Method, system, equipment and medium for fine division of vibration zones of hydropower units

Abstract

The present invention relates to the technical field of digital data processing and analysis, and in particular to a method, system, equipment and medium for finely dividing the vibration zones of hydropower units. The method comprises the following steps: based on multi-source vibration and pressure pulsation monitoring data, by screening key characteristic quantities characterizing the vibration characteristics of the unit, using a three-dimensional Gaussian mixture model to achieve operating zone segmentation, combining a density-based noise point spatial clustering algorithm for multi-scale data cleaning, synchronously eliminating abnormal data and optimizing data set integration; establishing a Gaussian process regression model through double optimization, fusing the cleaned multi-dimensional related data to achieve dynamic fine fitting of the vibration zone, and combining three-dimensional data visualization to accurately divide the vibration zone boundaries; dynamically analyzing the correlation of multi-source data, supporting automatic start and stop of the hydropower unit and precise load regulation to achieve rapid crossing of the vibration range, and combining a vibration avoidance economic dispatching strategy to improve the safety, stability and operating efficiency of the unit.

Description

Technical Field

[0001] The present invention relates to the technical field of digital data processing and analysis, and in particular to a method, system, equipment and medium for finely dividing vibration zones of a hydropower unit. Background Art

[0002] While hydropower units currently undergo stability testing within a planned head range during operation, verifying their operational stability under specific head conditions, detailed analysis of head and load remains insufficient, making it difficult to cover the full range of unit operating conditions. With the large-scale integration of renewable energy into the grid, the randomness and high-frequency switching characteristics of hydropower unit loads have significantly increased, posing particular challenges when participating in deep peak-shaving and frequency regulation. Existing stability test results fail to fully reflect the efficiency and stability characteristics of a unit under all operating conditions. Furthermore, opportunities for real-machine testing with refined vibration zone demarcation are limited, and the vibration zone ranges provided by manufacturers are overly broad. This leads to significant problems in actual operation: First, insufficient vibration zone demarcation accuracy causes units to frequently cross vibration zones during startup, shutdown, and load adjustments, exacerbating wear on mechanical components, shortening service life, and causing economic losses. Second, broad vibration zone demarcations hinder optimal load scheduling at power plants, leading to reduced unit efficiency and even forced operation within the vibration zone, threatening safety and economic efficiency. Third, the frequent participation of large units in peak and frequency regulation accelerates operating mode switching, causing dynamic evolution of the vibration zone, making the existing demarcations difficult to meet the requirements of economic optimization. Fourth, different units have different characteristics due to manufacturing and installation differences, but the existing vibration zone demarcation standard is single, weakening the effectiveness of unit coordinated scheduling. Fifth, due to environmental interference such as hydraulic and electromagnetic interference, the unit's multi-source monitoring data contains a high number of outliers, making it difficult to accurately represent the actual operating status. These problems collectively hinder the goal of safe, stable, efficient, and economical operation of hydropower units. Summary of the Invention

[0003] In view of the above problems in the prior art, the present invention is proposed.

[0004] Therefore, the technical problem to be solved by the present invention is: how to select characteristic quantities to characterize the vibration of the unit, eliminate abnormal values ​​in the historical data set of the unit operation, establish a vibration fitting model of the unit-related working conditions based on the normal set of the historical data of the unit operation, and realize the refined division of the vibration zone of the hydropower unit.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a method for fine-grained division of vibration zones of hydropower units, comprising: obtaining a data set of vibration characteristic quantities of the hydropower units, analyzing correlations, and selecting vibration characteristic quantities of the units; dividing the unit head-power-vibration historical operation data set into multiple operating condition zones using a three-dimensional Gaussian mixture model optimized by a maximum expectation algorithm; cleaning the operating condition data set using a density-based noise spatial clustering algorithm, removing abnormal data sets, merging normal data sets, and dividing the data set into dense and sparse operating condition zones of the units using a density-based noise spatial clustering algorithm; establishing a Gaussian process regression model for fitting the operating condition-vibration probability of the unit operating condition in dense and sparse zones using dual optimization of maximum likelihood estimation and cross-validation; drawing three- and two-dimensional heat maps of the operating condition-vibration of the unit operating condition in dense and sparse zones based on the established Gaussian process regression model for fitting the operating condition-vibration probability of the unit operating condition in dense and sparse zones based on the test set data, and finely dividing the vibration zones of the unit based on the two-dimensional heat maps of the operating condition-vibration of the unit operating condition in dense and sparse zones.

