Generator state prediction method, device, equipment, medium and product
By combining digital twin technology with the acquisition and analysis of visual and acoustic field data, the problems of low detection efficiency and high risk of missed detection in traditional wind turbine generator sets have been solved, achieving efficient and low-miss-detection fault detection and diagnosis, and simplifying the operation and maintenance process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-22
- Publication Date
- 2026-04-07
AI Technical Summary
Traditional wind turbine testing methods are inefficient, have a high risk of missing detections, are complex to operate and maintain, and lack early diagnostic capabilities.
By employing digital twin technology, visual and acoustic data from inside the generator compartment are collected. Then, time-series prediction models and digital twin models are used to simulate fault conditions, and operational strategies are adjusted to optimize fault detection and diagnosis.
It achieves efficient fault detection and diagnosis with low missed detection, simplifies the operation and maintenance process, and improves detection accuracy and early diagnosis capabilities.
Smart Images
Figure CN121803418A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wind power generation technology, specifically to a generator state prediction method, device, equipment, medium, and product. Background Technology
[0002] Wind turbines are typically installed in harsh environments, such as offshore wind farms, high-altitude mountains, or arid deserts. Facing extreme weather and complex operating conditions, the generators must withstand the combined effects of high humidity, salt spray corrosion, and long-term vibration. These factors accelerate the aging and failure of equipment components. Traditional wind turbine inspection methods suffer from low efficiency and a high risk of missed inspections; vibration analysis and maintenance are complex; electrical parameter analysis and identification capabilities are insufficient, feature extraction is difficult, and early diagnostic capabilities are lacking. Summary of the Invention
[0003] This invention provides a generator condition detection method, device, equipment, and product based on digital twins, to solve the problems of low efficiency, high risk of missed detection, complex operation and maintenance, and lack of early diagnostic capabilities in traditional wind turbine generator detection methods.
[0004] In a first aspect, the present invention provides a generator state prediction method, comprising: Collect visual and acoustic data from inside the generator compartment; Visual data and sound field data are input into a pre-established time-series prediction model to obtain visual prediction time-series data and sound field prediction time-series data of the generator compartment within the target time period. The visual prediction time series data and the sound field prediction time series data are synchronized to the generator digital twin model. The generator digital twin model is used to simulate the fault state of the generator compartment. The generator digital twin model is built based on the generator basic structure data. Different operating strategies of the generator are simulated in the digital twin model of the generator to obtain the fault state optimization results corresponding to different strategies, and the operating strategy is adjusted according to the fault state optimization results.
[0005] Visual and acoustic field prediction time-series data are obtained by predicting future visual and acoustic field data. Based on this data and a digital twin model, future generator fault states are simulated. Then, the adjusted operating strategy is used to predict the adjusted fault state. The optimal operating strategy can be derived based on the fault state. Simultaneously, current generator fault states can be detected and diagnosed by capturing current visual and acoustic field data. Both visual and acoustic data are non-contact detection methods, resulting in low efficiency, low risk of missed detections, and relatively simple operation and maintenance.
[0006] In one alternative implementation, visual and acoustic data within the generator compartment are acquired, including: Collect raw visual data and raw sound field data; The original visual data and original sound field data are denoised separately to obtain the visual data and sound field data.
[0007] This embodiment illustrates the process of denoising data. The denoising operation improves the quality of the data, thereby increasing the accuracy of subsequent calculation results.
[0008] In one optional implementation, the original visual data includes the wavelengths of each pixel at different time frames. The original visual data is then denoised to obtain the final visual data, which includes: Construct spectral reflectance curves for different pixels based on the wavelength of each pixel in different time frames; The spectral reflectance curve of each pixel is decomposed into low-frequency data and high-frequency data; High-frequency data is denoised using a pre-established threshold function to obtain denoised high-frequency data. By combining low-frequency data and high-frequency denoised data, the spectral reflectance curves of different pixels are constructed using the wavelengths of each pixel at different time frames. Visual data is determined based on the spectral reflectance curves of different pixels.
