A method, system, equipment, and medium for intelligent monitoring of dust status of excitation fan filter screen in hydropower stations.

Abstract

This invention relates to the field of hydropower station equipment testing technology, specifically to a method, system, equipment, and medium for intelligent monitoring of dust status on excitation fan filters in hydropower stations. The method provided by this invention mainly includes: outputting a current status index based on the rate of change and a filter status detection model; acquiring image data of the current filter; constructing an image recognition model; identifying the image data using the image recognition model to obtain a coverage index of the current filter; outputting a replacement index of the current filter based on the status index and coverage index through a fusion model; setting a judgment threshold; and outputting a judgment result based on the judgment threshold and the replacement index. Through the above method, by comprehensively analyzing and judging multiple data sources, a comprehensive result is given regarding whether or not to replace the filter. This avoids the problem of not being able to dynamically adjust the replacement timing according to changes in the plant's air environment and the unit's operating status, enabling timely filter replacement.

Description

Technical Field

[0001] This invention relates to the field of hydropower station equipment testing technology, and more specifically, to a method, system, equipment, and medium for intelligent monitoring of dust status of excitation fan filter screen in hydropower stations. Background Technology

[0002] Generator excitation systems commonly employ forced air cooling to dissipate heat from the power cabinet. Excitation fans are equipped with filters to trap airborne dust, preventing it from entering the power cabinet and adhering to the surfaces of core components such as thyristors and heat sinks, thus affecting heat dissipation efficiency and equipment operational stability. Currently, the industry widely adopts a fixed-cycle replacement method for excitation fan filters, relying on the experience of maintenance personnel to set the replacement interval, lacking real-time monitoring methods for the actual dust accumulation on the filters.

[0003] The existing fixed-period replacement method has significant drawbacks: on the one hand, if excessive dust accumulates on the filter and it is not replaced in time, it will lead to increased ventilation resistance, reduced heat dissipation, and excessive temperature rise in the power cabinet, affecting the stable operation of the excitation system; on the other hand, if the filter is replaced prematurely with only a small amount of dust accumulation, it will result in waste of filter consumables and increase maintenance costs. It also fails to dynamically adjust the replacement timing based on changes in the plant's air environment and the unit's operating status, which does not meet the requirements of lean equipment management. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, equipment, and medium for intelligent monitoring of dust status of excitation fan filter screen in hydropower stations, in order to solve the problems in the prior art.

[0005] This invention is achieved through the following technical solution: A method for intelligent monitoring of dust status on the filter screen of an excitation fan in a hydropower station includes: Obtain the basic parameters of the current dust, and normalize the basic parameters of the dust to obtain the target parameters; The rate of change of several target parameters is obtained respectively, a filter state detection model is constructed, and the current state index is output based on the rate of change and the filter state detection model. Acquire the image data of the current filter, construct an image recognition model, and use the image recognition model to identify the image data to obtain the coverage index of the current filter; A fusion model is constructed, and the replacement index of the current filter is output through the fusion model based on the state index and the coverage index. A judgment threshold is set, and the judgment result is output according to the judgment threshold and the replacement index.

[0006] Preferably, the normalization of the basic parameters of the dust includes:

[0007] Obtaining the rate of change of several target parameters includes:

[0008] In the formula, For the normalized first Basic parameters, For the original number Basic parameters, For the first Basic reference values ​​for various basic parameters For the rate of change, For the normalized first The first basic parameter Data at any given time For the first The first basic parameter Data at any given time For the first time, For the first time.

[0009] Preferably, constructing the filter state detection model includes:

[0010] In the formula, For state index, and All are the first calculation weights, and their sum is 1. The rate of change of temperature, The change rate of dust concentration.

[0011] Preferably, the construction of the image recognition model includes: Input layer, used to acquire raw image data Convolutional layers are used to extract visual features from raw image data; The activation layer is used to filter out ineffective features and enhance the filter structure and the features of the covered dust. Pooling layers are used to compress feature map sizes while preserving dust features. Fully connected layers are used to flatten the two-dimensional feature maps obtained from multi-layer convolution and pooling into one-dimensional feature vectors, thus completing feature fusion. The output layer is used to output the dust identification features and the confidence level of the current dust identification features; Historical filter images with different levels of dirt, different lighting conditions, and different angles were collected. The dust in the different historical filter images was labeled, and the historical filter images were divided into training set, validation set, and test set. The image recognition model was trained based on the training set, validation set, and test set to obtain the optimal image recognition model.

[0012] Preferably, the step of obtaining the coverage index of the current filter by recognizing the image data based on the image recognition model includes: After the image data is converted to grayscale, denoised, and normalized, it is input into the image recognition model; The image recognition model outputs dust recognition features, performs semantic segmentation on the image data with dust recognition features, and obtains the number of pixels covered by the dust recognition features and the number of original pixels of the filter in the image data, respectively. The current coverage index is obtained based on the number of covered pixels and the original number of pixels.

