Hydroelectric generating set operation state monitoring and evaluating cloud side cooperation device

Abstract

The invention relates to the field of hydropower station equipment monitoring, in particular to a hydroelectric generating set operation state monitoring and evaluation cloud edge collaboration device which comprises a multi-channel data acquisition module, a data processing communication module and an operation state evaluation module. Real-time monitoring of running state data of the mixed-flow water turbine set is achieved; caching, noise reduction preprocessing and feature extraction are carried out on the collected fast variable signals, and feature information reflecting the running state of the unit is sent to the running state evaluation module through an optical fiber cable by utilizing a network communication protocol; on the basis of obtaining unit operation condition data sent by a power station on-line monitoring system and feature information transmitted by a data processing communication module, computational fluid mechanics analysis, unit runner pressure high-fidelity agent model construction, unit operation state evaluation and evaluation result sending tasks are carried out. By deploying the cloud edge collaboration device, the running state of the water turbine can be monitored and evaluated in real time.

Description

Technical Field

[0001] The present application relates to the field of hydropower station equipment monitoring, and specifically to a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit. Background Art

[0002] As the core equipment of a hydropower plant, the operating status of a hydropower unit is directly related to the power generation efficiency and safety of the hydropower plant. However, due to the complex operating environment of the hydropower unit, various faults such as cavitation erosion, fatigue cracks, and insufficient preload of bolts 8 are prone to occur during operation, resulting in equipment damage and reduced power generation efficiency. In order to ensure the stable operation of the hydropower unit, improve power generation efficiency, and reduce operation and maintenance costs, it is necessary to monitor and evaluate the operating status of the hydropower unit in real time.

[0003] Traditional monitoring and evaluation methods mainly rely on manual inspections and wired sensors for data collection, which has certain limitations. On the one hand, manual inspections are labor-intensive, inefficient, and unable to grasp the status of equipment in real time; on the other hand, wired sensors have complex wiring, are difficult to install and maintain, and are susceptible to environmental interference. In addition, as the scale of hydropower plants continues to expand, the amount of monitoring data is also increasing, and traditional data processing methods can no longer meet the requirements of real-time and accuracy.

[0004] In response to the above problems, the development of cloud computing and edge computing technologies in recent years has provided a new solution for monitoring and evaluating the operating status of hydropower units. Cloud computing has powerful data processing and storage capabilities, which can realize the rapid processing and analysis of massive monitoring data; edge computing can migrate some computing tasks from the cloud to the device side, reduce data transmission delays, and improve real-time performance. Therefore, combining cloud computing and edge computing technologies to build a cloud-edge collaborative hydropower unit operating status monitoring and evaluation device is of great significance to improving the operating efficiency and safety of hydropower units. Summary of the invention

[0005] In order to realize real-time monitoring and evaluation of the operating status of hydropower units, improve power generation efficiency and safety, and reduce operation and maintenance costs, it is necessary to study a cloud-edge collaborative device for monitoring and evaluating the operating status of hydropower units to solve the above-mentioned problems.

[0006] According to a first aspect of the present invention, the present invention claims protection for a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit, comprising:

[0007] A multi-channel data acquisition module located at the client, a data processing and communication module located at the edge, and an operation status evaluation module located in the cloud;

[0008] The multi-channel data acquisition module includes a plurality of sensors, which are deployed at the component measuring point positions of each turbine of the turbine group to be monitored, and synchronously collect and monitor the vibration signals, pressure pulsation signals and pressure signals of the deployment positions of the sensor points in real time;

[0009] The data processing and communication module realizes data caching, data processing and feature signal extraction by deploying distributed data processing and communication cabinets on site, and its two ends are respectively connected to the multi-channel data acquisition module and the operation status evaluation module, and is used to perform data processing and feature extraction on the vibration signal, pressure pulsation signal and pressure signal monitored and collected by the multi-channel data acquisition module, and send them to the operation status evaluation module;

[0010] The operation status assessment module combines the operation condition data of the existing online monitoring system of the power station and the characteristics of the transmission of the data processing and communication module to carry out computational fluid dynamics analysis and construct a high-fidelity proxy model that describes the flow channel pressure of the turbine unit to carry out the unit operation status assessment.

