State monitoring method and system of hydraulic power plant computer monitoring system

Abstract

The embodiment of the invention provides a state monitoring method and system for a hydraulic power plant computer monitoring system, and the method comprises the steps: carrying out the semantic embedded coding and global significance fusion of each state evaluation index and each assembly working condition data of the hydraulic power plant computer monitoring system through employing an artificial intelligence technology based on deep learning; according to the method, the global semantic feature representation of the state evaluation index and the component working condition is obtained, and then fine-grained semantic interaction matching is carried out on the state evaluation index and the component working condition, so that the working state of the hydraulic power plant computer monitoring system is comprehensively understood, and the state grade is intelligently evaluated. Therefore, potential faults of the hydraulic power plant computer monitoring system can be found in time, corresponding measures can be taken in time to prevent further expansion of the faults, and reliability and safety of the hydraulic power plant computer monitoring system are improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of intelligent monitoring, and more specifically, to a method and system for monitoring the state of a computer monitoring system in a hydropower plant. Background Art

[0002] With the rapid development of information technology, as an important part of the country's energy supply, the automation level and intelligence degree of hydropower plants have been significantly improved. The computer monitoring system is increasingly widely used in hydropower plants, which can not only monitor the working state of generator sets in real time, but also effectively manage water resources to ensure the efficiency and safety of power production. However, with the increase in system complexity, how to accurately and timely evaluate the operating state of the computer monitoring system has become an urgent problem to be solved.

[0003] Traditional methods for monitoring the state of computer monitoring systems in hydropower plants mainly rely on manual experience and simple threshold judgment, and such methods have obvious limitations. On the one hand, due to the lack of a comprehensive understanding and in-depth analysis of the system state, traditional methods are difficult to detect potential fault hazards; on the other hand, with the explosion of data volume, the ability of traditional methods to process big data is also stretched.

[0004] Therefore, an optimized system and method for monitoring the state of a computer monitoring system in a hydropower plant are expected. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention is proposed. The embodiments of the present invention provide a method and system for monitoring the state of a computer monitoring system in a hydropower plant, which perform semantic embedding encoding and global saliency fusion on various state evaluation indicators and component working condition data of the computer monitoring system in the hydropower plant by using artificial intelligence technology based on deep learning to obtain global semantic feature representations of the state evaluation indicators and component working conditions. Furthermore, through fine-grained semantic interaction matching between the two, a comprehensive understanding of the working state of the computer monitoring system in the hydropower plant is realized, thereby intelligently evaluating its state level. In this way, potential faults of the computer monitoring system in the hydropower plant can be discovered in time, so as to take corresponding measures in time to prevent the further expansion of the faults, thereby improving the reliability and safety of the computer monitoring system in the hydropower plant.

[0006] In a first aspect, the embodiments of the present invention provide a method for monitoring the state of a computer monitoring system in a hydropower plant, which includes:

[0007] Count various state evaluation indicators of the computer monitoring system in the hydropower plant, and various component working condition data of the computer monitoring system in the hydropower plant;

[0008] Semantically embed and encode each of the state evaluation indicators and each component operating condition data to obtain a sequence of state evaluation indicator semantic embedding encoding vectors and a sequence of component operating condition data semantic embedding encoding vectors;

[0009] Perform feature significant aggregation on the sequence of state evaluation indicator semantic embedding encoding vectors and the sequence of component operating condition data semantic embedding encoding vectors respectively to obtain a state evaluation indicator global significant semantic aggregation representation vector and a component operating condition global significant semantic aggregation representation vector;

[0010] Perform feature matching interaction on the state evaluation indicator global significant semantic aggregation representation vector and the component operating condition global significant semantic aggregation representation vector to obtain a state evaluation indicator-component operating condition optimal alignment fusion feature vector;

[0011] Based on the state evaluation indicator-component operating condition optimal alignment fusion feature vector, determine the state level label of the computer monitoring system of the hydropower plant.

[0012] In some possible embodiments, the performing feature significant aggregation on the sequence of state evaluation indicator semantic embedding encoding vectors and the sequence of component operating condition data semantic embedding encoding vectors respectively to obtain a state evaluation indicator global significant semantic aggregation representation vector and a component operating condition global significant semantic aggregation representation vector includes:

[0013] Perform clustering analysis on the sequence of state evaluation indicator semantic embedding encoding vectors to obtain a state evaluation indicator self-supervised clustering representation vector;

[0014] Calculate the implicit clustering contribution factor of each state evaluation indicator semantic embedding encoding vector in the sequence of state evaluation indicator semantic embedding encoding vectors relative to the state evaluation indicator self-supervised clustering representation vector to obtain a sequence of state evaluation indicator semantic feature implicit clustering contribution factors;

[0015] Based on the sequence of state evaluation indicator semantic feature implicit clustering contribution factors, perform feature aggregation on the sequence of state evaluation indicator semantic embedding encoding vectors to obtain the state evaluation indicator global significant semantic aggregation representation vector.

[0016] In some possible embodiments, the calculating the implicit clustering contribution factor of each state evaluation indicator semantic embedding encoding vector in the sequence of state evaluation indicator semantic embedding encoding vectors relative to the state evaluation indicator self-supervised clustering representation vector to obtain a sequence of state evaluation indicator semantic feature implicit clustering contribution factors includes:

[0017] Calculate the dot-division vector between the semantic embedding coding vector of the state evaluation index and the self-supervised clustering representation vector of the state evaluation index, and calculate the base-2 logarithm of the absolute value of each eigenvalue in the dot-division vector to obtain the implicit clustering contribution weight vector;

[0018] Calculate the dot-product vector between the implicit clustering contribution weight vector and the semantic embedding coding vector of the state evaluation index, and calculate the exponential function value with base e and the sum of each eigenvalue of the dot-product vector as the exponent to obtain the implicit clustering contribution factor of the semantic feature of the state evaluation index.

