Hydropower station data processing method and device, computer equipment, readable storage medium and program product

Abstract

The application relates to a hydropower station data processing method and device, computer equipment, a readable storage medium and a program product. The method is applied to an edge controller and comprises the following steps: obtaining target data through a container deployed in the edge controller, wherein the target data comprises target voiceprint data, target visual data and target vibration data; analyzing the target voiceprint data, the target visual data and the target vibration data through the container deployed in the edge controller to obtain corresponding analysis results; in the case that a main programmable logic controller is abnormal, generating corresponding control instructions for the analysis results through a backup programmable logic controller, and executing corresponding control strategies based on the control instructions; and the backup programmable logic controller and the main programmable logic controller store the same data. The method can effectively avoid the interruption of the control process caused by the failure of the main programmable logic controller and reduce the equipment downtime.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pumped storage power stations, and in particular to a hydropower station data processing method and device, computer equipment, a computer readable storage medium and a computer program product. BACKGROUND

[0002] With the development of the power system, pumped storage power stations, as an important part of the power system, bear the key task of regulating power grid load and improving energy utilization efficiency. By using excess power to pump water to high reservoirs during low peak periods and releasing water flow for power generation during peak periods, they play an irreplaceable role in balancing power supply and demand and enhancing power grid stability.

[0003] In traditional technology, programmable logic controllers (PLC) are deployed in hydropower station applications to control control processes such as maintenance.

[0004] However, the current hydropower station data processing method may cause the entire control process to be interrupted once a fault occurs, seriously affecting the operation of the power station. SUMMARY

[0005] Therefore, it is necessary to provide a hydropower station data processing method, device, computer equipment, computer readable storage medium and computer program product that can process data without interrupting the process in view of the above technical problems.

[0006] In a first aspect, the present application provides a hydropower station data processing method applied in an edge controller, the method comprising:

[0007] The target data includes target voiceprint data, target visual data and target vibration data.

[0008] The target voiceprint data, the target visual data and the target vibration data are analyzed by the container deployed in the edge controller to obtain corresponding analysis results.

[0009] In the case of an abnormal main programmable logic controller, the backup programmable logic controller generates corresponding control instructions for each analysis result and executes a corresponding control strategy based on the control instructions; the backup programmable logic controller and the main programmable logic controller store the same data.

[0010] In one of the embodiments, the container includes a voiceprint container, a visual container, and a vibration container; the voiceprint container, the visual container, and the vibration container are independent of each other; the analysis result includes an identification result, an extraction result, and a processing result; the container deployed in the edge controller analyzes the target voiceprint data, the target visual data, and the target vibration data to obtain corresponding analysis results, including:

[0011] The identification model deployed in the voiceprint container processes the target voiceprint data to obtain an identification result of the identification model, and the identification result represents a health state of the target voiceprint data; the identification model is a neural network based on an attention mechanism;

[0012] The extraction model deployed in the visual container processes the target visual data to obtain an extraction result of the extraction model, and the extraction result represents dashboard data;

[0013] The vibration monitoring module deployed in the vibration container processes the target vibration data to obtain a processing result, and the processing result represents a health state of the target vibration data.

[0014] In one of the embodiments, the container deployed in the edge controller obtains target data, including:

[0015] The container deployed in the edge controller obtains initial data, and the initial data includes initial voiceprint data, initial visual data, and initial vibration data; the initial voiceprint data is collected based on a voiceprint sensor, the initial visual data is collected based on an image collection device, and the initial vibration data is collected based on a vibration sensor;

[0016] The container deployed in the edge controller performs noise reduction, echo cancellation, and enhancement preprocessing on the initial voiceprint data to obtain the target voiceprint data;

[0017] The container deployed in the edge controller performs enhancement processing on the initial visual data to obtain the target visual data;

[0018] The container deployed in the edge controller performs biasing, amplification, filtering, and integral noise reduction preprocessing on the initial vibration data to obtain preprocessed initial vibration data; and the preprocessed initial vibration data is converted into standard data to obtain the target vibration data.

[0019] In one of the embodiments, the target swing data includes a generator vibration sensor signal, a swing sensor signal and a key phase sensor signal; the target swing data is processed by a swing monitoring module deployed in the swing container to obtain a processing result, including:

[0020] The generator vibration sensor signal, the swing sensor signal and the key phase sensor signal are calculated and analyzed by the swing monitoring module deployed in the swing container to obtain a processing result.

[0021] In one of the embodiments, the target visual data is processed by the extraction model deployed in the visual container to obtain an extraction result of the extraction model, including:

[0022] The target visual data is cropped and background-removed by a target detection algorithm deployed in the visual container to obtain a dashboard area;

[0023] The dashboard area is binarized and contour-finding by a contour processing algorithm deployed in the visual container to obtain a key contour; the key contour includes a scale line contour, a pointer contour and an interference point contour;

[0024] Based on the key contour, a center of the dashboard area and a pointer straight line are obtained;

[0025] Based on the center of the dashboard area, the pointer straight line and a zero scale direction, the extraction result is obtained.