[0011] Furthermore, the device also includes:

[0012] The component measurement point positions of the multi-channel data acquisition module at least include:

[0013] Upper and lower frames, volute, draft tube and bolts 8;

[0014] The sensors at least include a multi-channel vibration sensor, a multi-channel pressure pulsation sensor and a multi-channel pressure sensor;

[0015] The monitoring signal of the sensor is amplified and collected by a corresponding multi-channel data acquisition card;

[0016] The multi-channel vibration sensor and the multi-channel pressure pulsation sensor both have four channels;

[0017] Four vibration sensors are arranged at four mutually perpendicular positions of the upper and lower frames of the turbine;

[0018] Four-way pressure pulsation sensors are arranged at the draft tube and the volute inlet and outlet of the turbine;

[0019] The multi-way pressure sensor is arranged at the outlet of the tailwater pipe; the vibration sensor of bolt 8 is arranged at bolt 8 and bolt 8 fastener 7;

[0020] The pressure pulsation signal acquisition card in the multi-channel data acquisition module adopts the same-starting point full-cycle sampling method, and the collected turbine pressure pulsation signal covers the abnormal operation state and transition process of the turbine.

[0021] Furthermore, the device also includes:

[0022] The data processing and communication module includes a plurality of on-site data processing and communication cabinets;

[0023] The data processing communication module is connected with a signal acquisition card, a data buffer device and a data processing device via a signal cable;

[0024] The data processing and communication module is connected in sequence to an on-site photoelectric converter, an on-site optical fiber distribution frame, a remote optical fiber distribution frame and a remote photoelectric converter through an optical fiber cable; the on-site photoelectric converter is connected to the data processing and communication module through a network cable; and the remote photoelectric converter is connected to the operation status evaluation module through a network cable;

[0025] Each of the on-site data processing and communication cabinets is composed of a data buffer device and a data processing device, and is arranged at different locations according to the requirements of data collection;

[0026] The data processing device in the data communication processing module processes the collected sensor signals, performs noise reduction and feature extraction on the fast variable pressure pulsation signal and the vibration signal, and retains the original features of the slow variable pressure signal without filtering.

[0027] Furthermore, the device also includes:

[0028] The operation status evaluation module is the cloud server of the cloud-edge collaboration device, and the cloud server is deployed in the computer room of the power station. The operation status evaluation module includes computational fluid dynamics, machine learning model training, turbine performance evaluation and evaluation result sending function modules;

[0029] The computational fluid dynamics function combines the operating condition data transmitted by the existing online monitoring system of the power station and the characteristic signal output by the data processing and communication module to carry out computational fluid dynamics analysis, calculate the corresponding operating condition data, and prepare for the agent model simulation;

[0030] The machine model learning and training function module uses the operating data of the fluid mechanics calculation to carry out agent model training. Based on the integrated model, the ability of the artificial neural network and the support vector machine is synergistically combined, so that the agent model can handle complex nonlinear relationships while minimizing the risk of overfitting and improving the ability to evaluate the operating status of the turbine.

[0031] Furthermore, when the device carries out the unit operation status assessment, the assessment is performed based on the splash water image at the outlet of the turbine, and specifically includes:

[0032] Collecting the water turbine units to be evaluated that need to be evaluated for their running status, collecting splashing videos of the water turbine units to be evaluated, and obtaining a plurality of splashing video frame images of the water turbine units to be evaluated;

[0033] Preprocessing the splashing video frame image of the water turbine set to be evaluated to obtain a water discharge image of the water turbine set to be evaluated and analyzed of the splashing video frame image of the water turbine set to be evaluated;

[0034] Collecting the dry-wet balance value in the water outlet image of the water turbine unit to be evaluated and analyzed, and determining the operating state of the splash video frame image of the water turbine unit to be evaluated;

[0035] Fault assessment statistics are performed on the operating status of multiple consecutive water discharge images of the water turbine unit to be evaluated and analyzed collected by the splashing video to obtain a fault assessment of the operating status of the water turbine unit to be evaluated.