[0019] In some possible embodiments, the feature aggregation of the sequence of the semantic embedding coding vectors of the state evaluation index based on the sequence of the implicit clustering contribution factors of the semantic features of the state evaluation index to obtain the globally significant semantic aggregation representation vector of the state evaluation index includes:

[0020] Arrange the sequence of the implicit clustering contribution factors of the semantic features of the state evaluation index into a state evaluation index semantic feature clustering contribution field distribution vector;

[0021] Perform explicit modeling based on the self-attention mechanism on the state evaluation index semantic feature clustering contribution field distribution vector to obtain a state evaluation index semantic feature clustering contribution field modulation weight vector;

[0022] Use each eigenvalue in the state evaluation index semantic feature clustering contribution field modulation weight vector as a weight to calculate the position-wise weighted sum of the sequence of the semantic embedding coding vectors of the state evaluation index to obtain the globally significant semantic aggregation representation vector of the state evaluation index.

[0023] In some possible embodiments, the optimal pairing multi-dimensional interaction of the set of the state evaluation index sub-component semantic feature vectors and the set of the component working condition sub-component semantic feature vectors to obtain the set of the state evaluation index-component working condition optimal sub-component paired semantic fusion feature vectors includes:

[0024] Perform optimal feature sub-component paired screening on the set of the state evaluation index sub-component semantic feature vectors and the set of the component working condition sub-component semantic feature vectors to obtain a set of optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairings;

[0025] Input each optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairing in the set of the optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairings into a feature multi-dimensional interaction module to obtain the set of the state evaluation index-component working condition optimal sub-component paired semantic fusion feature vectors.

[0026] In some possible embodiments, the optimal feature sub-component pairing screening of the set of semantic feature vectors of the state evaluation index sub-components and the set of semantic feature vectors of the component working condition sub-components to obtain the set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings includes:

[0027] For each semantic feature vector of the state evaluation index sub-components in the set of semantic feature vectors of the state evaluation index sub-components, calculate the hyperbolic space distance metric factor between it and each semantic feature vector of the component working condition sub-components in the set of semantic feature vectors of the component working condition sub-components, and select the semantic feature vector of the component working condition sub-components corresponding to the smallest hyperbolic space distance metric factor as the pairing object to obtain the set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings.

[0028] In some possible embodiments, the inputting each optimal {semantic feature vector of state evaluation index sub-components, semantic feature vector of component working condition sub-components} pairing in the set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings into the feature multi-dimensional interaction module to obtain the set of semantic fusion feature vectors of the optimal sub-component pairings of the state evaluation index - component working condition includes:

[0029] Calculate the element-wise addition, element-wise subtraction, and element-wise multiplication between the semantic feature vector of the state evaluation index sub-components and the semantic feature vector of the component working condition sub-components in the optimal {semantic feature vector of state evaluation index sub-components, semantic feature vector of component working condition sub-components} pairing respectively to obtain the first state evaluation index - component working condition optimal pairing interaction feature vector, the second state evaluation index - component working condition optimal pairing interaction feature vector, and the third state evaluation index - component working condition optimal pairing interaction feature vector;

[0030] Calculate the element-wise weighted sum between the first state evaluation index - component working condition optimal pairing interaction feature vector, the second state evaluation index - component working condition optimal pairing interaction feature vector, and the third state evaluation index - component working condition optimal pairing interaction feature vector to obtain the semantic fusion feature vector of the optimal sub-component pairing of the state evaluation index - component working condition.

[0031] In some possible embodiments, the sub-component feature semantic fusion of the set of semantic fusion feature vectors of the optimal sub-component pairings of the state evaluation index - component working condition to obtain the optimal alignment fusion feature vector of the state evaluation index - component working condition includes:

[0032] Cascade the set of state evaluation index-component operating condition optimal sub-allocation paired semantic fusion feature vectors to obtain the state evaluation index-component operating condition optimal alignment fusion feature vector.

[0033] In some possible embodiments, determining the state level label of the computer monitoring system of the hydropower plant based on the state evaluation index-component operating condition optimal alignment fusion feature vector includes:

[0034] Input the state evaluation index-component operating condition optimal alignment fusion feature vector into the state evaluation module based on the classifier to obtain a state evaluation result, and the state evaluation result is used to represent the state level label of the computer monitoring system of the hydropower plant.

[0035] In a second aspect, an embodiment of the present invention provides a state monitoring system for a computer monitoring system of a hydropower plant, including:

[0036] A data statistics module, configured to count each state evaluation index of the computer monitoring system of the hydropower plant and each component operating condition data of the computer monitoring system of the hydropower plant;

[0037] A semantic embedding encoding module, configured to perform semantic embedding encoding on each of the state evaluation indexes and each of the component operating condition data respectively to obtain a sequence of state evaluation index semantic embedding encoding vectors and a sequence of component operating condition data semantic embedding encoding vectors;

[0038] A feature significant aggregation module, configured to perform feature significant aggregation on the sequence of state evaluation index semantic embedding encoding vectors and the sequence of component operating condition data semantic embedding encoding vectors respectively to obtain a state evaluation index global significant semantic aggregation representation vector and a component operating condition global significant semantic aggregation representation vector;

[0039] A feature matching interaction module, configured to perform feature matching interaction on the state evaluation index global significant semantic aggregation representation vector and the component operating condition global significant semantic aggregation representation vector to obtain a state evaluation index-component operating condition optimal alignment fusion feature vector;

[0040] A state level label determination module, configured to determine the state level label of the computer monitoring system of the hydropower plant based on the state evaluation index-component operating condition optimal alignment fusion feature vector.