[0006] As a preferred solution of the method for finely dividing vibration zones of a hydropower unit described in the present invention, the characteristic quantity data set includes a historical operating condition data set of the hydropower unit and a characteristic quantity data set characterizing the vibration of the unit.

[0007] As a preferred embodiment of the method for finely dividing vibration zones of a hydropower unit according to the present invention, the method further comprises: selecting the vibration characteristic quantities of the unit including selecting the horizontal vibration of the water guide bearing of the hydropower unit, the horizontal vibration and the corresponding vertical vibration of the top cover, and the runner pressure as characteristic quantities; calculating the Pearson correlation coefficient of the unit power, water head, opening, the horizontal vibration of the water guide bearing, the horizontal vibration and the corresponding vertical vibration of the top cover, and the runner pressure; and obtaining the target characteristic quantity as a characterization indicator of the vibration characteristics of the unit.

[0008] As a preferred solution of the method for fine-grained division of vibration zones of a hydropower unit described in the present invention, the method further comprises dividing the unit head-power-vibration historical operation data set into multiple operating zones, including inputting the operating parameters and vibration characteristic quantities into a three-dimensional Gaussian mixture model, adopting an expectation maximization algorithm for iterative optimization, iteratively optimizing and estimating the mean vector, covariance matrix and weight parameters of the Gaussian distribution, fitting the three-dimensional spatial distribution of the data set, and dividing the unit head-power-vibration historical operation data set into multiple operating zones according to the actual distribution of the data set.

[0009] As a preferred solution of the method for fine-grained division of vibration zones of hydropower units described in the present invention, the density-based noise spatial clustering algorithm is adopted, which includes determining the neighborhood radius and minimum sample number of the density-based noise spatial clustering algorithm for cleaning data sets of each operating condition area, cleaning each operating condition area, and marking and removing abnormal points in each operating condition area; merging the normal data sets after cleaning the operating condition areas, and setting the adaptive neighborhood radius and minimum sample number of the density-based noise spatial clustering algorithm according to the operating characteristics of the unit, and dividing the normal data sets of the unit operation history into dense and sparse areas of the unit operation conditions.

[0010] As a preferred solution of the method for fine-grained division of vibration zones of a hydropower unit described in the present invention, the method adopts dual optimization of maximum likelihood estimation and cross-validation to establish a Gaussian process regression model for fitting the unit's operating conditions in dense and sparse areas and the vibration probability, including standardizing the data of the dense and sparse areas of the unit's operating conditions; dividing the training set and test set of the data of the dense and sparse areas of the unit's operating conditions; defining a cross-validation function in combination with the mean square error; using maximum likelihood estimation to optimize hyperparameters and combining cross-validation to evaluate performance; and using the optimal parameters obtained by the dual optimization to train the final Gaussian process regression model for fitting the unit's operating conditions in dense and sparse areas and the vibration probability.

[0011] As a preferred solution of the method for finely dividing the vibration zones of a hydropower unit described in the present invention, the finely dividing the vibration zones of the unit according to the two-dimensional thermal maps of the unit's operating conditions in dense and sparse areas includes inputting the test set data of the unit's operating conditions in dense areas and sparse areas into an established operating condition-vibration probability fitting Gaussian process regression model, performing cubic interpolation on the vibration output of the test set, drawing three- and two-dimensional thermal maps of the unit's operating conditions in dense areas and sparse areas according to the interpolation results, obtaining two-dimensional thermal maps of the unit's operating conditions in dense areas and sparse areas according to the three-dimensional thermal maps of the unit's operating conditions in dense areas and sparse areas, and finely dividing the vibration zones of the unit according to the two-dimensional thermal maps of the unit's operating conditions in dense areas and sparse areas.

[0012] Another object of the present invention is to provide a system for fine-grained division of vibration zones of hydropower units. By integrating a full-process collaborative architecture of a vibration feature quantity selection module, an abnormal data cleaning module, a unit vibration fitting model module and a vibration zone fine-grained division module, combined with a three-dimensional Gaussian mixture model multi-operating condition partitioning based on an expectation maximization algorithm and a density-based noise space clustering data cleaning technology, a dynamic screening strategy for the vibration characterization feature quantities of the hydropower unit is realized; a Gaussian process regression model driven by dual optimization of hyperparameters is adopted to improve the vibration probability prediction accuracy of the unit's dense and sparse operating condition areas; through the visual coupling analysis of calibrated cubic interpolation and three- and two-dimensional thermal maps, a multi-dimensional spatial mapping mechanism with adaptive vibration zone thresholds is established, which effectively solves the problems of high sensitivity of traditional methods to non-uniform operating condition data and fuzzy vibration boundary characterization, and provides a highly robust intelligent hierarchical control system for the safe operation of hydropower units.