[0026] In one of the embodiments, based on the key contour, the center of the dashboard area and the pointer straight line are obtained, including:

[0027] The scale line contour is extracted from the key contour by a first preset condition; the first preset condition includes at least one of a distance condition, an aspect ratio condition and an area condition;

[0028] The center of the dashboard area is obtained according to the intersection of the scale line contour;

[0029] The key contour is filtered by a second preset condition to obtain a pointer contour, and a pointer straight line is obtained based on the pointer contour; the second preset condition includes excluding the contour other than the pointer contour.

[0030] In a second aspect, the application further provides a hydropower station data processing device applied to an edge controller, the device including:

[0031] An acquisition module is configured to acquire target data by a container deployed in the edge controller, the target data including target voiceprint data, target visual data and target swing data;

[0032] an analysis module configured to analyze the target voiceprint data, the target visual data, and the target swing data by a container deployed in the edge controller to obtain corresponding analysis results;

[0033] a generation execution module configured to generate corresponding control instructions for each of the analysis results by a backup programmable logic controller in the case of an exception of a main programmable logic controller, and execute a corresponding control strategy based on the control instructions; the backup programmable logic controller and the main programmable logic controller store the same data.

[0034] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described above when executing the computer program.

[0035] In a fourth aspect, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program implements the steps of the method described above when executed by a processor.

[0036] In a fifth aspect, the present application further provides a computer program product comprising a computer program, and the computer program implements the steps of the method described above when executed by a processor.

[0037] The above water power station data processing method and device, computer device, computer readable storage medium, and computer program product, first, the target data is obtained by the container deployed in the edge controller, and the target data comprises target voiceprint data, target visual data, and target swing data; second, the target voiceprint data, the target visual data, and the target swing data are analyzed by the container deployed in the edge controller to obtain corresponding analysis results; and finally, the backup programmable logic controller and the main programmable logic controller store the same data, and in the case of an exception of the main programmable logic controller, the backup programmable logic controller generates corresponding control instructions for each of the analysis results, and executes a corresponding control strategy based on the control instructions. The backup PLC monitors the state of the main PLC in real time, ensures that the control process is uninterrupted when the main PLC fails, and maintains normal operation of the system. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0039] Figure 1 An application environment diagram of the hydropower station data processing method in an embodiment;

[0040] Figure 2 A flowchart diagram of the hydropower station data processing method in an embodiment;

[0041] Figure 3 A flowchart diagram of a step of analyzing target data by a container to obtain a corresponding analysis result in an embodiment;

[0042] Figure 4 A structural diagram of a recognition model in an embodiment;

[0043] Figure 5 A flowchart diagram of a step of obtaining target data by a container in an embodiment;

[0044] Figure 6 A flowchart diagram of a step of obtaining an extraction result by an extraction model in an embodiment;

[0045] Figure 7 A flowchart diagram of a step of obtaining a center of a dashboard area and a straight line of a pointer in an embodiment;

[0046] Figure 8 A structural block diagram of a hydropower station data processing apparatus in an embodiment;

[0047] Figure 9 An internal structural diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0048] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0049] The hydropower station data processing method provided by the embodiments of the present application can be applied to, for example, Figure 1The sensor communicates with the edge controller through the network, the edge controller communicates with the host computer or remote monitoring device through the network, and the sensor communicates with the host computer or remote monitoring device through the network. The target data includes target voice data, target visual data and target vibration data. The target voice data, target visual data and target vibration data are analyzed by the container deployed in the edge controller to obtain corresponding analysis results. In the case of abnormality of the main programmable logic controller, the backup programmable logic controller generates corresponding control instructions according to the analysis results, and executes corresponding control strategies based on the control instructions. The backup programmable logic controller and the main programmable logic controller store the same data. The sensor can be but is not limited to a vibration sensor, a voiceprint sensor and an image acquisition device. The image acquisition device can be a camera.

[0050] In one exemplary embodiment, as Figure 2 shown, a hydropower station data processing method is provided, which is applied to Figure 1 the edge controller as an example, including the following steps S202 to S206. Wherein:

[0051] Step S202, obtaining target data by a container deployed in the edge controller.

[0052] The target data includes target voice data, target visual data and target vibration data. The target data can be data collected by the sensor, or data collected by the sensor and preprocessed. The target voice data can be data collected by the voiceprint sensor, or preprocessed voiceprint data collected by the voiceprint sensor, or sound spectrum features emitted by the generator set during operation, reflecting fluid dynamics, mechanical friction and electrical discharge states. The target visual data can be data collected by the image acquisition device, or preprocessed visual data collected by the image acquisition device, which is also the instrument panel data, reflecting real-time monitoring data of the generator set operating parameters. The target vibration data can be data collected by the vibration sensor, or preprocessed vibration data collected by the vibration sensor, which also represents the dynamic signal of mechanical vibration and shaft system swing.