[0041] Compared with the prior art, a method and system for monitoring the state of a computer monitoring system in a hydropower plant provided by an embodiment of the present invention perform semantic embedding encoding and global saliency fusion on each state evaluation index and each component operating condition data of the computer monitoring system in the hydropower plant by using artificial intelligence technology based on deep learning to obtain a global semantic feature representation of the state evaluation index and the component operating condition. Furthermore, through fine-grained semantic interaction matching between the two, a comprehensive understanding of the working state of the computer monitoring system in the hydropower plant is achieved, so as to intelligently evaluate its state level. In this way, potential faults in the computer monitoring system of the hydropower plant can be detected in a timely manner, so as to take corresponding measures in a timely manner to prevent the further expansion of the faults, thereby improving the reliability and security of the computer monitoring system in the hydropower plant. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 It is a flowchart of the method for monitoring the state of a computer monitoring system in a hydropower plant according to an embodiment of the present invention;

[0044] Figure 2 It is a schematic diagram of data flow of the method for monitoring the state of a computer monitoring system in a hydropower plant according to an embodiment of the present invention;

[0045] Figure 3 It is a flowchart of sub-step S4 of the method for monitoring the state of a computer monitoring system in a hydropower plant according to an embodiment of the present invention;

[0046] Figure 4 It is a block diagram of the system for monitoring the state of a computer monitoring system in a hydropower plant according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below in conjunction with the drawings and specific embodiments. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the described embodiments of the present invention belong to the scope of protection of the present invention.

[0048] Unless otherwise specifically stated, the technical terms or scientific terms used in the embodiments of the present invention should have the ordinary meanings understood by those of ordinary skill in the art to which the present invention pertains. The use of "including" or "comprising" in the embodiments of the present invention neither limits the mentioned shapes, numbers, steps, actions, operations, components, elements and / or their groups, nor excludes the occurrence or addition of one or more other different shapes, numbers, steps, actions, operations, components, elements and / or their groups, or the addition of these. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity and order of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, the meaning of "a plurality" is two or more, unless otherwise specifically and clearly defined.

[0049] Unless otherwise specifically stated, the relative settings, numerical expressions and numerical values of the components and steps set forth in these embodiments do not limit the scope of the present invention. At the same time, it should be understood that for the sake of convenience of description, the sizes of the various parts shown in the drawings are not drawn in actual proportional relationships. For technologies, methods and devices known to those of ordinary skill in the relevant art, they may not be discussed in detail, but in appropriate cases, the shown technologies, methods and devices should be regarded as part of the authorization specification. In all the examples shown and discussed here, any specific other examples may have different values. It should be noted that: similar symbols and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0050] In the description of the embodiments of the present invention, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In the embodiments of the present invention, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in the embodiments of the present invention and the features of different embodiments or examples.

[0051] In the present invention, a flowchart is used to illustrate the operations performed by the system according to an embodiment of the present invention. It should be understood that the operations before or after do not necessarily need to be executed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0052] Next, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments of the present invention. It should be understood that the present invention is not limited by the exemplary embodiments described herein.

[0053] The state monitoring method of the traditional hydropower plant computer monitoring system mainly relies on manual experience and simple threshold judgment, and this method has obvious limitations. On the one hand, due to the lack of a comprehensive understanding and in-depth analysis of the system state, the traditional method is difficult to discover potential fault hazards; on the other hand, with the rapid increase in the amount of data, the ability of the traditional method to process big data is also stretched. Therefore, an optimized state monitoring system and method for the hydropower plant computer monitoring system are expected.

[0054] In the technical solution of the embodiment of the present invention, a state monitoring method for a hydropower plant computer monitoring system is proposed. Figure 1 It is a flowchart of the state monitoring method for the hydropower plant computer monitoring system according to an embodiment of the present invention. Figure 2 It is a schematic diagram of data flow of the state monitoring method for the hydropower plant computer monitoring system according to an embodiment of the present invention. As Figure 1 and Figure 2 shown, the state monitoring method for the hydropower plant computer monitoring system according to an embodiment of the present invention includes the steps of: S1, counting each state evaluation index of the hydropower plant computer monitoring system and the working condition data of each component of the hydropower plant computer monitoring system; S2, respectively performing semantic embedding encoding on each state evaluation index and each component working condition data to obtain a sequence of state evaluation index semantic embedding encoding vectors and a sequence of component working condition data semantic embedding encoding vectors; S3, respectively performing feature significant aggregation on the sequence of state evaluation index semantic embedding encoding vectors and the sequence of component working condition data semantic embedding encoding vectors to obtain a state evaluation index global significant semantic aggregation representation vector and a component working condition global significant semantic aggregation representation vector; S4, performing feature matching interaction on the state evaluation index global significant semantic aggregation representation vector and the component working condition global significant semantic aggregation representation vector to obtain a state evaluation index-component working condition optimal alignment fusion feature vector; S5, based on the state evaluation index-component working condition optimal alignment fusion feature vector, determining the state level label of the hydropower plant computer monitoring system.

[0055] Specifically, in S1, each status evaluation index of the computer monitoring system of the hydropower plant and the operating condition data of each component of the computer monitoring system of the hydropower plant are statistically analyzed. It should be understood that the computer monitoring system of the hydropower plant is a highly complex system, including multiple subsystems and components. By collecting and statistically analyzing each status evaluation index and component operating condition data, the actual operation of the system can be comprehensively and meticulously reflected, providing sufficient data support for subsequent analysis and evaluation. Among them, the status evaluation indexes include, but are not limited to, parameters and standards such as system availability, throughput, error rate, resource utilization rate, etc., while the component operating condition data involves the working status information of key components such as CPU usage rate, memory usage, temperature, voltage, etc.

[0056] Specifically, in S2, semantic embedding encoding is respectively performed on each of the status evaluation indexes and the component operating condition data to obtain a sequence of status evaluation index semantic embedding encoding vectors and a sequence of component operating condition data semantic embedding encoding vectors. Considering the diversity of the data forms of the status evaluation indexes and component operating condition data, including numerical data, text descriptions, etc. In order to convert these different types of unstructured or semi-structured data into a unified data form and extract their semantic information, in the technical solution of the present invention, semantic embedding encoding is further respectively performed on each of the status evaluation indexes and the component operating condition data, mapping them to a high-dimensional semantic feature space, converting them into high-dimensional dense numerical vector representations, and retaining their semantic features for subsequent data correlation analysis. In the embodiment of the present invention, the Word2Vec model is used to train and generate a status evaluation index embedding encoding matrix and an operating condition data embedding encoding matrix, and semantic embedding encoding is respectively performed on each of the status evaluation indexes and the component operating condition data with this to capture the inherent semantic meanings of each item of data, thereby generating a sequence of status evaluation index semantic embedding encoding vectors and a sequence of component operating condition data semantic embedding encoding vectors.