[0036] The present application relates to the field of hydropower station equipment monitoring, and specifically to a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit, including a multi-channel data acquisition module, a data processing and communication module, and an operating status evaluation module. By synchronously acquiring the signals of the sensing points installed at the mixed flow turbine unit, the operating status data of the mixed flow turbine unit can be monitored in real time; the acquired fast variable signals are cached, pre-processed for noise reduction, and feature extracted, and the characteristic information reflecting the operating status of the unit is sent to the operating status evaluation module via the optical fiber cable using the network communication protocol; on the basis of acquiring the unit operating condition data sent by the power station online monitoring system and the characteristic information transmitted by the data processing and communication module, computational fluid dynamics analysis, construction of a high-fidelity proxy model of the unit flow channel pressure, unit operating status evaluation, and sending of evaluation results are carried out. By deploying the cloud-edge collaborative device, the present invention can monitor and evaluate the operating status of the turbine in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A structural block diagram of a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit provided in an embodiment of the present invention;

[0038] Figure 2 A cross-sectional view of the deployment positions of the sensing points of the vibration sensor, the pressure pulsation sensor and the pressure sensor in a multi-channel data acquisition module of a cloud-edge collaborative device for monitoring and evaluating the operation status of a hydropower unit provided in an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of the deployment position of the vibration sensor measuring point of bolt 8 and the connection of the signal cable in a multi-channel data acquisition module of a cloud-edge collaborative device for monitoring and evaluating the operation status of a hydropower unit provided in an embodiment of the present invention;

[0040] Figure 4 A schematic diagram of the hardware device of a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit provided in an embodiment of the present invention;

[0041] Figure 5A distributed layout diagram of distributed on-site data processing and communication cabinets in a power station in a data processing and communication module of a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit provided in an embodiment of the present invention;

[0042] Figure 6 A schematic diagram of a data processing device in a data processing communication module of a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit provided in an embodiment of the present invention;

[0043] Figure 7 A schematic diagram of the operation status assessment module of a cloud-edge collaborative device for monitoring and assessing the operation status of a hydropower unit provided in an embodiment of the present invention, which performs computational fluid dynamics analysis, builds a high-fidelity proxy model of the unit flow channel pressure, assesses the operation status of the unit, and sends the assessment result;

[0044] Figure markings: upper frame vibration measuring point-1, lower frame vibration measuring point-2, volute inlet pressure pulsation measuring point-3, volute outlet pressure pulsation measuring point-4, tailwater cone pressure pulsation measuring point-5, tailwater pipe pressure measuring point-6, fastener-7, bolt-8. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0046] The terms "first", "second" and "third" in this application are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications in the embodiments of the present application (such as up, down, left, right, front, back...) are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0047] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0048] According to the first aspect of the present invention, the present invention claims a cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit, referring to Figure 1 ,include:

[0049] A multi-channel data acquisition module located at the client, a data processing and communication module located at the edge, and an operation status evaluation module located in the cloud;

[0050] The multi-channel data acquisition module includes a plurality of sensors, which are deployed at the component measuring point positions of each turbine of the turbine group to be monitored, and synchronously collect and monitor the vibration signals, pressure pulsation signals and pressure signals of the deployment positions of the sensor points in real time;

[0051] The data processing and communication module realizes data caching, data processing and feature signal extraction by deploying distributed data processing and communication cabinets on site, and its two ends are respectively connected to the multi-channel data acquisition module and the operation status evaluation module, and is used to perform data processing and feature extraction on the vibration signal, pressure pulsation signal and pressure signal monitored and collected by the multi-channel data acquisition module, and send them to the operation status evaluation module;

[0052] The operation status assessment module combines the operation condition data of the existing online monitoring system of the power station and the characteristics of the transmission of the data processing and communication module to carry out computational fluid dynamics analysis and construct a high-fidelity proxy model that describes the flow channel pressure of the turbine unit to carry out the unit operation status assessment.