[0013] In order to solve the above technical problems, the present invention provides the following technical solutions: a system for fine division of vibration zones of hydropower units, comprising: a vibration feature quantity selection module, an abnormal data cleaning module, a unit vibration fitting model module and a vibration zone fine division module;

[0014] The vibration characteristic quantity selection module obtains a set of vibration characteristic quantities of the hydropower unit under related working conditions, calculates the Pearson correlation coefficient of each characteristic quantity, and selects the vibration characteristic quantity of the unit;

[0015] The abnormal data cleaning module uses a three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm to divide the unit head-power-vibration historical operation data set into multiple operating conditions; uses a density-based noise point spatial clustering algorithm to clean the head-power-vibration three-dimensional data of each operating condition zone, and eliminates abnormal data sets in each operating condition zone; merges the normal data sets of each operating condition zone, and uses a density-based noise point spatial clustering algorithm to divide the normal data sets into dense and sparse unit operating condition zones;

[0016] The unit vibration fitting model module adopts dual optimization of maximum likelihood estimation and cross-validation to establish a Gaussian process regression model for unit operating conditions in dense and sparse areas - vibration probability fitting;

[0017] The vibration zone fine division module fits the Gaussian process regression model based on the established unit operating conditions-intensive and sparse zone operating conditions-vibration probability, draws the unit operating conditions-intensive and sparse zone operating conditions-vibration three- and two-dimensional thermal maps based on the test set data, and finely divides the unit vibration zone based on the unit operating conditions-intensive and sparse zone operating conditions-vibration two-dimensional thermal maps.

[0018] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method for finely dividing the vibration zones of a hydropower unit as described above are implemented.

[0019] A computer-readable storage medium stores a computer program thereon, wherein when the computer program is executed by a processor, the steps of the method for finely dividing the vibration zones of a hydropower unit as described above are implemented.

[0020] Beneficial effects of the present invention: The present invention proposes a method and system for fine-grained division of vibration zones of hydropower units, aiming to regularly and dynamically explore the correlation between multi-source vibration and pressure pulsation monitoring data of hydropower units, and at the same time perform multi-scale cleaning of multi-source vibration and pressure pulsation monitoring data, and based on the cleaned multi-source correlated vibration and pressure pulsation data, realize dynamic fine-grained fitting division of vibration zones of hydropower units under related working conditions, guide hydropower units to automatically start and stop, increase and reduce loads, accurately and quickly cross vibration zones, and economically optimize scheduling operations to avoid vibration, thereby improving the safety, stability and efficiency of the units, and providing technical support for efficient, safe and stable operation of giant hydropower units. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 A flow chart of a method for finely dividing vibration zones of a hydropower unit provided in one embodiment of the present invention.

[0023] Figure 2 A Pearson correlation coefficient diagram of various characteristic quantities of a method for finely dividing vibration zones of a hydropower unit provided by one embodiment of the present invention.

[0024] Figure 3 A three-dimensional spatial distribution diagram of a historical operation data set of a method for finely dividing vibration zones of a hydropower unit provided by one embodiment of the present invention.

[0025] Figure 4 A multi-operating zone diagram of a method for finely dividing vibration zones of a hydropower unit provided by one embodiment of the present invention.

[0026] Figure 5 This is a distribution diagram of abnormal points in a method for finely dividing vibration zones of a hydropower unit provided by one embodiment of the present invention.

[0027] Figure 6A distribution diagram of dense and sparse areas of unit operating conditions for a method for finely dividing vibration zones of a hydropower unit provided in one embodiment of the present invention.

[0028] Figure 7 A three-dimensional thermodynamic map of the vibration of a unit in an operating condition-intensive area, according to a method for finely dividing the vibration zones of a hydropower unit provided by an embodiment of the present invention.

[0029] Figure 8 A two-dimensional thermodynamic map of the vibration of a unit in an operating condition-intensive area, according to a method for finely dividing the vibration zones of a hydropower unit provided by an embodiment of the present invention.

[0030] Figure 9 A three-dimensional thermodynamic map of the vibration of a sparse zone of a hydropower unit operating condition is provided in accordance with an embodiment of the present invention.