[0057] Specifically, in S3, feature significant aggregation is respectively performed on the sequence of status evaluation index semantic embedding encoding vectors and the sequence of component operating condition data semantic embedding encoding vectors to obtain a status evaluation index global significant semantic aggregation representation vector and a component operating condition global significant semantic aggregation representation vector. Since the status evaluation indexes and the component operating condition data are multi-dimensional and interrelated, a one-sided analysis alone cannot comprehensively understand the working status of the system. Therefore, in order to better capture the overall state of the system, feature aggregation processing is further respectively performed on the sequence of component operating condition data semantic embedding encoding vectors and the sequence of status evaluation index semantic embedding encoding vectors to aggregate these local and scattered feature information into global and high-level feature representations. In the technical solution of the present invention, in order to improve the effect of feature aggregation, a dominant aggregation method based on feature contribution degree is proposed.

[0058] Specifically, in the embodiments of the present invention, taking the sequence of the state evaluation index semantic embedding coding vectors as an example, feature significant aggregation is performed on the sequence of the state evaluation index semantic embedding coding vectors to obtain a state evaluation index global significant semantic aggregation representation vector, including: First, performing clustering analysis on the sequence of the state evaluation index semantic embedding coding vectors to obtain a state evaluation index self-supervised clustering representation vector; by using an unsupervised learning method to perform clustering analysis on the sequence to reveal the internal correlation distribution among local information and generate a state evaluation index self-supervised clustering representation vector, so as to represent the global semantic distribution characteristics of the state evaluation index; then, calculating the implicit clustering contribution factors of each state evaluation index semantic embedding coding vector in the sequence of the state evaluation index semantic embedding coding vectors relative to the state evaluation index self-supervised clustering representation vector to obtain a sequence of state evaluation index semantic feature implicit clustering contribution factors; by calculating the implicit clustering contribution factors of each state evaluation index semantic embedding coding vector relative to the state evaluation index self-supervised clustering representation vector, to reflect its contribution degree to the clustering features and quantify its role in the clustering process, so as to provide a basis for subsequent feature selection and weight assignment; furthermore, based on the sequence of the state evaluation index semantic feature implicit clustering contribution factors, performing feature aggregation on the sequence of the state evaluation index semantic embedding coding vectors to obtain the state evaluation index global significant semantic aggregation representation vector; here, the calculated implicit clustering contribution factors are integrated into a vector form, and a self-attention mechanism is used to perform explicit modeling on it, and through global self-correlation analysis to generate a more targeted weight vector, so as to realize the weighted aggregation of the features of each state evaluation index and obtain the state evaluation index global significant semantic aggregation representation vector.

[0059] Among them, the process of calculating the implicit clustering contribution factors of each state evaluation index semantic embedding coding vector in the sequence of the state evaluation index semantic embedding coding vectors relative to the state evaluation index self-supervised clustering representation vector to obtain a sequence of state evaluation index semantic feature implicit clustering contribution factors includes: calculating the dot-division vector between the state evaluation index semantic embedding coding vector and the state evaluation index self-supervised clustering representation vector, and calculating the base-2 logarithm of the absolute value of each eigenvalue in the dot-division vector to obtain an implicit clustering contribution weight vector; calculating the dot-product vector between the implicit clustering contribution weight vector and the state evaluation index semantic embedding coding vector, and calculating the exponential function value with e as the base and the sum of the eigenvalues of the dot-product vector as the exponent to obtain the state evaluation index semantic feature implicit clustering contribution factor.

[0060] More specifically, the process of feature aggregation on the sequence of the semantic embedding encoding vectors of the state evaluation indicators based on the sequence of the implicit clustering contribution factors of the semantic features of the state evaluation indicators to obtain the globally significant semantic aggregation representation vector of the state evaluation indicators includes: arranging the sequence of the implicit clustering contribution factors of the semantic features of the state evaluation indicators into a state evaluation indicator semantic feature clustering contribution field distribution vector; performing explicit modeling based on the self-attention mechanism on the state evaluation indicator semantic feature clustering contribution field distribution vector to obtain a state evaluation indicator semantic feature clustering contribution field modulation weight vector; using each eigenvalue in the state evaluation indicator semantic feature clustering contribution field modulation weight vector as a weight to calculate the position-wise weighted sum of the sequence of the semantic embedding encoding vectors of the state evaluation indicators to obtain the globally significant semantic aggregation representation vector of the state evaluation indicators. Among them, performing explicit modeling based on the self-attention mechanism on the state evaluation indicator semantic feature clustering contribution field distribution vector to obtain a state evaluation indicator semantic feature clustering contribution field modulation weight vector includes: respectively using a query embedding matrix, a key embedding matrix, and a value embedding matrix to perform embedding transformation on the state evaluation indicator semantic feature clustering contribution field distribution vector to obtain a query vector, a key vector, and a value vector; multiplying the query vector by the transpose vector of the key vector and then dividing by the square root of the scale of the key vector to obtain an attention score matrix; multiplying the attention score matrix by the value vector after passing through the softmax function to obtain the state evaluation indicator semantic feature clustering contribution field modulation weight vector.

[0061] Similarly, for the sequence of the semantic embedding encoding vectors of the component working condition data, the above method is also used for processing to generate a globally significant semantic aggregation representation vector of the component working condition. In this way, it can be ensured that the contribution degree of each item of data to the overall state evaluation of the system is fully considered during the aggregation process, so as to obtain a more representative global feature representation.