[0053] The edge controller includes a container and a PLC. The container includes a plurality of independent containers, which are generated by different containers through virtualization Docker technology. The PLC includes a main PLC and a backup PLC.

[0054] With the reform of energy structure and power market, the traditional data acquisition and processing system has been difficult to support the efficient and intelligent operation of hydropower station, and cannot simultaneously collect equipment voiceprint, instrument data and vibration signal.

[0055] Optionally, the sensor collects initial data, including initial voiceprint data, initial visual data and initial swing data, and transmits the initial data to containers in the edge controller, different containers process respective initial data, and each container obtains respective target data. For example, the voiceprint container processes the initial voiceprint data to obtain target voiceprint data.

[0056] Optionally, the sensor collects target data, including target voiceprint data, target visual data and target swing data, and transmits the target data to containers in the edge controller, and different containers simultaneously obtain respective target data. For example, the voiceprint container obtains target voiceprint data.

[0057] In step S204, the target voiceprint data, the target visual data and the target swing data are analyzed by the containers deployed in the edge controller to obtain corresponding analysis results.

[0058] Optionally, the edge controller analyzes the target voiceprint data by the voiceprint container deployed in the edge controller to obtain the analysis result corresponding to the target voiceprint data, analyzes the target visual data by the visual container deployed in the edge controller to obtain the analysis result corresponding to the target visual data, and analyzes the target swing data by the swing container deployed in the edge controller to obtain the analysis result corresponding to the target swing data.

[0059] In step S206, in the case of an abnormality of the main programmable logic controller, the backup programmable logic controller generates corresponding control instructions for each analysis result and executes a corresponding control strategy based on the control instructions. The backup programmable logic controller and the main programmable logic controller store the same data.

[0060] Optionally, the containers transmit each analysis result and the target data to the main programmable logic controller, and the main PLC is synchronized in real time to the backup PLC through a high-speed synchronization link. The backup PLC continuously monitors the state of the main PLC, and as soon as a failure of the main PLC is detected, it immediately switches to a working state and continues to execute the step of generating corresponding control instructions for each analysis result, such as adjusting the operating parameters of the water pump, starting the backup device or issuing further detection instructions, based on the stored data. and executes a corresponding control strategy. This ensures uninterrupted device control and real-time monitoring of the main and backup PLC communication link through a heartbeat detection mechanism to ensure stable data transmission.

[0061] Optionally, the main PLC also collects and obtains regular electrical signals. At this time, the main PLC aggregates the target voiceprint data, the target visual data, the target swing data and the regular electrical signals.

[0062] Optionally, the visual, voiceprint, and vibration signals are collected by the sensors, and the collected raw signals are simultaneously uploaded to the corresponding containers of the monitoring configuration platform and the edge controller. In the containers, real-time analysis is performed through algorithms, and the analysis results are sent to the main PLC and the monitoring configuration platform. The main PLC uploads the generated control instructions and control strategies to the upper computer or the monitoring configuration platform through the network. The upper computer or the monitoring configuration platform can receive and save all collected raw signals or target data, algorithm analysis results, and control instructions for abnormal situations, providing users with real-time monitoring and historical data query convenience, greatly improving the efficiency and reliability of hydropower station operation and management. At the same time, the platform displays the working status of the main and standby PLCs in real time, and when the main and standby switching is detected, an alarm is sent in time, and the switching time, reason, and other information are recorded for easy tracing and analysis by the operation and maintenance personnel.

[0063] Optionally, during the operation of the main PLC, the target data and the corresponding analysis results of the target data are synchronized to the standby PLC in real time through NAPro software configuration. The standby programmable logic controller and the main programmable logic controller store the same data, ensuring the consistency and integrity of the main and standby data. The edge controller monitors the state of the main PLC in real time through the heartbeat detection mechanism of the standby PLC. In the case of abnormality of the main programmable logic controller, the standby programmable logic controller generates corresponding control instructions for each analysis result, such as target visual data for instrument panel data. When the pressure gauge reading is too high, i.e., exceeds the threshold, the relief valve needs to be turned on for pressure relief operation. At this time, a control instruction for turning on the relief valve is generated, and the corresponding control strategy is executed based on the control instruction, such as pressure relief operation according to the control instruction for turning on the relief valve.

[0064] It should be noted that the analysis results corresponding to the target voice data, the analysis results corresponding to the target vibration data, and the conventional electrical signals all generate control instructions through threshold setting and execute corresponding control operations based on the control instructions.

[0065] In the above hydropower station data processing method, first, the target data, including target voice data, target visual data, and target vibration data, are obtained through the container deployed in the edge controller. Second, the target voice data, target visual data, and target vibration data are analyzed through the container deployed in the edge controller to obtain corresponding analysis results. Finally, the standby programmable logic controller and the main programmable logic controller store the same data. In the case of abnormality of the main programmable logic controller, the standby programmable logic controller generates corresponding control instructions for each analysis result, and executes corresponding control strategies based on the control instructions. The standby PLC monitors the state of the main PLC in real time, ensures that the main PLC can be quickly switched in case of failure, guarantees the uninterrupted control flow, and maintains the normal operation of the system.