[0031] Figure 10 A two-dimensional thermodynamic map of the vibration of a sparse zone of a hydropower unit operating condition is provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0032] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0033] Example 1, with reference to Figures 1-10 , is an embodiment of the present invention, which provides a method for finely dividing the vibration zones of a hydropower unit, comprising:

[0034] S1: Obtain a data set of vibration characteristics of hydropower units, analyze the correlation, and select the vibration characteristics of the units.

[0035] It should be noted that if Figure 1 As shown in step S1 in the figure, a historical operating condition dataset of the hydropower unit and a characteristic quantity dataset representing the vibration of the unit are obtained from the computer monitoring system of the hydropower station, and the correlation between the characteristic quantities of the historical operating condition dataset of the unit is analyzed to select the characteristic quantities that can effectively represent the vibration characteristics of the unit.

[0036] Furthermore, the water guide bearing level of the hydropower unit is selected Directional vibration, top cover level Direction and The vertical Z-direction vibration corresponding to the direction, the wheel pressure is used as the characteristic quantity; the computer group power, head, opening, water guide bearing level Directional vibration, top cover level Direction and The vertical direction Pearson correlation coefficient of directional vibration and runner pressure; the characteristic quantity with the smallest sum of the absolute values ​​of the differences between this indicator and other indicators is selected as the characteristic quantity to characterize the vibration characteristics of the unit.

[0037] Specifically, select the water guide bearing level of a hydropower unit Directional vibration , top cover level direction and The vertical direction Directional vibration , runner pressure as a characteristic quantity ; Get unit power , water head , opening , water guide bearing level Directional vibration, top cover level Direction and The vertical direction The Pearson correlation coefficient of each characteristic quantity is calculated based on the historical operating data of directional vibration and runner pressure, such as Figure 2 As shown in the figure, the characteristic quantity with the smallest sum of the absolute value of the difference between the correlation coefficients of each indicator and other indicators is selected as the characteristic quantity to characterize the vibration characteristics of the unit. The characteristic quantity obtained by calculation is: Directional vibration ; The three-dimensional spatial distribution of its historical operation data set is as follows Figure 3 shown.

[0038] S2: A three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm is used to divide the unit head-power-vibration historical operation data set into multiple operating conditions.

[0039] It should be noted that if Figure 1 As shown in step S2 in , a three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm is used to divide the unit head-power-vibration historical operation data set into multiple operating condition areas.

[0040] Furthermore, the operating parameters power, head and vibration characteristics are input into the three-dimensional Gaussian mixture model. At the same time, the expectation maximization algorithm is used for iterative optimization to estimate the mean vector, covariance matrix and weight parameters of each Gaussian distribution, fit the three-dimensional spatial true distribution of the data set, and then divide the unit head-power-vibration historical operation data set into multiple operating condition areas according to the true distribution of the data set.

[0041] Specifically, the operating parameters power, water head and vibration characteristics are input into the three-dimensional Gaussian mixture model; in three-dimensional space, the probability density function of a single Gaussian distribution is Expressed as:

[0042] ,

[0043] in, is the probability density function of a single Gaussian distribution, is a data point in three-dimensional space, is the mean vector of the Gaussian distribution, which is also a three-dimensional vector. is the covariance matrix of the Gaussian distribution, which is a The matrix, is the determinant of the covariance matrix, is the square of the Mahalanobis distance, used to measure the data points With the mean vector The distance between Expressed as a natural constant The exponential function with base , is the covariance matrix of the Gaussian distribution The inverse matrix of

[0044] The Gaussian mixture model assumes that the data is generated by a mixture of multiple Gaussian distributions, and its probability density function is Expressed as:

[0045] ,

[0046] in, is the number of Gaussian distributions, also known as the number of mixture components, For the The mixture weights of Gaussian components satisfy and , For the The probability density function of a Gaussian distribution, parameterized by the mean vector and the covariance matrix , 、 and Represent the set of mean vectors, covariance matrices, and mixing weights of all Gaussian components respectively;

[0047] The expectation maximization algorithm is used to iteratively optimize the mean vector, covariance matrix and weight parameters of each Gaussian distribution to fit the three-dimensional spatial distribution of the data set. The goal of the algorithm is to maximize the log-likelihood function of the observed data. Expressed as:

[0048] ,

[0049] in, represents the observation data, represents the latent variable, represents the model parameters, represents an entirety, expressed as a log-likelihood function;

[0050] ,

[0051] in, Indicates the The parameter estimates at iteration , represents the conditional expectation of the latent variable given the current parameter estimate and observed data, Expressed as an overall function maximization function, Indicates the number of iterations;

[0052] exist In the step, by maximizing function to update parameter estimates , expressed as:

[0053] ,

[0054] in, represents a mathematical function that returns the complex argument, Indicates the maximum value of the model parameter;

[0055] According to the actual distribution of the data set, the unit head-power-vibration historical operation data set is divided into multiple operating condition areas. Figure 4 shown.