[0013] Preferably, the current coverage index is obtained by combining the number of covered pixels and the original number of pixels:

[0014] In the formula, To cover the index, This represents the original number of pixels. For the first The number of pixels covered by each dust recognition feature. For the first Confidence level of each dust identification feature The number of dust identification features.

[0015] Preferably, the construction of the fusion model includes:

[0016] In the formula, To change the index, and The second weight is calculated, and its sum is 1.

[0017] Secondly, the present invention also provides an intelligent monitoring system for the dust status of excitation fan filters in hydropower stations, used to execute the aforementioned intelligent monitoring method for the dust status of excitation fan filters in hydropower stations, including: The data processing module is configured to acquire the basic parameters of the current dust, normalize the basic parameters of the dust to obtain target parameters; acquire the change rate of several target parameters respectively, construct a filter state detection model, output the current state index based on the change rate and the filter state detection model; acquire the image data of the current filter, construct an image recognition model, and identify the image data based on the image recognition model to obtain the coverage index of the current filter. The result output module is configured to construct a fusion model, output the current filter replacement index based on the state index and coverage index, set a judgment threshold, and output the judgment result based on the judgment threshold and the replacement index.

[0018] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the above-mentioned intelligent monitoring method for dust status of excitation fan filter screen in hydropower stations.

[0019] Fourthly, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station.

[0020] The technical solution of the present invention has at least the following advantages and beneficial effects: The method provided by this invention mainly includes: outputting a current state index based on the change rate and filter state detection model; acquiring image data of the current filter; constructing an image recognition model; identifying the image data using the image recognition model to obtain a coverage index of the current filter; outputting a replacement index of the current filter based on the state index and coverage index through a fusion model; setting a judgment threshold; and outputting a judgment result based on the judgment threshold and the replacement index. Through this method, by comprehensively analyzing and judging multiple data sources, a comprehensive result is given regarding whether or not to replace the filter. This avoids the problem of not being able to dynamically adjust the replacement timing according to changes in the plant's air environment and the unit's operating status, enabling timely filter replacement. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0024] The module division in this application is a logical division. In actual application, there may be other division methods. For example, multiple modules may be combined into or integrated into another system, or some features may be ignored or not executed.

[0025] Furthermore, the connection, coupling, or communication in this application can be a direct connection, coupling, or communication between related objects, or an indirect connection, coupling, or communication through other devices. Moreover, the connection, coupling, or communication between objects can be electrical or other similar forms, and this application does not impose any limitations on these. Independently described modules or sub-modules may or may not be physically separated; they may be implemented in software or hardware, and some modules or sub-modules may be implemented in software, with the processor calling the software to implement the function of these modules or sub-modules, while other modules or sub-modules may be implemented in hardware, such as through hardware circuits. Furthermore, some or all of the modules can be selected to achieve the purpose of this application's solution according to actual needs.

[0026] Please refer to Figures 1-2 The present invention provides an intelligent monitoring method for dust status of excitation fan filter screen in hydropower stations, comprising: S1 obtains the current basic parameters of the dust, and normalizes the basic parameters of the dust to obtain the target parameters; In this embodiment, the basic parameters mainly include the power cabinet temperature parameter and dust concentration parameter. The dust concentration and temperature signals collected by the sensor are converted into code values ​​(corresponding to the actual physical quantities), and upper and lower limit thresholds are set to filter abnormal data to ensure data validity.

[0027] Redundant data processing: The average value of the data from the three dust sensors is taken as the valid data. When a single or dual channel fails, the data of the faulty channel is automatically removed. When all three channels fail, the output value is zero.

[0028] S2: Obtain the rate of change of several target parameters respectively, construct a filter state detection model, and output the current state index based on the rate of change and the filter state detection model; S3: Obtain the image data of the current filter, construct an image recognition model, and use the image recognition model to identify the image data to obtain the coverage index of the current filter; S4: Construct a fusion model, output the current filter replacement index based on the state index and coverage index, set a judgment threshold, and output the judgment result based on the judgment threshold and the replacement index.