[0053] Furthermore, the device also includes:

[0054] Reference Figure 2 and 3 , the component measurement point positions of the multi-channel data acquisition module at least include:

[0055] Upper and lower frames, volute, draft tube and bolts 8;

[0056] The upper and lower frames include an upper frame vibration measuring point 1 and a lower frame vibration measuring point 2;

[0057] The volute includes a volute inlet pressure pulsation measuring point 3 and a volute outlet pressure pulsation measuring point 4;

[0058] The tailwater pipe includes a tailwater cone pressure pulsation measuring point 5 and a tailwater pipe pressure measuring point 6;

[0059] The sensors at least include a multi-channel vibration sensor, a multi-channel pressure pulsation sensor and a multi-channel pressure sensor;

[0060] The monitoring signal of the sensor is amplified and collected by a corresponding multi-channel data acquisition card;

[0061] The multi-channel vibration sensor and the multi-channel pressure pulsation sensor both have four channels;

[0062] Four vibration sensors are arranged at four mutually perpendicular positions of the upper and lower frames of the turbine;

[0063] Four-way pressure pulsation sensors are arranged at the draft tube and the volute inlet and outlet of the turbine;

[0064] The multi-way pressure sensor is arranged at the outlet of the tailwater pipe; the vibration sensor of bolt 8 is arranged at bolt 8 and bolt 8 fastener 7;

[0065] The pressure pulsation signal acquisition card in the multi-channel data acquisition module adopts the same-starting point full-cycle sampling method, and the collected turbine pressure pulsation signal covers the abnormal operation state and transition process of the turbine.

[0066] Among them, in this embodiment, the vibration acceleration sensor selects an integrated vibration transmitter sensor, whose response frequency is 5-1000Hz, accuracy is ±0.3%, and working voltage is 24VDC. The pressure pulsation sensor is, for example, a Keller pressure transmitter, whose response speed is 0.25ms, accuracy is ±0.3%, and working voltage is 8-28VDC. The pressure sensor is, for example, an HDP500 pressure transmitter, with a response range of 0-1MPa, accuracy of ±0.1%, and working voltage of 9-36VDC. The sampling frequency of the bolt 8 vibration sensor is 12.8KHz, and the number of sampling blocks is selected as 16 or 32 blocks according to actual needs. The acquisition card in the data processing communication module is integrated from multiple corresponding acquisition cards, including a vibration signal acquisition card and a pressure pulsation signal acquisition card. The sampling rate of the vibration signal acquisition card is 1kSPS, and the sampling rate of the pressure pulsation signal acquisition card is 4MSPS. Each acquisition card is dual-channel, the channel voltage is ±10V, and the ADC is 16 bits.

[0067] Furthermore, the pressure pulsation signal acquisition card adopts the same-starting point full-cycle sampling method. The sampling frequency of the acquisition card is 256 times the industrial frequency, and the highest analysis frequency can reach 128 times the industrial frequency. The collected turbine pressure pulsation signal can cover the abnormal operation state of the turbine and the transition process of the turbine.

[0068] Further, refer to Figure 4-6, the device further comprises:

[0069] The data processing and communication module includes a plurality of on-site data processing and communication cabinets;

[0070] The data processing communication module is connected with a signal acquisition card, a data buffer device and a data processing device via a signal cable;

[0071] The data processing and communication module is connected in sequence to an on-site photoelectric converter, an on-site optical fiber distribution frame, a remote optical fiber distribution frame and a remote photoelectric converter through an optical fiber cable; the on-site photoelectric converter is connected to the data processing and communication module through a network cable; and the remote photoelectric converter is connected to the operation status evaluation module through a network cable;

[0072] Each of the on-site data processing and communication cabinets is composed of a data buffer device and a data processing device, and is arranged at different locations according to the requirements of data collection;

[0073] The data processing device in the data communication processing module processes the collected sensor signals, performs noise reduction and feature extraction on the fast variable pressure pulsation signal and the vibration signal, and retains the original features of the slow variable pressure signal without filtering.