[0066] In one exemplary embodiment, as shown in Figure 3 The container includes a voiceprint container, a visual container, and a vibration container; the voiceprint container, the visual container, and the vibration container are independent of each other; the analysis result includes an identification result, an extraction result, and a processing result; the target voiceprint data, the target visual data, and the target vibration data are analyzed by the container deployed in the edge controller to obtain the corresponding analysis result, including steps S302 to S306. Wherein:

[0067] In step S302, the target voiceprint data is processed by the identification model deployed in the voiceprint container to obtain the identification result of the identification model.

[0068] The identification result represents the health status of the target voiceprint data; the identification model is determined based on a convolutional neural network with attention mechanism. The convolutional neural attention module (CBAM) with channel attention and spatial attention is added, which can better focus on the detailed features of the target, reduce the redundant feature information generated in feature extraction, and improve the representation ability.

[0069] In practical applications, considering that different types of signal data processing tasks may conflict with environmental requirements, different containers are generated by virtualizing Docker technology to provide independent processing environments for voiceprints, vision, and vibration signals.

[0070] Optionally, before deploying the identification model, the identification model is trained, and the model parameters of the identification model are quantized and compressed, so that the model can process voice data in real time on the edge side. The model analyzes the input audio data to obtain the classification of the target voiceprint data, which includes normal and abnormal, and generates an alarm message in the case of abnormality. In this way, it is determined whether the device is running normally.

[0071] As shown in Figure 4 The identification model extracts feature maps for the input target voiceprint data, inputs the extracted feature maps into the convolutional neural attention (CBAM) module, processes the feature maps through channel attention and spatial attention, and then processes the output vectors through two convolution layers and pooling layers in an alternating manner. The output vectors are input into a fully connected layer 1, a regular layer (Dropout), and a fully connected layer 2, and finally classified by a softmax classifier, either abnormal or normal.

[0072] In step S304, the target visual data is processed by the extraction model deployed in the visual container to obtain the extraction result of the extraction model.

[0073] The extraction result is used to represent the dashboard data. The extraction model includes a target detection model and a data extraction model.

[0074] Optionally, the edge controller processes the target visual data through a target detection algorithm deployed in the target detection model in the visual container to obtain processed target visual data, and processes the processed target visual data through an extraction algorithm deployed in the data extraction model to obtain an extraction result of the data extraction model.

[0075] In step S306, the target vibration data is processed through a vibration monitoring module deployed in the vibration container to obtain a processing result.

[0076] The processing result represents the health status of the target vibration data. The vibration monitoring module includes an online detection system based on an ARM processor and a data acquisition system based on an FPGA. The ARM is a master control unit of a measuring instrument, used for controlling key phase, vibration signal detection, calculation and communication, responding to PLC master module configuration download, time setting and data query commands through an internal communication interface, and providing an Ethernet interface to realize high-speed transmission of raw sample data with a computer. The high-speed vibration data acquisition module realizes synchronization and whole-cycle sampling control of multiple sensor signals through FPGA chip logic.

[0077] Optionally, the edge controller processes the target vibration data through a vibration monitoring module deployed in the vibration container to obtain a processing result. The processing includes real-time calculation of characteristic parameters such as amplitude, phase, frequency spectrum, axis trajectory, gap voltage, rotational speed and order analysis for the target vibration data.

[0078] In this embodiment, real-time algorithm analysis is mainly performed on target voiceprint data, target visual data and target vibration data, and abnormal conditions are processed in a timely manner, which can improve data sharing and break through data silos.

[0079] In one exemplary embodiment, as shown in Figure 5 The target data is obtained through the container deployed in the edge controller, including steps S502 to S508. Wherein:

[0080] In step S502, initial data is obtained through the container deployed in the edge controller. The initial data includes initial voiceprint data, initial visual data and initial vibration data.

[0081] The initial voiceprint data is collected based on a voiceprint sensor, the initial visual data is collected based on an image acquisition device, and the initial vibration data is collected based on a vibration sensor.

[0082] Optionally, for devices including water pumps, oil pumps, pointer instruments and generator sets, the initial voiceprint data of the devices is collected based on a voiceprint sensor. The initial visual data of the devices is collected based on an image acquisition device such as a camera. The initial vibration data of the devices is collected based on a vibration sensor.

[0083] In step S504, the initial voiceprint data is denoised, echo-canceled and enhanced by the container deployed in the edge controller to obtain target voiceprint data.