[0029] The method provided by this invention mainly includes: outputting a current state index based on the change rate and filter state detection model; acquiring image data of the current filter; constructing an image recognition model; identifying the image data using the image recognition model to obtain a coverage index of the current filter; outputting a replacement index of the current filter based on the state index and coverage index through a fusion model; setting a judgment threshold; and outputting a judgment result based on the judgment threshold and the replacement index. Through this method, by comprehensively analyzing and judging multiple data sources, a comprehensive result is given regarding whether or not to replace the filter. This avoids the problem of not being able to dynamically adjust the replacement timing according to changes in the plant's air environment and the unit's operating status, enabling timely filter replacement. In one exemplary embodiment of the present invention, the normalization of the basic parameters of the dust includes:

[0030] Obtaining the rate of change of several target parameters includes:

[0031] In the formula, For the normalized first Basic parameters, For the original number Basic parameters, For the first Basic reference values ​​for various basic parameters For the rate of change, For the normalized first The first basic parameter Data at any given time For the first The first basic parameter Data at any given time For the first time, For the first time.

[0032] In this embodiment, the rate of change of temperature and concentration are mainly considered. If the temperature rises rapidly, it indicates that the blockage is more serious. If the temperature drops, it indicates that the blockage is not serious. At the same time, the rapid increase in concentration also indicates that the dust is accumulating rapidly.

[0033] Constructing the filter state detection model includes:

[0034] In the formula, For state index, and All are the first calculation weights, and their sum is 1. The rate of change of temperature, The change rate of dust concentration.

[0035] In one exemplary embodiment of the present invention, the construction of the image recognition model includes: Input layer, used to acquire raw image data

[0036] In the formula, To output the feature map in coordinates The convolution result value at that point, For the feature map of the previous layer in coordinates The pixel grayscale value at that location, The kernel size is [size]. The kernel size is the convolution kernel size. and The coordinates are within the convolution kernel.

[0037] Convolutional layers are used to extract visual features from raw image data; The activation layer is used to filter out ineffective features and enhance the filter structure and the features of the covered dust. Pooling layers are used to compress feature map sizes while preserving dust features. Fully connected layers are used to flatten the two-dimensional feature maps obtained from multi-layer convolution and pooling into one-dimensional feature vectors, thus completing feature fusion.

[0038] In the formula, This is the combined feature vector output by the fully connected layer. This is the weight matrix. One-dimensional feature vector The output layer is used to output the dust identification features and the confidence level of the current dust identification features; Historical filter images with different levels of dirt, different lighting conditions, and different angles were collected. The dust in the different historical filter images was labeled, and the historical filter images were divided into training set, validation set, and test set. The image recognition model was trained based on the training set, validation set, and test set to obtain the optimal image recognition model.

[0039] In one exemplary embodiment of the present invention, the step of obtaining the coverage index of the current filter by recognizing the image data based on an image recognition model includes: After the image data is converted to grayscale, denoised, and normalized, it is input into the image recognition model; The image recognition model outputs dust recognition features, performs semantic segmentation on the image data with dust recognition features, and obtains the number of pixels covered by the dust recognition features and the number of original pixels of the filter in the image data, respectively. The current coverage index is obtained based on the number of covered pixels and the original number of pixels.

[0040] Preferably, the current coverage index is obtained by combining the number of covered pixels and the original number of pixels:

[0041] In the formula, To cover the index, This represents the original number of pixels. For the first The number of pixels covered by each dust recognition feature. For the first Confidence level of each dust identification feature The number of dust identification features.

[0042] In this embodiment, the impact of confidence level on the entire dust identification feature is taken into account to improve the accuracy of the final calculation result as much as possible.

[0043] In one exemplary embodiment of the present invention, the construction of the fusion model includes:

[0044] In the formula, To change the index, and The second weight is calculated, and its sum is 1.

[0045] The higher the replacement index, the more severe the filter blockage, and the more necessary it is to be replaced. The judgment threshold can be set according to the standards of the specific working environment. That is, under the current standard, when a filter needs to be replaced, the current temperature data, dust concentration data, and image data are collected, and the current replacement index is calculated through the fusion model as the judgment threshold.

[0046] Secondly, the present invention also provides an intelligent monitoring system for the dust status of excitation fan filters in hydropower stations, used to execute the aforementioned intelligent monitoring method for the dust status of excitation fan filters in hydropower stations, including: The data processing module is configured to acquire the basic parameters of the current dust, normalize the basic parameters of the dust to obtain target parameters; acquire the change rate of several target parameters respectively, construct a filter state detection model, output the current state index based on the change rate and the filter state detection model; acquire the image data of the current filter, construct an image recognition model, and identify the image data based on the image recognition model to obtain the coverage index of the current filter. The result output module is configured to construct a fusion model, output the current filter replacement index based on the state index and coverage index, set a judgment threshold, and output the judgment result based on the judgment threshold and the replacement index.

[0047] The hardware system is integrated into the control box and is wall-mounted. Core components include: PLC controller: Schneider M340 series PLC is selected, configured with one 16-channel digital input module, one 16-channel digital output module and two 8-channel analog input modules, responsible for data acquisition, logic operation and control signal output.

[0048] Human-computer interaction module: Equipped with a touch screen for real-time data display, trend chart display, alarm prompts, parameter settings, and manual operation (forced start, forced stop, and reset).