[0074] Further, refer to Figure 7 , the device further comprises:

[0075] The operation status evaluation module is the cloud server of the cloud-edge collaboration device, and the cloud server is deployed in the computer room of the power station. The operation status evaluation module includes computational fluid dynamics, machine learning model training, turbine performance evaluation and evaluation result sending function modules;

[0076] The computational fluid dynamics function combines the operating condition data transmitted by the existing online monitoring system of the power station and the characteristic signal output by the data processing and communication module to carry out computational fluid dynamics analysis, calculate the corresponding operating condition data, and prepare for the agent model simulation;

[0077] The machine model learning and training function module uses the operating data of the fluid mechanics calculation to carry out agent model training. Based on the integrated model, the ability of the artificial neural network and the support vector machine is synergistically combined, so that the agent model can handle complex nonlinear relationships while minimizing the risk of overfitting and improving the ability to evaluate the operating status of the turbine.

[0078] Among them, in this embodiment, the operation status evaluation module data is composed of the unit operation condition data collected by the existing monitoring system of the hydropower station and the characteristic information transmitted by the data processing and communication module. On the basis of obtaining the above data, nearly half of the noise reduction extraction is performed on it, and the selected sample space is obtained by combining parameter weight analysis and random sample sampling. Computational fluid dynamics analysis is performed using the operating condition data, and a mixed flow turbine simulation model is constructed to perform flow field simulation to obtain a mixed data set after the actual data and the simulation data are mixed and enhanced. The sample space is selected, and the operating condition parameters of the proxy model are selected according to the sample space. The comprehensive value of the input characteristic information of the proxy model is obtained according to the weight distribution of the mixed data of the sample space, and the proxy model is trained to obtain the proxy model training result. The training results are used to evaluate the operation status of the turbine, and the evaluation results are sent to the power station monitoring host computer through the network cable and communication protocol.

[0079] In this embodiment, the operation status evaluation module is arranged in a remote computer room, and its operation status evaluation unit is, for example, a precision T3630 workstation, and its dynamic display unit is, for example, a 19-inch 9U industrial LCD display.

[0080] Furthermore, when the device carries out the unit operation status assessment, the assessment is performed based on the splash water image at the outlet of the turbine, and specifically includes:

[0081] Collecting the water turbine units to be evaluated that need to be evaluated for their running status, collecting splashing videos of the water turbine units to be evaluated, and obtaining a plurality of splashing video frame images of the water turbine units to be evaluated;

[0082] Preprocessing the splashing video frame image of the water turbine set to be evaluated to obtain a water discharge image of the water turbine set to be evaluated and analyzed of the splashing video frame image of the water turbine set to be evaluated;

[0083] Collecting the dry-wet balance value in the water outlet image of the water turbine unit to be evaluated and analyzed, and determining the operating state of the splash video frame image of the water turbine unit to be evaluated;

[0084] Fault assessment statistics are performed on the operating status of multiple consecutive water discharge images of the water turbine unit to be evaluated and analyzed collected by the splashing video to obtain a fault assessment of the operating status of the water turbine unit to be evaluated.

[0085] In this embodiment, the step of collecting the water turbine unit to be evaluated that needs to be evaluated for its operating status, collecting splashing videos of the water turbine unit to be evaluated, and obtaining a plurality of splashing video frame images of the water turbine unit to be evaluated further comprises:

[0086] A camera is used to collect images of the splash outlet and splash ground of the turbine unit. Each splash outlet and splash ground of the turbine is divided into multiple rectangular areas for image collection, and the area of ​​each area is kept consistent.

[0087] The splashing video of each turbine splashing outlet and each area of ​​the splashing ground is collected three times, and each splashing video is collected by performing pre-photographing, mid-photographing and post-photographing processing on each turbine splashing outlet and each area of ​​the splashing ground;

[0088] The splash video of the first turbine unit to be evaluated is acquired by early photographing and splash video processing, the splash video of the second turbine unit to be evaluated is acquired by mid-term photographing and splash video processing, and the splash video of the third turbine unit to be evaluated is acquired by late photographing and splash video processing.