[0062] In summary, in the above embodiment, still taking the sequence of the semantic embedding encoding vectors of the state evaluation indicators as an example, the process of significantly aggregating the features of the sequence of the semantic embedding encoding vectors of the state evaluation indicators to obtain the globally significant semantic aggregation representation vector of the state evaluation indicators includes: processing the set of the semantic embedding encoding vectors of the state evaluation indicators with the following feature significant aggregation formula to obtain the globally significant semantic aggregation representation vector of the state evaluation indicators, where the feature significant aggregation formula is:

[0063] X = {x 1 , x 2 ,..., x k ,..., x n}

[0064]

[0065] v d = {d 1 ; d 2 ;...; d i ;...; d n}

[0066]

[0067] v q = W q v d

[0068] v k = W k v d

[0069] v v = W b v d

[0070]

[0071] Among them, X represents the set of semantic embedding coding vectors of the state evaluation indicators, x 1 , x 2 , x i , x k and x n are respectively the first, second, i-th, k-th, and n-th state evaluation indicator semantic embedding coding vectors in the set of semantic embedding coding vectors of the state evaluation indicators. The value of n is the number of semantic embedding coding vectors of the state evaluation indicators. x c represents the self-supervised clustering representation vector of the state evaluation indicator, represents the j-th eigenvalue of the i-th state evaluation indicator semantic embedding coding vector, represents the j-th eigenvalue of the self-supervised clustering representation vector of the state evaluation indicator. |·| represents the absolute value, log represents the logarithmic function with base 2, m represents the length of the semantic embedding coding vector of the state evaluation indicator, exp(·) represents the exponential function with base e, d 1 , d 2 , d i and d n represent the first, second, i-th, and n-th state evaluation indicator semantic feature implicit clustering contribution factors respectively. v d represents the state evaluation indicator semantic feature clustering contribution field distribution vector. W q , W k and W b represent the query embedding matrix, key embedding matrix, and value embedding matrix respectively. v q , v k and vv respectively represent the query vector, the key vector, and the value vector, (·) T represents the transpose of a vector, and Softmax represents the normalized exponential function represents matrix multiplication, v w represents the field modulation weight vector of the semantic feature clustering contribution of the state evaluation index represents the i-th eigenvalue in the field modulation weight vector of the semantic feature clustering contribution of the state evaluation index, x f represents the global significant semantic aggregation representation vector of the state evaluation index

[0072] Specifically, in step S4, a feature matching interaction is performed on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition to obtain an optimal alignment and fusion feature vector of the state evaluation index - component working condition. Since there is an interaction relationship between the state evaluation index of the monitoring system and the component working condition, for example, the abnormality of a certain state evaluation index may be caused by the combined action of multiple components. Therefore, in order to understand the working state of the system more deeply, the present invention further performs a semantic interaction matching analysis on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition to reveal the internal connection and mutual influence between the two. It is worth mentioning that, in order to improve the accuracy and efficiency of the semantic interaction matching, in the technical solution of the present invention, a fine-grained semantic feature interaction matching method is proposed, and by performing fine-grained decoupling and interaction matching calculations on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition, the optimal alignment and fusion between the state evaluation index and the state evaluation index are realized. In a specific example of the present invention, as Figure 3 shown, step S4 includes: S41, performing fine-grained feature decoupling on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition based on a predetermined scale to obtain a set of semantic feature vectors of the sub-components of the state evaluation index and a set of semantic feature vectors of the sub-components of the component working condition; S42, performing an optimal sub-component pairing multi-dimensional interaction on the set of semantic feature vectors of the sub-components of the state evaluation index and the set of semantic feature vectors of the sub-components of the component working condition to obtain a set of optimal sub-component pairing semantic fusion feature vectors of the state evaluation index - component working condition; S43, performing sub-component feature semantic fusion on the set of optimal sub-component pairing semantic fusion feature vectors of the state evaluation index - component working condition to obtain the optimal alignment and fusion feature vector of the state evaluation index - component working condition

[0073] Specifically, in step S41, based on a predetermined scale, fine-grained feature decoupling is performed on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition to obtain a set of state evaluation index sub-component semantic feature vectors and a set of component working condition sub-component semantic feature vectors. Through fine-grained feature decoupling, the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition are decomposed into multiple sub-components, forming a set of state evaluation index sub-component semantic feature vectors and a set of component working condition sub-component semantic feature vectors, so as to facilitate feature comparison at a lower level, thereby capturing more subtle semantic associations and improving the accuracy of feature matching and interaction.

[0074] Specifically, in step S42, optimal sub-component pairing and multi-dimensional interaction are performed on the set of state evaluation index sub-component semantic feature vectors and the set of component working condition sub-component semantic feature vectors to obtain a set of state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors. In the technical solution of the present invention, after feature decoupling, the hyperbolic space distance metric is used to calculate the hyperbolic space distance metric factor between any state evaluation index sub-component semantic feature vector and component working condition sub-component semantic feature vector, so as to capture the semantic similarity between the state evaluation index and the component working condition, thereby providing a guiding basis for subsequent optimal feature matching queries. Subsequently, based on the hyperbolic space distance metric factor, optimal feature sub-component pairing screening is performed on the set of state evaluation index sub-component semantic feature vectors and the set of component working condition sub-component semantic feature vectors to screen out a pair of feature vectors with the smallest distance in the hyperbolic space, so as to ensure that the most valuable feature information is retained in the subsequent interaction and fusion process and achieve optimal matching. Furthermore, through multi-dimensional interaction of the selected optimal pairings, deeper feature associations between the two are learned to enhance the expression ability of the features.

[0075] In an embodiment of the present invention, the process of performing optimal pairing multi-dimensional interaction on the set of semantic feature vectors of the state evaluation index sub-components and the set of semantic feature vectors of the component working condition sub-components to obtain a set of state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors includes: performing optimal feature sub-component pairing screening on the set of semantic feature vectors of the state evaluation index sub-components and the set of semantic feature vectors of the component working condition sub-components to obtain a set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings; inputting each optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairing in the set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings into a feature multi-dimensional interaction module to obtain the set of state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors.

[0076] Among them, the process of performing optimal feature sub-component pairing screening on the set of semantic feature vectors of the state evaluation index sub-components and the set of semantic feature vectors of the component working condition sub-components to obtain a set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings includes: for each semantic feature vector of the state evaluation index sub-components in the set of semantic feature vectors of the state evaluation index sub-components, calculating the hyperbolic space distance metric factor between it and each semantic feature vector of the component working condition sub-components in the set of semantic feature vectors of the component working condition sub-components, and selecting the semantic feature vector of the component working condition sub-components corresponding to the smallest hyperbolic space distance metric factor as the pairing object to obtain the set of optimal {semantic feature vectors of state evaluation index sub-components, semantic feature vectors of component working condition sub-components} pairings.

