to pay attention to the change of the time-series vectors of other measurement points adjacent to measurement point A. By analyzing the changes of adjacent measurement points, judge whether it is the cause of the failure of this point. If so, continue to analyze the degree to which it causes the failure of measurement point A. Which measurement points jointly cause the failure and the degree of association of each measurement point with the failure.

[0038] In the present invention, through cleaning, denoising, standardizing, and data interpolation preprocessing operations on the collected analysis and anomaly detection time series data, deep learning models of long short-term memory network (LSTM) and gated recurrent unit (GRU) are used to model and analyze the time series data, and extract the dam state features. Based on the features extracted by the deep learning model, combined with the reconstruction error and statistical methods, anomaly detection of the time series data is realized to identify potential safety hazards. The present invention combines multiple machine learning algorithms such as random forest, gradient boosting decision tree (GBDT), and support vector machine (SVM) to construct an ensemble learning model to achieve accurate diagnosis of dam faults. At the same time, according to the fault diagnosis results, combined with the fault impact analysis, emergency response, and historical data in the dam safety knowledge graph, warning information is generated, and the fault diagnosis and warning results are fed back to the dam safety knowledge graph to realize the dynamic update of the knowledge graph. At the same time, by establishing communication with the terminal, device, and visualization platform, the warning and diagnosis result data are displayed, so as to comprehensively improve the monitoring accuracy of the dam.

[0039] Finally, it should be noted that the above is only a preferred embodiment of the present invention, and does not impose any form of limitation on the present invention; any ordinary technical personnel in the industry can smoothly implement the present invention according to the instructions and the above description. Any equivalent changes made by using the technical content disclosed above and making some modifications, decorations, and evolutions are equivalent embodiments of the present invention; at the same time, any equivalent changes, modifications, and evolutions made to the above embodiments based on the essential technology of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A dam safety monitoring method based on knowledge association structure, characterized in that: The specific steps include: S1. Based on the monitoring instruments in the dam, the multi-source heterogeneous data of the dam is collected, and the knowledge graph technology is used to structure the multi-source heterogeneous data to obtain the dam safety knowledge graph containing the nodes of dam components, monitoring indicators, fault types and their mutual relationships; S2. Preprocess the time series data in the dam safety knowledge graph, use the deep learning model to model and analyze the preprocessed time series data, and obtain the dam status characteristics; Perform anomaly detection based on dam status characteristics to identify potential safety hazards of the dam; S3. Combine multiple machine learning algorithms to build an integrated learning model. The integrated learning model performs fault diagnosis based on the potential safety hazards of the dam. According to the fault diagnosis results and the dam safety knowledge graph, it generates dam warning information and feeds the dam warning information back to the dam safety knowledge graph. Combined with historical data, the dam safety knowledge graph is dynamically updated to monitor the safety of the dam.

2. The dam safety monitoring method based on knowledge association structure according to claim 1 is characterized in that: In S1, the multi-source heterogeneous data include deformation monitoring data, seepage monitoring data, water condition monitoring data, environmental quantity monitoring data and stress monitoring data.

3. The dam safety monitoring method based on knowledge association structure according to claim 2 is characterized in that: In S1, the collection of multi-source heterogeneous data also includes the use of association rule mining algorithms to verify and screen the collected data; the nodes in the dam safety knowledge graph include dam components, monitoring indicators, and fault types, and the edges are the relationships between nodes.

4. The dam safety monitoring method based on knowledge association structure according to claim 1 is characterized in that: In S2, the time series data in the dam safety knowledge graph is formed by classifying multi-source heterogeneous data, entering the time series measurement points or curves of the multi-source heterogeneous data in chronological order, and aggregating the time series measurement points or curves of the multi-source heterogeneous data.

5. The dam safety monitoring method based on knowledge association structure according to claim 4 is characterized in that: In S2, preprocessing the time series data in the dam safety knowledge graph includes: cleaning, denoising, standardizing and interpolating the time series data.

6. The dam safety monitoring method based on knowledge association structure according to claim 1 is characterized in that: In S2, the deep learning models used include a long short-term memory network model and a gated recurrent unit model.

7. The dam safety monitoring method based on knowledge association structure according to claim 6 is characterized in that: In S2, based on the dam status features extracted by the deep learning model, combined with the reconstruction error and statistical methods, anomaly detection of time series data is realized, and a dam anomaly threshold is set. When it is detected that the real-time data of the dam exceeds the anomaly threshold, an alarm mechanism is triggered.

8. The dam safety monitoring method based on knowledge association structure according to claim 1 is characterized in that: In the S3, the multiple machine learning algorithms combined in the integrated learning model include random forest, gradient boosting tree and support vector machine.

9. The dam safety monitoring method based on knowledge association structure according to claim 1 is characterized in that: In S3, the dam warning information is communicated with the corresponding terminal device, and the data corresponding to the dam safety monitoring is visualized.

10. A system based on the dam safety monitoring method based on knowledge association structure according to any one of claims 1 to 9, characterized in that: It includes data acquisition and graph construction module, data analysis module and fault diagnosis and early warning module, among which: The data collection and graph construction module is used to collect multi-source heterogeneous data of the dam, use knowledge graph technology to structure the multi-source heterogeneous data, and obtain the dam safety knowledge graph; The data analysis module is used to pre-process the time series data in the dam safety knowledge graph, use the deep learning model to model and analyze the pre-processed time series data, obtain the dam status characteristics, perform anomaly detection, and identify potential safety hazards of the dam; The fault diagnosis and early warning module is used to perform fault diagnosis based on the potential safety hazards of the dam using an integrated learning model, generate dam early warning information based on the fault diagnosis results and the dam safety knowledge graph, and feed the dam early warning information back to the dam safety knowledge graph. The dam safety knowledge graph is dynamically updated in combination with historical data to conduct dam safety monitoring.

Images