[0049] Dust monitoring module: Each unit is equipped with 3 air quality transmitters (laser scattering principle), with a measurement range of 0 to 1000 μg / m³, and outputs 4-20mA analog signal to achieve redundant dust concentration acquisition.

[0050] Temperature monitoring module: Equipped with an infrared thermometer (working band 8~14μm, measurement range 0~500℃) to collect the temperature of the power cabinet air outlet, supplemented by an ambient temperature sensor, and outputs a 4-20mA analog signal.

[0051] Auxiliary components include relays, power modules, indicator lights, and related hardware circuits. The power module provides a stable 24V DC power supply to each component, and the relays output replacement indicator signals.

[0052] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0053] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. This computer software product, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0054] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent monitoring of dust status in the filter screen of an excitation fan in a hydropower station, characterized in that, include: Obtain the basic parameters of the current dust, and normalize the basic parameters of the dust to obtain the target parameters; The rate of change of several target parameters is obtained respectively, a filter state detection model is constructed, and the current state index is output based on the rate of change and the filter state detection model. Acquire the image data of the current filter, construct an image recognition model, and use the image recognition model to identify the image data to obtain the coverage index of the current filter; A fusion model is constructed, and the replacement index of the current filter is output through the fusion model based on the state index and the coverage index. A judgment threshold is set, and the judgment result is output according to the judgment threshold and the replacement index.

2. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 1, characterized in that, The basic parameters of the dust were normalized, including: Obtaining the rate of change of several target parameters includes: In the formula, For the normalized first Basic parameters, For the original number Basic parameters, For the first Basic reference values ​​for various basic parameters For the rate of change, For the normalized first The first basic parameter Data at any given time For the first The first basic parameter Data at any given time For the first time, For the first time.

3. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 2, characterized in that, Constructing the filter state detection model includes: In the formula, For state index, and All are the first calculated weights, and their sum is 1. The rate of change of temperature, The change rate of dust concentration.

4. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 3, characterized in that, The image recognition model is constructed by: Input layer, used to acquire raw image data Convolutional layers are used to extract visual features from raw image data; The activation layer is used to filter out ineffective features and enhance the filter structure and the features of the covered dust. Pooling layers are used to compress feature map sizes while preserving dust features. Fully connected layers are used to flatten the two-dimensional feature maps obtained from multi-layer convolution and pooling into one-dimensional feature vectors, thus completing feature fusion. The output layer is used to output the dust identification features and the confidence level of the current dust identification features; Historical filter images with different levels of dirt, different lighting conditions, and different angles were collected. The dust in the different historical filter images was labeled, and the historical filter images were divided into training set, validation set, and test set. The image recognition model was trained based on the training set, validation set, and test set to obtain the optimal image recognition model.

5. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 4, characterized in that, The process of identifying the image data based on an image recognition model to obtain the coverage index of the current filter includes: After the image data is converted to grayscale, denoised, and normalized, it is input into the image recognition model; The image recognition model outputs dust recognition features, performs semantic segmentation on the image data with dust recognition features, and obtains the number of pixels covered by the dust recognition features and the number of original pixels of the filter in the image data, respectively. The current coverage index is obtained based on the number of covered pixels and the original number of pixels.

6. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 5, characterized in that, The current coverage index is obtained by combining the number of covered pixels and the number of original pixels, including: In the formula, To cover the index, This represents the original number of pixels. For the first The number of pixels covered by each dust recognition feature. For the first Confidence level of each dust identification feature The number of dust identification features.

7. The intelligent monitoring method for dust status of excitation fan filter screen in a hydropower station according to claim 6, characterized in that, The construction of the fusion model includes: In the formula, To change the index, and The second weight is calculated, and its sum is 1.

8. An intelligent monitoring system for dust status of excitation fan filter screen in a hydropower station, characterized in that, A method for intelligent monitoring of dust status of excitation fan filter screen in a hydropower station, as described in any one of claims 1-7, includes: The data processing module is configured to acquire the basic parameters of the current dust, normalize the basic parameters of the dust to obtain target parameters; acquire the change rate of several target parameters respectively, construct a filter state detection model, output the current state index based on the change rate and the filter state detection model; acquire the image data of the current filter, construct an image recognition model, and identify the image data based on the image recognition model to obtain the coverage index of the current filter. The result output module is configured to construct a fusion model, output the current filter replacement index based on the state index and coverage index, set a judgment threshold, and output the judgment result based on the judgment threshold and the replacement index.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the intelligent monitoring method for dust status of the excitation fan filter screen in a hydropower station as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the intelligent monitoring method for dust status of excitation fan filter screen in hydropower stations as described in any one of claims 1-7.

Images