[0084] Optionally, the edge controller performs array signal denoising, echo cancellation and data enhancement preprocessing on the initial voiceprint data by the voiceprint container deployed in the edge controller to obtain target voiceprint data, including: the initial voiceprint data is an original voiceprint signal with environmental noise, and a denoised signal is obtained through a beam forming and spatial filtering algorithm; an enhanced signal is obtained by processing the denoised signal through time domain warping and impulse response simulation; target voiceprint data is obtained by processing the enhanced signal through adaptive filtering and nonlinear residual suppression. The target voiceprint data is a pure voiceprint signal.

[0085] In step S506, the initial visual data is enhanced by the container deployed in the edge controller to obtain target visual data.

[0086] Optionally, the edge controller enhances the initial visual data by the visual container deployed in the edge controller to obtain target visual data. The enhancement of the initial visual data to obtain the target visual data includes: spatial domain enhancement of the initial visual data to obtain a basic image; frequency domain enhancement and depth increase of the basic image to obtain the target visual data.

[0087] In step S508, the initial swing data is preprocessed by biasing, amplifying, filtering and integral denoising by the container deployed in the edge controller to obtain preprocessed initial swing data; and the preprocessed initial swing data is converted into standard data to obtain target swing data.

[0088] Optionally, to ensure signal detection accuracy, the initial swing data is preprocessed by biasing, amplifying, filtering and integral denoising by the swing container deployed in the edge controller to obtain preprocessed initial swing data; and the preprocessed initial swing data is converted into standard data such as a standard input signal for A / D sampling by the swing container deployed in the edge controller to obtain target swing data.

[0089] In this embodiment, the initial data is preprocessed by the algorithm in the container, which can improve the correctness of the processing result.

[0090] In an example embodiment, the target vibration data includes a generator vibration sensor signal, a swing sensor signal, and a key phase sensor signal; the target vibration data is processed by a vibration monitoring module deployed in a vibration container to obtain a processing result, including: the vibration monitoring module deployed in the vibration container performs calculation and analysis on the generator vibration sensor signal, the swing sensor signal, and the key phase sensor signal to obtain the processing result.

[0091] The generator vibration sensor signal measures the vibration of the generator bearing seat, frame, or shell. The vibration is mechanical vibration caused by rotor imbalance, misalignment, bearing damage, etc. It represents the mechanical vibration acceleration of the rotating parts (rotor, bearing) of the unit and reflects the dynamic imbalance state.

[0092] The swing sensor signal measures the radial displacement (swing) of the generator main shaft. It reflects the amount of radial displacement change of the main shaft within the bearing clearance and reflects the shaft system bending and oil film state.

[0093] The key phase sensor signal represents the synchronous pulse generated by the rotor for each revolution, providing an angular reference.

[0094] Optionally, the vibration monitoring module deployed in the vibration container extracts vibration features of the generator vibration sensor signal, extracts swing features of the swing sensor signal, and extracts key phase features of the key phase sensor signal; the vibration features, swing features, and key phase features are fused to obtain a two-dimensional holographic spectrum, i.e., the processing result.

[0095] In this embodiment, the processing result reflecting the mechanical state of the generator unit is obtained by jointly analyzing the three signals.

[0096] In an example embodiment, as shown in Figure 6 the target visual data is processed by an extraction model deployed in a visual container to obtain an extraction result of the extraction model, including steps S602 to S608. Wherein:

[0097] Step S602, the target visual data is cropped and background-removed by a target detection algorithm deployed in the visual container to obtain a dashboard area.

[0098] Optionally, the edge controller uses a detection box to position the target visual data by the target detection algorithm deployed in the visual container to obtain an initial dashboard area, extracts the ROI area of interest of the initial dashboard by mask technology, removes the background of the ROI area, and crops other areas to obtain the dashboard area, e.g., a regular circular area.

[0099] Step S604: The dashboard area is binarized and contours are found using a contour processing algorithm deployed in the visual container to obtain the key contours.

[0100] The key contours include the scale line contour, the pointer contour, and the interference point contour.

[0101] Contour processing algorithms tend to focus more on upper-level semantic objects when dealing with object boundaries in the dashboard area. For example, the findContours() function in OpenCV obtains each contour and stores it as a point vector. In addition, it also obtains the topological information of the image, namely the index numbers of the next contour, the previous contour, the parent contour, and the inset contours of a given contour.

[0102] Optionally, the edge controller binarizes the dashboard area using the `findContours()` function in OpenCV deployed within the vision container, obtaining a binarized image. Contour detection is then performed on the binarized image using contour lookup, and contour analysis is conducted using the `drawContours()` function. Through contour analysis, various effective attribute information of the contours can be obtained, outputting key contours such as tick mark contours, pointer contours, and interference point contours.

[0103] Step S606: Based on the key contour, obtain the center of the dashboard area and the pointer line.

[0104] Optionally, the edge controller obtains the center of the dashboard area, such as the center of a circle and the pointer line, based on the key contours of the OpenCV deployed in the vision container.

[0105] Step S608: Based on the center of the dashboard area, the pointer line, and the zero mark direction, the extraction result is obtained.