[0056] S3: A density-based noise spatial clustering algorithm is used to clean the operating condition data set, remove abnormal data sets, merge normal data sets, and divide the data set into dense and sparse areas of unit operating conditions using a density-based noise spatial clustering algorithm.

[0057] It should be noted that if Figure 1 As shown in step S3 in , a density-based noise spatial clustering algorithm is used to clean the head-power-vibration three-dimensional data of each operating area, and the abnormal data sets of each operating area are eliminated; the normal data sets of each operating area are merged, and the algorithm is used to divide them into dense and sparse areas of unit operation conditions.

[0058] Furthermore, the neighborhood radius and minimum number of samples of the algorithm for cleaning the data set of each operating condition area are determined, each operating condition area is cleaned, and the abnormal points in each operating condition area are marked and removed; the normal data sets after cleaning of each operating condition area are merged, and according to the operating characteristics of the unit, the adaptive neighborhood radius and minimum number of samples of the algorithm are set, and the normal data sets of the unit operation history are divided into dense and sparse areas of the unit operation conditions.

[0059] Specifically, the neighborhood radius and minimum number of samples of the data set cleaning algorithm for each working area are determined, each working area is cleaned one by one, and the abnormal points in each working area are marked and removed, such as Figure 5 As shown, the algorithm includes: initializing clusters, determining cluster density thresholds and minimum points ; Read any point in the data set, according to and Determine whether the point is a core point. If it is not a core point but is located at a core point, Points within the neighborhood are recorded as boundary points, otherwise they are recorded as noise points; for the core point, all points whose density can be reached are recorded as a cluster. ; Read other unvisited core points in the data set, determine the point set with which its density can be reached, and obtain new clusters; repeat the above steps until all points in the data set are determined; the cluster density threshold and minimum number of points in each working condition area are shown in Table 1;

[0060] Table 1 Cluster density threshold and minimum number of points in each operating area

[0061] ,

[0062] The cleaned normal data sets of each operating condition area are merged, and according to the unit operation characteristics, the adaptive neighborhood radius and the minimum number of samples of the algorithm are set, as shown in Table 2. The unit operation history normal data set is divided into the unit operation condition dense area and sparse area, as shown in Table 2. Figure 6 As shown;

[0063] Table 2 Cluster density thresholds and minimum number of points in dense and sparse areas of unit operating conditions

[0064] ,

[0065] S4: Using maximum likelihood estimation and cross-validation dual optimization, a Gaussian process regression model for unit operating conditions in dense and sparse areas and vibration probability fitting is established.

[0066] It should be noted that if Figure 1 As shown in step S4 in , a Gaussian process regression model for fitting the unit's operating conditions in dense and sparse areas and vibration probability is established by using maximum likelihood estimation and cross-validation dual optimization.

[0067] Furthermore, the data of the dense and sparse areas of the unit operating conditions are standardized; the data of the dense and sparse areas of the unit operating conditions are divided into training sets and test sets; the cross-validation function is defined in combination with the mean square error; the hyperparameters are optimized using maximum likelihood estimation and the performance is evaluated in combination with cross-validation; the optimal parameters obtained by double optimization are used to train the final Gaussian process regression model for the dense and sparse areas of the unit operating conditions-vibration probability fitting.

[0068] Specifically, Gaussian process regression is a non-parametric Bayesian regression method that assumes that the observed data is generated by a Gaussian process, that is, the output function The mean is , the covariance is Gaussian process; the model is expressed as: ,in is the noise term, which is assumed to be Gaussian noise;

[0069] Training phase:

[0070] Given a training dataset ,in is the input variable, is the corresponding output variable;

[0071] Calculate the covariance matrix of the training dataset ,in The elements of Calculated;

[0072] Prediction stage:

[0073] For new input points , predict its corresponding output value ;

[0074] According to the properties of the Gaussian process, the predicted distribution is a Gaussian distribution with a mean of and variance Expressed as: ,

[0075] in is the covariance matrix between the new input point and the training dataset, is the covariance matrix of the training data set;