[0077] More specifically, the process of inputting each optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pair in the set of the optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairs into the feature multi-dimensional interaction module to obtain the set of state evaluation index-component working condition optimal sub-component pair semantic fusion feature vectors includes: respectively calculating the pointwise addition, pointwise subtraction, and pointwise multiplication between the state evaluation index sub-component semantic feature vector and the component working condition sub-component semantic feature vector in the optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pair to obtain a first state evaluation index-component working condition optimal pair interaction feature vector, a second state evaluation index-component working condition optimal pair interaction feature vector, and a third state evaluation index-component working condition optimal pair interaction feature vector; calculating the weighted sum by position between the first state evaluation index-component working condition optimal pair interaction feature vector, the second state evaluation index-component working condition optimal pair interaction feature vector, and the third state evaluation index-component working condition optimal pair interaction feature vector to obtain the state evaluation index-component working condition optimal sub-component pair semantic fusion feature vector.

[0078] In summary, in the above embodiments, feature matching interaction is performed on the state evaluation index global significant semantic aggregation representation vector and the component working condition global significant semantic aggregation representation vector to obtain a state evaluation index-component working condition optimal alignment fusion feature vector, including: processing the state evaluation index global significant semantic aggregation representation vector and the component working condition global significant semantic aggregation representation vector with the following feature fine-grained matching interaction formula to obtain the state evaluation index-component working condition optimal alignment fusion feature vector, where the feature fine-grained matching interaction formula is:

[0079] Decouple(v 1 )={v 11 ,v 12 ,...,v 1i ,...,v 1n}

[0080] Decouple(v 2 )={v 21 ,v 22 ,...,v 2j ,...,v 2n}

[0081]

[0082] v f =【v p1 ;v p2 ;...vpn

[0083] Among them, v 1 represents the global significant semantic aggregation representation vector of the state evaluation index, and v 2 represents the global significant semantic aggregation representation vector of the component working condition. Decouple(·) represents decoupling processing, and v 11 , v 12 , v 1i and v 1n respectively represent the semantic feature vectors of the first, second, i-th, and n-th sub-components of the state evaluation index. The value of n is the number of semantic feature vectors of the sub-components of the state evaluation index. v 21 , v 22 , v 2j , v 2n respectively represent the semantic feature vectors of the first, second, j-th, and n-th sub-components of the component working condition. arcosh(·) represents the inverse hyperbolic cosine function in the hyperbolic space, ‖·‖ represents the norm of the feature vector, and d P (v 1i , v 2j ) represents the hyperbolic space distance metric factor between the i-th semantic feature vector of the state evaluation index sub-component and the j-th semantic feature vector of the component working condition sub-component. represents the index for finding the semantic feature vector of the component working condition sub-component that minimizes the hyperbolic space distance metric factor. k represents the index of the semantic feature vector of the component working condition sub-component corresponding to the minimum hyperbolic space distance metric factor. α, β, and γ are different weight parameters, and ⊙ represents dot product. represents point addition. represents point subtraction, and v p1 , v p2 , v pi and v pn respectively represent the semantic fusion feature vectors of the optimal sub-component pairings of the first, second, i-th, and n-th state evaluation index-component working conditions. [·,·,·] represents the concatenation operation, and v f represents the optimal alignment fusion feature vector of the state evaluation index-component working condition.

[0084] ​Specifically, in step S43, sub-component feature semantic fusion is performed on the set of state evaluation index-component operating condition optimal sub-component paired semantic fusion feature vectors to obtain the state evaluation index-component operating condition optimal alignment fusion feature vector. In a specific example of the present invention, the set of state evaluation index-component operating condition optimal sub-component paired semantic fusion feature vectors is cascaded to obtain the state evaluation index-component operating condition optimal alignment fusion feature vector. By cascading, the interaction information of all sub-components is integrated, thereby generating a global state evaluation index-component operating condition optimal alignment fusion feature vector, so as to more comprehensively characterize the semantic association between the state evaluation index and the component operating condition.

[0085] It is worth mentioning that in other specific examples of the present invention, the state evaluation index- component operating condition optimal alignment fusion feature vector can also be obtained by other methods of feature matching interaction between the state evaluation index global significant semantic aggregation representation vector and the component operating condition global significant semantic aggregation representation vector. For example: input the state evaluation index global significant semantic aggregation representation vector and the component operating condition global significant semantic aggregation representation vector; perform standardization processing on the two vectors using L2 normalization; use cosine similarity to calculate the similarity between the two vectors; generate a matching matrix according to the similarity measurement result, and each element in the matrix represents the similarity between a certain feature in the state evaluation index vector and a certain feature in the component operating condition vector; design an attention module, which can receive the matching matrix as input and output an attention weight matrix; calculate the attention weight through a certain mechanism (such as the softmax function) to ensure that the weight value is between 0 and 1, and the sum of all weight values is 1; use the attention weight matrix to perform weighted fusion on the state evaluation index vector and the component operating condition vector; specifically, the attention weight matrix can be multiplied by the state evaluation index vector and the component operating condition vector respectively, and then the results are added to obtain the state evaluation index-component operating condition optimal alignment fusion feature vector.

[0086] Particularly, in step S5, based on the state evaluation index-component operating condition optimal alignment fusion feature vector, the state level label of the hydropower plant computer monitoring system is determined. In a specific example of the present invention, the state evaluation index-component operating condition optimal alignment fusion feature vector is input into the state evaluation module based on a classifier to obtain a state evaluation result, and the state evaluation result is used to represent the state level label of the hydropower plant computer monitoring system.