[0106] Optionally, the edge controller, using OpenCV deployed within the vision container, calculates the angle between the pointer line and the zero-scale direction based on the center of the dashboard area, such as the center of a circle, the pointer line, and the zero-scale direction, using an angle formula. Calculate the reading based on the instrument's scale division: Reading = ×(range / 360°) yields the extraction result. The included angle formula is shown in formula (1).

[0107] =arccos((V_zero·V_pointer) / (|V_zero||V_pointer|)) Formula (1)

[0108] In the formula, The angle between the pointer line and the zero mark direction, with a value ranging from 0 to 2. ; V_zero is a zero scale vector, the physical meaning is a vector from the center of the instrument to the zero scale point; V_pointer is a pointer vector, the physical meaning is a vector from the center of the instrument to the tip of the pointer; | | represents the modulus.

[0109] In this embodiment, the instrument panel data can be accurately extracted through the target detection algorithm and the contour processing algorithm.

[0110] In one exemplary embodiment, as shown in Figure 7 the center and the pointer straight line of the instrument panel region are obtained based on the key contour, including steps S702 to S706. Among them:

[0111] In step S702, the scale line contour is extracted from the key contour through the first preset condition.

[0112] Among them, the first preset condition includes at least one of the distance condition, the aspect ratio condition and the area condition. For example, the first preset condition excludes the contour with an area <10px² or an aspect ratio >5:1.

[0113] Optionally, the edge controller extracts a plurality of scale line contours from the key contour based on the first preset condition of excluding the contour with an area <10px² or an aspect ratio >5:1 through the deployed OpenCV deployed in the visual container, wherein the plurality of scale line contours includes at least two scale line contours.

[0114] In step S704, the center of the instrument panel region is obtained according to the intersection of the scale line contours.

[0115] Optionally, the edge controller uses the least square method to fit the scale line contours for the at least two scale line contours through the deployed OpenCV deployed in the visual container, obtains at least two scale line straight lines, calculates the intersection of the at least two scale line straight lines, takes the median value of the plurality of intersection coordinates to eliminate errors, and obtains the center of the instrument panel region.

[0116] In step S706, the key contour is filtered through the second preset condition to obtain the pointer contour, and the pointer straight line is obtained based on the pointer contour.

[0117] Among them, the second preset condition includes excluding the contour other than the pointer contour.

[0118] Optionally, the edge controller filters the key contour based on the second preset condition of excluding the contour other than the pointer contour through the deployed OpenCV deployed in the visual container, and the remaining contour is the pointer contour. The least square method is used to fit the pointer contour to obtain the pointer straight line.

[0119] In this embodiment, the center and the pointer straight line are extracted through the key contour, which provides a basis for subsequent extraction of the instrument panel data.

[0120] It should be understood that, although the steps in the flowcharts involved in the embodiments described above are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately or alternately executed with at least part of other steps or steps or stages in other steps.

[0121] Based on the same inventive concept, the embodiments of the present application also provide a hydropower station data processing device for implementing the above-mentioned hydropower station data processing method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more hydropower station data processing device embodiments provided below can refer to the limitations of the hydropower station data processing method described above, which will not be repeated here.

[0122] In one exemplary embodiment, as shown in Figure 8 a hydropower station data processing device is provided, comprising: an acquisition module 801, an analysis module 802, and a generation and execution module 803, wherein:

[0123] The acquisition module 801 is configured to acquire target data through a container deployed in an edge controller, the target data including target voiceprint data, target visual data, and target vibration data.

[0124] The analysis module 802 is configured to analyze the target voiceprint data, the target visual data, and the target vibration data through the container deployed in the edge controller to obtain corresponding analysis results.

[0125] The generation and execution module 803 is configured to, in the case of an exception of a main programmable logic controller, generate corresponding control instructions for each analysis result through a backup programmable logic controller, and execute a corresponding control strategy based on the control instructions; the backup programmable logic controller and the main programmable logic controller store the same data.

[0126] In an example embodiment, the container includes a voiceprint container, a visual container, and a swing container; the voiceprint container, the visual container, and the swing container are independent of each other; the analysis result includes an identification result, an extraction result, and a processing result; the analysis module 802 is further configured to process the target voiceprint data by using an identification model deployed in the voiceprint container to obtain an identification result of the identification model, the identification result representing a health state of the target voiceprint data; the identification model is determined based on a neural network of an attention mechanism; process the target visual data by using an extraction model deployed in the visual container to obtain an extraction result of the extraction model, the extraction result representing the dashboard data; and process the target swing data by using a swing monitoring module deployed in the swing container to obtain a processing result, the processing result representing a health state of the target swing data.