[0076] The specific optimization steps include: standardizing the data of the dense and sparse areas of the unit's operating conditions; randomly sampling the data of the dense and sparse areas of the unit's operating conditions and dividing them into training sets and test sets according to a 1:4 ratio; selecting the radial basis kernel function as the Gaussian process regression kernel function, randomly dividing the divided training set into 4 parts, training the Gaussian process regression model, and calculating the mean square error of the output of each training set, using grid search to automatically optimize the Gaussian process regression model hyperparameters, and cross-validating to evaluate the model's general performance; combining the automatically searched and optimized Gaussian process regression model hyperparameters with maximum likelihood estimation to optimize the hyperparameters, and training the final Gaussian process regression model for the unit's dense and sparse areas of operating conditions-vibration probability fitting.

[0077] S5: Based on the established Gaussian process regression model for the unit's dense operating conditions and sparse operating conditions-vibration probability, draw the three- and two-dimensional heat maps of the unit's dense operating conditions and sparse operating conditions-vibration based on the test set data, and finely divide the unit's vibration zones based on the two-dimensional heat maps of the unit's dense operating conditions and sparse operating conditions-vibration.

[0078] It should be noted that if Figure 1 As shown in step S5, according to the established unit operating condition intensive and sparse area working condition-vibration probability fitting Gaussian process regression model, based on the test set data, the unit operating condition intensive and sparse area working condition-vibration three-dimensional and two-dimensional heat maps are drawn, and the unit vibration zone is refined according to the unit operating condition intensive and sparse area working condition-vibration two-dimensional heat maps.

[0079] Furthermore, the test set data of the unit's operating condition dense area and sparse area are input into the established operating condition-vibration probability fitting Gaussian process regression model, and the vibration output of the test set is interpolated three times. According to the interpolation results, the three-dimensional and two-dimensional heat maps of the unit's operating condition dense area and sparse area operating condition-vibration are drawn. According to the three-dimensional heat maps of the unit's operating condition dense area and sparse area operating condition-vibration, the two-dimensional heat maps of the unit's operating condition dense area and sparse area operating condition-vibration are obtained. Finally, according to the two-dimensional heat maps of the unit's operating condition dense area and sparse area operating condition-vibration, the unit vibration zone is finely divided.

[0080] Specifically, the test data of the unit operating condition intensive area is input into the established working condition intensive area working condition-vibration probability fitting Gaussian process regression model, and the vibration output of the test set is interpolated three times. Based on the interpolation results, the three-dimensional and two-dimensional thermal maps of the unit operating condition intensive area working condition-vibration are drawn. The results are as follows: Figure 7 and Figure 8 As shown;

[0081] At the same time, the test data of the sparse area of ​​the unit operating condition is input into the established sparse area of ​​the operating condition-vibration probability fitting Gaussian process regression model, and the vibration output of the test set is interpolated three times. According to the interpolation results, the three-dimensional and two-dimensional thermal maps of the sparse area of ​​the unit operating condition-vibration are drawn. The results are as follows: Figure 9 and Figure 10 As shown;

[0082] According to the two-dimensional thermal diagram of unit operating condition-vibration in dense and sparse areas, the unit operating condition-vibration degree partition diagram is obtained, such as Figure 8 and Figure 10 As shown in the figure, the refined division result of the vibration zone of the hydropower unit can be obtained.

[0083] The above is a schematic diagram of a method for finely dividing vibration zones of a hydroelectric generator unit according to this embodiment. It should be noted that the technical solution of the system for finely dividing vibration zones of a hydroelectric generator unit and the technical solution of the above-mentioned method for finely dividing vibration zones of a hydroelectric generator unit are based on the same concept. For details not described in detail in the technical solution of the system for finely dividing vibration zones of a hydroelectric generator unit according to this embodiment, please refer to the description of the technical solution of the above-mentioned method for finely dividing vibration zones of a hydroelectric generator unit.