[0089] The preprocessing operation of the splashing video frame image of the water turbine set to be evaluated to obtain the water discharge image of the water turbine set to be evaluated and analyzed of the splashing video frame image of the water turbine set to be evaluated also includes:

[0090] The splashing video frame image of the water turbine set to be evaluated is screened and processed. If the surface of the splashing video frame image of the water turbine set to be evaluated is affected by oil or other stains, resulting in uneven shadows other than dry and wet in the image, the splashing video frame image of the water turbine set to be evaluated is discarded and re-shot, or the area with large shadow color difference is eliminated, or the area with large shadow difference is processed by software to distinguish the dry and wet areas from the normal areas by shadow;

[0091] The splashing video frame images of the water turbine set to be evaluated are cropped and cut out so that the actual areas of the water turbine sets in all images are the same as the areas divided when the image data is collected, and the areas are the same, and the splashing video frame images of the water turbine set to be evaluated are processed into the same size;

[0092] The processed video frame images of the splashing of the turbine unit to be evaluated of the same size are binarized, the wet area in the binary image becomes high grayscale, the dry area position presents low grayscale, and the dry and wet areas are distinguished from the surrounding normal plane.

[0093] The collecting of the dry-wet balance value in the water outlet image of the water turbine unit to be evaluated and analyzed, and determining the operating state of the splashing video frame image of the water turbine unit to be evaluated, further includes:

[0094] Collecting the number N1 of high grayscale feature points and N2 of low grayscale feature points in the water discharge image of the turbine unit to be evaluated and analyzed;

[0095] The proportion of high grayscale feature points to the total feature points and the proportion of dry and wet areas on the outer surface of the turbine unit are calculated and expressed as the surface dry-wet balance value K. The calculation formula is shown in formula (3):

[0096]

[0097] Where: K—surface dry-wet balance value%;

[0098] N1—Number of high grayscale feature points;

[0099] N2—Number of low grayscale feature points;

[0100] Determining the operating state of the splashing video frame image of the turbine unit to be evaluated according to the surface dry-wet balance value;

[0101] The operating states include a first operating state, a second operating state, a third operating state, and a fourth operating state.

[0102] The performing fault assessment statistics on the operating status of a plurality of continuous water discharge images of the to-be-assessed and analyzed hydraulic turbine unit collected by the splashing video to obtain the operating status fault assessment of the to-be-assessed hydraulic turbine unit further includes:

[0103] Sequentially perform operating status detection on the splashing video of the first to-be-evaluated water turbine unit, the splashing video of the second to-be-evaluated water turbine unit, and the splashing video of the third to-be-evaluated water turbine unit;

[0104] Determine and obtain a first average operating state, a second average operating state, and a third average operating state in each of the splashing videos of the hydraulic turbine set to be evaluated according to the surface dry-wet balance values ​​of a plurality of splashing video frames of the hydraulic turbine set to be evaluated in the first splashing video of the hydraulic turbine set to be evaluated, the second splashing video of the hydraulic turbine set to be evaluated, and the third splashing video of the hydraulic turbine set to be evaluated;

[0105] The operating state fault assessment result of the water turbine set to be assessed is obtained according to the operating state fault weights of the splashing video of the first water turbine set to be assessed, the splashing video of the second water turbine set to be assessed, and the splashing video of the third water turbine set to be assessed.

[0106] The present invention takes a mixed flow turbine unit of a certain hydropower station as the research object, and installs a multi-channel data sensor on the upper and lower frames, volute, tailwater pipe and bolt 8 of the turbine to collect monitoring data. The data acquisition signal includes a vibration signal, a pressure pulsation signal and a pressure signal. At the same time, the collected turbine pressure pulsation signal is required to cover the abnormal operation state of the turbine and the transition process of the turbine, that is, including the violent high-frequency signal changes of the pressure pulsation during the start-up and shutdown process of the turbine, the abnormal periodic pressure pulsation signal changes when the blades are abnormally running, and other characteristic signals under working conditions.