[0087] In summary, the state monitoring method of the computer monitoring system of a hydropower plant according to the embodiments of the present invention is elucidated. By adopting artificial intelligence technology based on deep learning, semantic embedding coding and global saliency fusion are performed on each state evaluation index and each component working condition data of the computer monitoring system of the hydropower plant to obtain a global semantic feature representation of the state evaluation index and the component working condition. Furthermore, through fine-grained semantic interaction matching between the two, a comprehensive understanding of the working state of the computer monitoring system of the hydropower plant is achieved, thereby intelligently evaluating its state level. In this way, potential faults in the computer monitoring system of the hydropower plant can be detected in a timely manner, so as to take corresponding measures in a timely manner to prevent the further expansion of the faults, thereby improving the reliability and safety of the computer monitoring system of the hydropower plant.

[0088] Furthermore, a state monitoring system for a computer monitoring system of a hydropower plant is also provided.

[0089] Figure 4 It is a block diagram of a state monitoring system for a computer monitoring system of a hydropower plant according to an embodiment of the present invention. As Figure 4 shown, the state monitoring system 300 of the computer monitoring system of the hydropower plant according to the embodiment of the present invention includes: a data statistics module 310 for statistically calculating each state evaluation index of the computer monitoring system of the hydropower plant and each component working condition data of the computer monitoring system of the hydropower plant; a semantic embedding coding module 320 for respectively performing semantic embedding coding on each state evaluation index and each component working condition data to obtain a sequence of state evaluation index semantic embedding coding vectors and a sequence of component working condition data semantic embedding coding vectors; a feature saliency aggregation module 330 for respectively performing feature saliency aggregation on the sequence of state evaluation index semantic embedding coding vectors and the sequence of component working condition data semantic embedding coding vectors to obtain a state evaluation index global significant semantic aggregation representation vector and a component working condition global significant semantic aggregation representation vector; a feature matching interaction module 340 for performing feature matching interaction on the state evaluation index global significant semantic aggregation representation vector and the component working condition global significant semantic aggregation representation vector to obtain a state evaluation index-component working condition optimal alignment fusion feature vector; and a state level label determination module 350 for determining a state level label of the computer monitoring system of the hydropower plant based on the state evaluation index-component working condition optimal alignment fusion feature vector.

[0090] As described above, the status monitoring system 300 of the hydropower plant computer monitoring system according to the embodiments of the present invention can be implemented in various wireless terminals, such as a server having a status monitoring algorithm for the hydropower plant computer monitoring system. In a possible implementation manner, the status monitoring system 300 of the hydropower plant computer monitoring system according to the embodiments of the present invention can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the status monitoring system 300 of the hydropower plant computer monitoring system can be a software module in the operating system of the wireless terminal, or can be an application program developed for the wireless terminal; of course, the status monitoring system 300 of the hydropower plant computer monitoring system can also be one of the numerous hardware modules of the wireless terminal.

[0091] Alternatively, in another example, the status monitoring system 300 of the hydropower plant computer monitoring system and the wireless terminal can also be separate devices, and the status monitoring system 300 of the hydropower plant computer monitoring system can be connected to the wireless terminal through a wired and / or wireless network and transmit interaction information in accordance with a predefined data format.

[0092] It can be understood that the above embodiments are merely exemplary embodiments adopted to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered within the protection scope of the present invention.

Claims

1. A method for monitoring the status of a hydropower plant computer monitoring system, characterized in that: include: Counting various status evaluation indicators of the hydropower plant computer monitoring system and the operating data of various components of the hydropower plant computer monitoring system; Separately performing semantic embedding coding on each of the state evaluation indicators and each of the component operating condition data to obtain a sequence of state evaluation indicator semantic embedding coding vectors and a sequence of component operating condition data semantic embedding coding vectors; Performing feature significant aggregation on the sequence of the state evaluation index semantic embedding coding vectors and the sequence of the component working condition data semantic embedding coding vectors to obtain a state evaluation index global significant semantic aggregation representation vector and a component working condition global significant semantic aggregation representation vector; Performing feature matching interaction on the global significant semantic aggregate representation vector of the state evaluation index and the global significant semantic aggregate representation vector of the component working condition to obtain the optimal alignment fusion feature vector of the state evaluation index-component working condition, including: performing fine-grained feature decoupling on the global significant semantic aggregate representation vector of the state evaluation index and the global significant semantic aggregate representation vector of the component working condition based on a predetermined scale to obtain a set of state evaluation index sub-component semantic feature vectors and a set of component working condition sub-component semantic feature vectors; performing optimal sub-component pairing multi-dimensional interaction on the set of state evaluation index sub-component semantic feature vectors and the set of component working condition sub-component semantic feature vectors to obtain a set of state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors; performing sub-component feature semantic fusion on the set of state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors to obtain the state evaluation index-component working condition optimal alignment fusion feature vector; Based on the state evaluation index-component operating condition optimal alignment fusion feature vector, the state level label of the hydropower plant computer monitoring system is determined.

2. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 1, characterized in that: The method of performing feature significant aggregation on the sequence of the state evaluation index semantic embedding coding vectors and the sequence of the component working condition data semantic embedding coding vectors to obtain a state evaluation index global significant semantic aggregation representation vector and a component working condition global significant semantic aggregation representation vector comprises: Performing cluster analysis on the sequence of the state evaluation indicator semantic embedding coding vectors to obtain a state evaluation indicator self-supervised cluster representation vector; Calculating the implicit clustering contribution factor of each state evaluation indicator semantic embedding coding vector in the sequence of the state evaluation indicator semantic embedding coding vector relative to the state evaluation indicator self-supervised clustering representation vector to obtain a sequence of implicit clustering contribution factors of state evaluation indicator semantic features; Based on the sequence of implicit clustering contribution factors of the semantic features of the state evaluation indicator, feature aggregation is performed on the sequence of semantic embedding coding vectors of the state evaluation indicator to obtain a global significant semantic aggregation representation vector of the state evaluation indicator.

3. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 2, characterized in that: The step of calculating the implicit clustering contribution factor of each state evaluation indicator semantic embedding coding vector in the sequence of the state evaluation indicator semantic embedding coding vector relative to the state evaluation indicator self-supervised clustering representation vector to obtain a sequence of implicit clustering contribution factors of state evaluation indicator semantic features, including: Calculating a dotted vector between the state evaluation indicator semantic embedding encoding vector and the state evaluation indicator self-supervised clustering representation vector, and calculating a base-two logarithm of the absolute value of each eigenvalue in the dotted vector to obtain an implicit clustering contribution weight vector; The dot product vector between the implicit clustering contribution weight vector and the state evaluation indicator semantic embedding coding vector is calculated, and the exponential function value with e as the base and the sum of the eigenvalues ​​of the dot product vector as the exponent is calculated to obtain the implicit clustering contribution factor of the semantic feature of the state evaluation indicator.

4. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 3, characterized in that: The method of performing feature aggregation on the sequence of the state evaluation indicator semantic embedding coding vectors based on the sequence of implicit clustering contribution factors of the state evaluation indicator semantic features to obtain the state evaluation indicator global significant semantic aggregation representation vector comprises: Arranging the sequence of implicit clustering contribution factors of the state evaluation indicator semantic features into a state evaluation indicator semantic feature clustering contribution field distribution vector; Performing explicit modeling on the distribution vector of the state evaluation indicator semantic feature cluster contribution field based on the self-attention mechanism to obtain a state evaluation indicator semantic feature cluster contribution field modulation weight vector; Taking each eigenvalue in the state evaluation indicator semantic feature clustering contribution field modulation weight vector as a weight, the position-weighted sum of the sequence of the state evaluation indicator semantic embedding coding vector is calculated to obtain the state evaluation indicator global significant semantic aggregation representation vector.

5. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 4, characterized in that: The optimal pairing multi-dimensional interaction of the set of state evaluation index sub-component semantic feature vectors and the set of component operating condition sub-component semantic feature vectors is performed to obtain a set of state evaluation index-component operating condition optimal sub-component pairing semantic fusion feature vectors, including: Performing optimal feature sub-component pairing screening on the set of semantic feature vectors of the state evaluation index sub-component and the set of semantic feature vectors of the component working condition sub-component to obtain an optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairing set; Each optimal {state evaluation index sub-component semantic feature vector, component operating condition sub-component semantic feature vector} pair in the set of optimal {state evaluation index sub-component semantic feature vector, component operating condition sub-component semantic feature vector} pairs is input into the feature multi-dimensional interaction module to obtain a set of semantic fusion feature vectors of the state evaluation index-component operating condition optimal sub-component pairs.

6. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 5, characterized in that: The optimal feature sub-component pairing screening of the set of the state evaluation index sub-component semantic feature vectors and the set of the component working condition sub-component semantic feature vectors to obtain the optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairing set includes: For each state evaluation indicator sub-component semantic feature vector in the set of state evaluation indicator sub-component semantic feature vectors, the hyperbolic space distance measurement factor between it and each component operating condition sub-component semantic feature vector in the set of component operating condition sub-component semantic feature vectors is calculated, and the component operating condition sub-component semantic feature vector corresponding to the smallest hyperbolic space distance measurement factor is selected as the pairing object to obtain the optimal {state evaluation indicator sub-component semantic feature vector, component operating condition sub-component semantic feature vector} pairing set.

7. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 6, characterized in that: The step of inputting each optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pair in the set of the optimal {state evaluation index sub-component semantic feature vector, component working condition sub-component semantic feature vector} pairs into a feature multi-dimensional interaction module to obtain a set of the state evaluation index-component working condition optimal sub-component paired semantic fusion feature vectors, including: Respectively calculating the position point addition, position point subtraction and position point multiplication between the state evaluation indicator subcomponent semantic feature vector and the component working condition subcomponent semantic feature vector in the optimal {state evaluation indicator subcomponent semantic feature vector, component working condition subcomponent semantic feature vector} pairing to obtain a first state evaluation indicator-component working condition optimal pairing interaction feature vector, a second state evaluation indicator-component working condition optimal pairing interaction feature vector and a third state evaluation indicator-component working condition optimal pairing interaction feature vector; Calculate the position-weighted sum of the first state evaluation index-component operating condition optimal pairing interaction feature vector, the second state evaluation index-component operating condition optimal pairing interaction feature vector and the third state evaluation index-component operating condition optimal pairing interaction feature vector to obtain the state evaluation index-component operating condition optimal sub-component pairing semantic fusion feature vector.

8. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 7, characterized in that: The step of performing sub-component feature semantic fusion on the set of the state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors to obtain the state evaluation index-component working condition optimal alignment fusion feature vector includes: The set of the state evaluation index-component working condition optimal sub-component pairing semantic fusion feature vectors is cascaded to obtain the state evaluation index-component working condition optimal alignment fusion feature vector.

9. The method for monitoring the status of a hydropower plant computer monitoring system according to claim 8, characterized in that: The determining of the state level label of the hydropower plant computer monitoring system based on the state evaluation index-component operating condition optimal alignment fusion feature vector comprises: The state evaluation index-component operating condition optimal alignment fusion feature vector is input into a classifier-based state evaluation module to obtain a state evaluation result, and the state evaluation result is used to represent the state level label of the hydropower plant computer monitoring system.

10. A state monitoring system for a hydropower plant computer monitoring system, characterized in that: include: A data statistics module, used to count various status evaluation indicators of the hydropower plant computer monitoring system, as well as the working condition data of various components of the hydropower plant computer monitoring system; A semantic embedding coding module, used to perform semantic embedding coding on each of the state evaluation indicators and each of the component operating condition data to obtain a sequence of state evaluation indicator semantic embedding coding vectors and a sequence of component operating condition data semantic embedding coding vectors; A feature significant aggregation module is used to perform feature significant aggregation on the sequence of the state evaluation index semantic embedding coding vectors and the sequence of the component working condition data semantic embedding coding vectors to obtain a global significant semantic aggregation representation vector of the state evaluation index and a global significant semantic aggregation representation vector of the component working condition; A feature matching interaction module is used to perform feature matching interaction on the global significant semantic aggregation representation vector of the state evaluation index and the global significant semantic aggregation representation vector of the component working condition to obtain the state evaluation index-component working condition optimal alignment fusion feature vector; The status level label determination module is used to determine the status level label of the hydropower plant computer monitoring system based on the status evaluation index-component operating condition optimal alignment fusion feature vector.

Images