[0084] Example 2 is an embodiment of the present invention, which provides a system for finely dividing vibration zones of a hydropower unit, characterized by comprising: a vibration feature quantity selection module, an abnormal data cleaning module, a unit vibration fitting model module, and a vibration zone finely dividing module;

[0085] The vibration characteristic quantity selection module obtains a set of vibration characteristic quantities of the hydropower unit under related working conditions, calculates the Pearson correlation coefficient of each characteristic quantity, and selects the vibration characteristic quantity of the unit;

[0086] The abnormal data cleaning module uses a three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm to divide the unit head-power-vibration historical operation data set into multiple operating conditions; uses a density-based noise point spatial clustering algorithm to clean the head-power-vibration three-dimensional data of each operating condition zone, and eliminates abnormal data sets in each operating condition zone; merges the normal data sets of each operating condition zone, and uses a density-based noise point spatial clustering algorithm to divide the normal data sets into dense and sparse unit operating condition zones;

[0087] The unit vibration fitting model module adopts dual optimization of maximum likelihood estimation and cross-validation to establish a Gaussian process regression model for unit operating conditions in dense and sparse areas - vibration probability fitting;

[0088] The vibration zone fine division module fits the Gaussian process regression model based on the established unit operating conditions-intensive and sparse zone operating conditions-vibration probability, draws the unit operating conditions-intensive and sparse zone operating conditions-vibration three- and two-dimensional thermal maps based on the test set data, and finely divides the unit vibration zone based on the unit operating conditions-intensive and sparse zone operating conditions-vibration two-dimensional thermal maps.

[0089] This embodiment further provides a computing device applicable to a method for finely dividing vibration zones of a hydropower unit, including:

[0090] Memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement a method for fine-grained division of vibration zones of a hydropower unit as proposed in the above embodiment.

[0091] This embodiment further provides a storage medium storing a computer program, which, when executed by a processor, implements a method for finely dividing vibration zones of a hydropower unit as proposed in the above embodiment.

[0092] The storage medium proposed in this embodiment and the method for finely dividing the vibration zones of a hydropower unit proposed in the above embodiment belong to the same inventive concept. Technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0093] Example 3 is the third embodiment of the present invention, which differs from the first two embodiments in that:

[0094] If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0095] Logic and / or steps otherwise described herein, which may be considered, for example, as a sequenced list of executable instructions for implementing logical functions, may be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0096] It should be understood that various components of the present invention may be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof may be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0097] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for finely dividing the vibration zones of a hydropower unit, characterized by: include: Obtain a data set of vibration characteristics of hydropower units, analyze the correlation, and select the vibration characteristics of the units; A three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm is used to divide the unit head-power-vibration historical operation data set into multiple operating conditions. A density-based noise spatial clustering algorithm is used to clean the operating condition data set, remove abnormal data sets, merge normal data sets, and set the adaptive neighborhood radius and minimum sample number of the density-based noise spatial clustering algorithm according to the unit operation characteristics, dividing the unit operation history normal data set into dense and sparse unit operation condition areas. The density-based noise point spatial clustering algorithm includes initializing clusters, determining the cluster density threshold ε and the minimum number of points M; reading any point in the data set, determining the core point based on ε and M, and recording it as a boundary point if it is not a core point but is within the ε neighborhood of the core point; otherwise, it is recorded as a noise point; for the core point, the set of all points that are density-reachable is recorded as a cluster j; reading the remaining unvisited core points in the data set, determining the set of points that are density-reachable, and obtaining a new cluster; Using dual optimization of maximum likelihood estimation and cross-validation, a Gaussian process regression model for fitting vibration probability in dense and sparse operating conditions of the unit is established. According to the established Gaussian process regression model for fitting the unit's operating conditions-vibration probability of dense and sparse areas, three- and two-dimensional thermal maps of the unit's operating conditions-vibration of dense and sparse areas are drawn based on the test set data, and the unit vibration zone is finely divided according to the two-dimensional thermal map of the unit's operating conditions-vibration of dense and sparse areas.

2. The method for finely dividing the vibration zones of a hydropower unit according to claim 1, characterized in that: The characteristic quantity data set includes a historical operating condition data set of the hydropower unit and a characteristic quantity data set representing the vibration of the unit.

3. The method for finely dividing the vibration zones of a hydropower unit according to claim 2, characterized in that: The selecting of the unit vibration characteristic quantity includes selecting the horizontal vibration of the water guide bearing of the hydropower unit, the horizontal vibration and the corresponding vertical vibration of the top cover, and the runner pressure as characteristic quantities; calculating the Pearson correlation coefficient of the unit power, water head, opening, the horizontal vibration of the water guide bearing, the horizontal vibration and the corresponding vertical vibration of the top cover, and the runner pressure; and calculating the characteristic quantity with the minimum sum of the absolute values ​​of the differences between the Pearson correlation coefficients of each indicator and the remaining indicators as the characteristic quantity characterizing the vibration characteristics of the unit.