[0107] Based on the monitoring data of the power station, the original turbine signal data space is processed for noise reduction and feature extraction. For example, the wavelet transform and principal component analysis (PCA) methods are used to reduce noise and extract features of fast variable pressure pulsation signals and vibration signals, and the original features of slow variable pressure signals are retained without filtering. The calculation formulas of wavelet transform and principal component analysis are as follows:

[0108]

[0109] Z = XW (5);

[0110] Among them, X(t) is the wavelet transform result, c k is the wavelet coefficient, ψ k (t) is the wavelet basis function, Z is the reduced-dimensional data, X is the original data, and W is the projection matrix composed of eigenvectors.

[0111] Combined with the operating condition data transmitted by the existing online monitoring system of the power station and the characteristic signals output by the above data processing and communication module, computational fluid dynamics analysis is carried out to calculate the unit operating data under different conditions and prepare data for the agent model simulation. The calculation formula is as follows:

[0112]

[0113] P = η·ρ·g·H·Q (7);

[0114] Among them, V is the velocity vector, t is time, p is pressure, ρ is fluid density, f is volume force, μ is dynamic viscosity, P is power, η is unit efficiency, H is head, and Q is flow rate.

[0115] Based on the above processed data space, a mixed flow turbine proxy model is constructed. In order to improve the model prediction performance and regression ability, an advanced model based on neural network can be used for optimization and processing. For example, the artificial neural network (ANN) and support vector machine (SVR) models can be integrated, and the features learned from the artificial neural network are integrated into the input features of the support vector regression. This fusion utilizes the powerful nonlinear modeling ability of the artificial neural network, which can perform high-dimensional extraction and learning in the data space, and then use SVR to perform regression analysis on these learned features. The integration of these two algorithms, while making full use of their respective advantages, also makes up for their respective weaknesses, thereby improving the prediction performance, enhancing the model robustness, reducing the risk of overfitting, and obtaining a more powerful and flexible proxy model. The main technical parameters of the mixed flow turbine unit are shown in Table 1:

[0116] Table 1

[0117]

[0118]

[0119] In order to verify the superiority of this proxy model, this integrated model is compared with the ANN model, SVR model and GPR model. The mean relative error (MRE) is used as the model evaluation index. Each model is compared 10 times, and the average value is taken as the final evaluation index. The evaluation index formula is as follows:

[0120]

[0121] Where n is the number of samples, y i is the true value, Predicted values ​​for the model

[0122] The comparison results are shown in Table 2:

[0123] Table 2

[0124]

[0125] The study showed that all models achieved good results, with MRE values ​​below 2%. However, the integrated model ANN+SVR used in this study had the highest accuracy, with an MRE of only 0.6645%, achieving enhanced generalization and robustness. In addition, the integrated agent model of this study further improved computational efficiency, reduced operating costs, and could accurately predict the mechanical behavior of turbines under different operating conditions, improving state assessment capabilities.

[0126] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0127] In addition, each functional unit in each embodiment of the present application 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 above integrated unit can be implemented in the form of hardware or in the form of software functional units. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present application.

[0128] The specific implementation methods of the invention are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.

Claims

1. A cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit, characterized in that: include: A multi-channel data acquisition module located at the client, a data processing and communication module located at the edge, and an operation status evaluation module located in the cloud; The multi-channel data acquisition module includes a plurality of sensors, which are deployed at the component measuring point positions of each turbine of the turbine group to be monitored, and synchronously collect and monitor the vibration signals, pressure pulsation signals and pressure signals of the deployment positions of the sensor points in real time; The data processing and communication module realizes data caching, data processing and feature signal extraction by deploying distributed data processing and communication cabinets on site, and its two ends are respectively connected to the multi-channel data acquisition module and the operation status evaluation module, and is used to perform data processing and feature extraction on the vibration signal, pressure pulsation signal and pressure signal monitored and collected by the multi-channel data acquisition module, and send them to the operation status evaluation module; The operation status assessment module combines the operation condition data of the existing online monitoring system of the power station and the characteristics of the transmission of the data processing and communication module to carry out computational fluid dynamics analysis and construct a high-fidelity proxy model that describes the flow channel pressure of the turbine unit to carry out the unit operation status assessment.