[0127] In an example embodiment, the acquisition module 801 is further configured to acquire initial data by using a container deployed in the edge controller, the initial data including initial voiceprint data, initial visual data, and initial swing data; the initial voiceprint data is acquired based on a voiceprint sensor, the initial visual data is acquired based on an image acquisition device, and the initial swing data is acquired based on a swing sensor; perform noise reduction, echo cancellation, and enhancement preprocessing on the initial voiceprint data by using the container deployed in the edge controller to obtain the target voiceprint data; perform enhancement processing on the initial visual data by using the container deployed in the edge controller to obtain the target visual data; perform biasing, amplification, filtering, and integral noise reduction preprocessing on the initial swing data by using the container deployed in the edge controller to obtain preprocessed initial swing data; and convert the preprocessed initial swing data into standard data to obtain the target swing data.

[0128] In an example embodiment, the target swing data includes a generator vibration sensor signal, a swing sensor signal, and a key phase sensor signal; the analysis module 802 is further configured to calculate and analyze the generator vibration sensor signal, the swing sensor signal, and the key phase sensor signal by using a swing monitoring module deployed in the swing container to obtain the processing result.

[0129] In an example embodiment, the analysis module 802 is further configured to crop the target visual data, remove the background, and obtain a dashboard region by using a target detection algorithm deployed in the visual container; perform binarization and contour searching on the dashboard region by using a contour processing algorithm deployed in the visual container to obtain a key contour; the key contour includes a scale line contour, a pointer contour, and an interference point contour; based on the key contour, obtain a center of the dashboard region and a pointer straight line; and based on the center of the dashboard region, the pointer straight line, and a zero scale direction, obtain the extraction result.

[0130] In an example embodiment, the analysis module 802 is further configured to extract the scale line contour from the key contour by a first preset condition; the first preset condition comprises at least one of a distance condition, an aspect ratio condition, and an area condition; obtain the center of the dashboard region according to the intersection of the scale line contour; filter the key contour by a second preset condition to obtain the pointer contour, and obtain the pointer straight line based on the pointer contour; the second preset condition comprises excluding the contour outside the pointer contour.

[0131] The modules in the hydropower station data processing apparatus can be implemented by software, hardware, or a combination thereof. The modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the modules.

[0132] In an example embodiment, a computer device is provided, which can be a server, and an internal structure diagram of the computer device can be as shown in Figure 9 The computer device includes a processor, a memory, an input / output interface, and a communication interface. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store hydropower station data. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with terminals outside through a network connection. The computer program is executed by the processor to implement a hydropower station data processing method.

[0133] Those skilled in the art can understand that Figure 9 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0134] In an example embodiment, a computer device is provided, which can be a server, and an internal structure diagram of the computer device can be as shown in

[0135] In an embodiment, a computer readable storage medium is provided, which stores a computer program. The computer program is executed by a processor to implement the steps of the above method embodiments.

[0136] In an embodiment, a computer program product is provided, which includes a computer program. The computer program is executed by a processor to implement the steps of the above method embodiments.

[0137] A person of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, the computer program can include the processes of the above method embodiments. Any reference to a memory, a database or other medium in each embodiment provided by the present application can include at least one of a non-volatile memory and a volatile memory. The non-volatile memory can include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical storage, a high-density embedded non-volatile memory, a resistive random access memory (ReRAM), a magnetoresistive random access memory (MRAM), a ferroelectric random access memory (FRAM), a phase change memory (PCM), a graphene memory, etc. The volatile memory can include a random access memory (RAM) or an external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc. The database involved in each embodiment provided by the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., and is not limited thereto. The processor involved in each embodiment provided by the present application can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., and is not limited thereto.

[0138] Any technical features in the above embodiments can be combined, and for the sake of brevity, not all possible combinations are described above, however, any combination of these technical features is deemed to be within the scope of the present application.

[0139] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be pointed out that, for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A data processing method for a hydropower station, characterized in that, When applied to an edge controller, the method includes: Target data is acquired through containers deployed in an edge controller. The target data includes target acoustic data, target visual data, and target vibration data. The containers are generated using virtualization Docker technology and include an acoustic container, a visual container, and a vibration container. The acoustic container, the visual container, and the vibration container are independent of each other. The target vibration data includes generator vibration sensor signals, oscillation sensor signals, and key phase sensor signals. The target voiceprint data, target visual data, and target vibration data are analyzed by a container deployed in the edge controller to obtain corresponding analysis results. These results include recognition results, extraction results, and processing results. Specifically, the analysis includes: processing the target voiceprint data using a recognition model deployed in the voiceprint container to obtain the recognition result of the recognition model, which characterizes the health status of the target voiceprint data; the recognition model is determined based on an attention-based neural network; processing the target visual data using an extraction model deployed in the visual container to obtain the extraction result of the extraction model, which characterizes the dashboard data; and processing the target vibration data using a vibration monitoring module deployed in the vibration container to obtain a processing result, which characterizes the health status of the target vibration data. When the backup programmable logic controller (PLC) detects an anomaly in the main PLC, the backup PLC generates corresponding control instructions based on the analysis results and executes the corresponding control strategy based on the control instructions; the backup PLC and the main PLC store the same data. The method further includes: monitoring the communication link between the main programmable logic controller and the backup programmable logic controller in real time through a heartbeat detection mechanism.