4. A method for finely dividing the vibration zones of a hydropower unit according to claim 3, characterized in that: The method of dividing the unit head-power-vibration historical operation data set into multiple operating condition zones includes inputting the operating condition parameters and vibration characteristic quantities into a three-dimensional Gaussian mixture model, using an expectation maximization algorithm, iteratively optimizing and estimating the mean vector, covariance matrix, and weight parameters of the Gaussian distribution, fitting the three-dimensional spatial distribution of the data set, and dividing the unit head-power-vibration historical operation data set into multiple operating condition zones according to the actual distribution of the data set.

5. The method for finely dividing the vibration zones of a hydropower unit according to claim 4, characterized in that: The density-based noise spatial clustering algorithm includes determining the neighborhood radius and minimum sample number of the density-based noise spatial clustering algorithm for the cleaned data set of each operating condition area, cleaning each operating condition area, and marking and removing abnormal points in each operating condition area; merging the normal data sets after the cleaning of the operating condition areas, and setting the adaptive neighborhood radius and minimum sample number of the density-based noise spatial clustering algorithm according to the unit operation characteristics, and dividing the normal data sets of the unit operation history into dense and sparse unit operation condition areas.

6. A method for finely dividing vibration zones of a hydroelectric generator set according to claim 5, characterized in that: The dual optimization of maximum likelihood estimation and cross-validation is used to establish a Gaussian process regression model for fitting the unit's operating conditions in dense and sparse areas to the vibration probability, including standardizing the data in the dense and sparse areas of the unit's operating conditions; dividing the data in the dense and sparse areas of the unit's operating conditions into training sets and test sets; defining a cross-validation function in combination with the mean square error; using maximum likelihood estimation to optimize hyperparameters and combining cross-validation to evaluate performance; and using the optimal parameters obtained from the dual optimization to train the final Gaussian process regression model for fitting the unit's operating conditions in dense and sparse areas to the vibration probability.

7. The method for finely dividing the vibration zones of a hydroelectric generator set according to claim 6, characterized in that: The method of finely dividing the vibration zones of the unit according to the two-dimensional heat maps of the unit's operating conditions in dense and sparse areas includes inputting the test set data of the unit's operating conditions in dense areas and sparse areas into an established operating condition-vibration probability fitting Gaussian process regression model, performing cubic interpolation on the vibration output of the test set, drawing three- and two-dimensional heat maps of the unit's operating conditions in dense areas and sparse areas according to the interpolation results, obtaining two-dimensional heat maps of the unit's operating conditions in dense areas and sparse areas according to the three-dimensional heat maps of the unit's operating conditions in dense areas and sparse areas, and finely dividing the vibration zones of the unit according to the two-dimensional heat maps of the unit's operating conditions in dense areas and sparse areas.

8. A system for finely dividing vibration zones of a hydroelectric generator set, applying a method for finely dividing vibration zones of a hydroelectric generator set as claimed in any one of claims 1 to 7, characterized in that: include: Vibration feature selection module, abnormal data cleaning module, unit vibration fitting model module and vibration zone refinement division module; The vibration characteristic quantity selection module obtains a set of vibration characteristic quantities of the hydropower unit under related working conditions, calculates the Pearson correlation coefficient of each characteristic quantity, and selects the vibration characteristic quantity of the unit; The abnormal data cleaning module uses a three-dimensional Gaussian mixture model optimized by the maximum expectation algorithm to divide the unit head-power-vibration historical operation data set into multiple operating conditions; uses a density-based noise point spatial clustering algorithm to clean the head-power-vibration three-dimensional data of each operating condition zone, and eliminates abnormal data sets in each operating condition zone; merges the normal data sets of each operating condition zone, and uses a density-based noise point spatial clustering algorithm to divide the normal data sets into dense and sparse unit operating condition zones; The unit vibration fitting model module adopts dual optimization of maximum likelihood estimation and cross-validation to establish a Gaussian process regression model for unit operating conditions in dense and sparse areas - vibration probability fitting; The vibration zone fine division module fits the Gaussian process regression model based on the established unit operating conditions-intensive and sparse zone operating conditions-vibration probability, draws the unit operating conditions-intensive and sparse zone operating conditions-vibration three- and two-dimensional thermal maps based on the test set data, and finely divides the unit vibration zone based on the unit operating conditions-intensive and sparse zone operating conditions-vibration two-dimensional thermal maps.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the processor implements the steps of a method for finely dividing vibration zones of a hydropower unit according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a method for finely dividing vibration zones of a hydropower unit according to any one of claims 1 to 7 are implemented.

Images