2. A cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit according to claim 1, characterized in that: Also includes: The component measurement point positions of the multi-channel data acquisition module at least include: Upper and lower frames, volute, volute, draft tube and bolts (8); The sensors at least include a multi-channel vibration sensor, a multi-channel pressure pulsation sensor and a multi-channel pressure sensor; The monitoring signal of the sensor is amplified and collected by a corresponding multi-channel data acquisition card; The multi-channel vibration sensor and the multi-channel pressure pulsation sensor both have four channels; Four vibration sensors are arranged at four mutually perpendicular positions of the upper and lower frames of the turbine; Four-way pressure pulsation sensors are arranged at the draft tube and the volute inlet and outlet of the turbine; A multi-channel pressure sensor is arranged at the outlet of the tailwater pipe; a bolt (8) vibration sensor is arranged at the bolt (8) and the bolt (8) fastener (7); The pressure pulsation signal acquisition card in the multi-channel data acquisition module adopts the same-starting-point full-cycle sampling method, and the collected turbine pressure pulsation signal covers the abnormal operation state and turbine transition process of the turbine, that is, the characteristic signals under such working conditions as the drastic high-frequency signal changes of the pressure pulsation during the start-up and shutdown of the turbine, and the abnormal periodic pressure pulsation signal changes when the blades operate abnormally.

3. A cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit according to claim 1, characterized in that: Also includes: The data processing and communication module includes a plurality of on-site data processing and communication cabinets; The data processing communication module is connected with a signal acquisition card, a data buffer device and a data processing device via a signal cable; The data processing and communication module is connected in sequence to an on-site photoelectric converter, an on-site optical fiber distribution frame, a remote optical fiber distribution frame and a remote photoelectric converter through an optical fiber cable; the on-site photoelectric converter is connected to the data processing and communication module through a network cable; and the remote photoelectric converter is connected to the operation status evaluation module through a network cable; Each of the on-site data processing and communication cabinets is composed of a data buffer device and a data processing device, and is arranged at different locations according to the requirements of data collection; The data processing device in the data communication processing module processes the collected sensor signals, performs noise reduction and feature extraction on the fast variable pressure pulsation signal and vibration signal through wavelet transform and principal component analysis, and retains the original features of the slow variable pressure signal without filtering. The calculation formulas of wavelet transform and principal component analysis are as follows: Z=XW(2).

4. A cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit according to claim 1, characterized in that: Also includes: The operation status evaluation module is the cloud server of the cloud-edge collaboration device, and the cloud server is deployed in the computer room of the power station. The operation status evaluation module includes computational fluid dynamics, machine learning model training, turbine performance evaluation and evaluation result sending function modules; The computational fluid dynamics function combines the operating condition data transmitted by the existing online monitoring system of the power station and the characteristic signal output by the data processing and communication module to carry out computational fluid dynamics analysis, calculate the corresponding operating condition data, and prepare for the agent model simulation; The machine model learning and training function module uses the operating data of the fluid mechanics calculation to carry out agent model training. Based on the integrated model, the ability of the artificial neural network and the support vector machine is synergistically combined, so that the agent model can handle complex nonlinear relationships while minimizing the risk of overfitting and improving the ability to evaluate the operating status of the turbine.

5. A cloud-edge collaborative device for monitoring and evaluating the operating status of a hydropower unit according to claim 1, characterized in that: When the unit operation status assessment is carried out, the assessment is performed based on the splash water image of the turbine outlet, and specifically includes: Collecting the water turbine units to be evaluated that need to be evaluated for their running status, collecting splashing videos of the water turbine units to be evaluated, and obtaining a plurality of splashing video frame images of the water turbine units to be evaluated; Preprocessing the splashing video frame image of the water turbine set to be evaluated to obtain a water discharge image of the water turbine set to be evaluated and analyzed of the splashing video frame image of the water turbine set to be evaluated; Collecting the dry-wet balance value in the water outlet image of the water turbine unit to be evaluated and analyzed, and determining the operating state of the splash video frame image of the water turbine unit to be evaluated; Fault assessment statistics are performed on the operating status of multiple consecutive water discharge images of the water turbine unit to be evaluated and analyzed collected by the splashing video to obtain a fault assessment of the operating status of the water turbine unit to be evaluated.

Images