2. The method according to claim 1, characterized in that, The process of acquiring target data through containers deployed in the edge controller includes: Initial data is acquired through containers deployed in the edge controller. This initial data includes initial voiceprint data, initial visual data, and initial oscillation data. Specifically, the initial voiceprint data is acquired based on a voiceprint sensor, the initial visual data is acquired based on an image acquisition device, and the initial oscillation data is acquired based on an oscillation sensor. The initial voiceprint data is preprocessed by denoising, echo cancellation, and enhancement using containers deployed in the edge controller to obtain the target voiceprint data. The initial visual data is augmented by a container deployed in an edge controller to obtain the target visual data; The initial oscillation data is preprocessed by biasing, amplifying, filtering, and integrating noise reduction through a container deployed in the edge controller to obtain preprocessed initial oscillation data; the preprocessed initial oscillation data is then converted into standard data to obtain the target oscillation data.

3. The method according to claim 2, characterized in that, The target oscillation data includes generator vibration sensor signals, oscillation sensor signals, and key phase sensor signals; the oscillation monitoring module deployed in the oscillation container processes the target oscillation data to obtain processing results, including: The vibration monitoring module deployed in the vibration container calculates and analyzes the generator vibration sensor signal, the oscillation sensor signal, and the key phase sensor signal to obtain the processing results.

4. The method according to claim 1, characterized in that, The step of processing the target visual data using an extraction model deployed in the visual container to obtain the extraction result of the extraction model includes: The target visual data is cropped and background removed by a target detection algorithm deployed in the visual container to obtain the dashboard area. The dashboard area is binarized and contour searched by a contour processing algorithm deployed in the visual container to obtain key contours; the key contours include scale line contours, pointer contours and interference point contours. Based on the key contour, the center of the dashboard area and the pointer line are obtained; The extraction result is obtained based on the center of the dashboard area, the pointer line, and the zero mark direction.

5. The method according to claim 4, characterized in that, The process of obtaining the center and pointer line of the dashboard area based on the key contour includes: The scale line contour is extracted from the key contour by means of a first preset condition; the first preset condition includes at least one of distance condition, aspect ratio condition and area condition. The center of the instrument panel area is obtained from the intersection of the scale lines. The key contour is filtered by a second preset condition to obtain a pointer contour, and a pointer line is obtained based on the pointer contour; the second preset condition includes contours that exclude the pointer contour.

6. A data processing device for a hydropower station, characterized in that, The device, used in edge controllers, includes: The acquisition module is used to acquire target data through containers deployed in the edge controller. The target data includes target acoustic data, target visual data, and target vibration data. The containers are generated using virtualization Docker technology and include an acoustic container, a visual container, and a vibration container. The acoustic container, the visual container, and the vibration container are independent of each other. The target vibration data includes generator vibration sensor signals, oscillation sensor signals, and key phase sensor signals. An analysis module is used to analyze the target voiceprint data, target visual data, and target vibration data through containers deployed in an edge controller, and obtain corresponding analysis results. The analysis results include recognition results, extraction results, and processing results. Specifically, it includes: processing the target voiceprint data through a recognition model deployed in the voiceprint container to obtain the recognition result of the recognition model, which characterizes the health status of the target voiceprint data; the recognition model is determined based on an attention-based neural network; processing the target visual data through an extraction model deployed in the visual container to obtain the extraction result of the extraction model, which characterizes dashboard data; and processing the target vibration data through a vibration monitoring module deployed in the vibration container to obtain a processing result, which characterizes the health status of the target vibration data. The generation and execution module is used to generate corresponding control instructions based on the analysis results when the main programmable logic controller is detected to be abnormal by the backup programmable logic controller, and to execute the corresponding control strategy based on the control instructions; the backup programmable logic controller and the main programmable logic controller store the same data; the communication link between the main programmable logic controller and the backup programmable logic controller is monitored in real time through a heartbeat detection mechanism.

7. The apparatus according to claim 6, characterized in that, The acquisition module is further configured to acquire initial data through a container deployed in the edge controller. The initial data includes initial voiceprint data, initial visual data, and initial oscillation data. The initial voiceprint data is acquired based on a voiceprint sensor, the initial visual data is acquired based on an image acquisition device, and the initial oscillation data is acquired based on a oscillation sensor. The initial voiceprint data is preprocessed by noise reduction, echo cancellation, and enhancement through the container deployed in the edge controller to obtain target voiceprint data. The initial visual data is enhanced through the container deployed in the edge controller to obtain target visual data. The initial oscillation data is preprocessed by biasing, amplifying, filtering, and integral noise reduction through the container deployed in the edge controller to obtain preprocessed initial oscillation data. The preprocessed initial oscillation data is converted into standard data to obtain target oscillation data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images