[0009] A method for denoising visual data is presented, which improves the accuracy of the data and the accuracy of subsequent calculation results.
[0010] In one optional implementation, visual prediction time-series data and sound field prediction time-series data are synchronized to the generator digital twin model, and the generator digital twin model is used to simulate the fault state of the generator nacelle, including: Perform temporal and spatial consistency processing on visual prediction time-series data and sound field prediction time-series data; The processed visual prediction time series data and sound field prediction time series data are synchronized to the generator digital twin model, and the generator digital twin model is used to simulate the fault state of the generator compartment.
[0011] A method is presented to first align visual prediction time-series data and sound field prediction time-series data in terms of time and space, and then synchronize the aligned data to the generator digital twin model for fault simulation, thereby locating abnormal locations and making the fault state simulation results of the digital twin model more accurate.
[0012] In one optional implementation, different operating strategies of the generator are simulated in the digital twin model of the generator to obtain the fault state optimization results corresponding to different strategies, and the operating strategies are adjusted according to the fault state optimization results. The operating strategies include one or more of adjusting the speed, reducing the load, or dynamically adjusting the stress distribution of components.
[0013] Possible adjustment methods in the operation strategy are given. Different operation strategies are obtained according to different adjustment methods. By incorporating different operation strategies into the generator digital twin model, the fault state results under different adjustment methods can be obtained.
[0014] In one alternative implementation, it further includes: Determine the mapping relationship between each operating parameter in the operating strategy and the generator equipment; The generator equipment is adjusted according to the operating strategy and mapping relationship.
[0015] The mapping relationship between operating parameters and generator equipment is presented, and the process of adjusting the generator equipment according to the mapping relationship is described. This demonstrates the operational process for precisely adjusting the operating state of the generator equipment.
[0016] In a second aspect, the present invention provides a generator state prediction device, comprising: The data acquisition module is used to collect visual and acoustic data inside the generator compartment; The data prediction module is used to input the visual data and sound field data into a pre-established time-series prediction model to obtain the visual prediction time-series data and sound field prediction time-series data of the generator compartment within the target time period. The fault simulation module is used to synchronize the visual prediction time series data and the sound field prediction time series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator compartment. The strategy adjustment module is used to simulate different operating strategies of the generator in the generator digital twin model, obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results.
[0017] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the generator state prediction method of the first aspect or any corresponding embodiment described above.
[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the generator state prediction method of the first aspect or any corresponding embodiment thereof.
[0019] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the generator state prediction method of the first aspect or any corresponding embodiment described above. Attached Figure Description
[0020] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0021] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the first process of a generator state prediction method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a second process for a generator state prediction method according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the third process of a generator state prediction method according to an embodiment of the present invention; Figure 5 This is a structural block diagram of a generator state prediction device according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.
[0024] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0025] As an optional application scenario of this invention, such as Figure 1 As shown, the generator state prediction system may include at least one terminal device and at least one server. Figure 1 The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.
[0026] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.
[0027] Traditional wind turbine inspection methods suffer from low efficiency, high risk of missed detections, complex vibration analysis and maintenance, insufficient electrical parameter analysis and identification capabilities, difficulty in feature extraction, and a lack of early diagnostic capabilities. This invention provides a generator condition prediction method that assesses internal generator faults through both visual and auditory data. It predicts future visual and auditory data, constructs a generator twin model based on the predicted data, and achieves optimal results by adjusting different strategies and combining them with fault simulation results from the generator twin model.
[0028] According to an embodiment of the present invention, a generator state prediction method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.
[0029] This embodiment provides a generator state prediction method, which can be used on mobile terminals such as mobile phones and tablets. Figure 2 This is a flowchart of a generator state prediction method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Collect visual and acoustic data from inside the generator compartment.
[0030] Visual data refers to digital signal data obtained by capturing the deformation characteristics of key components and converting the deformation into digital signals. Visual data can reflect the operating status of equipment and accurately measure the micro-deformation and vibration values of the generator body caused by mechanical vibration, temperature changes, or load fluctuations during operation. Acoustic field data refers to abnormal acoustic field data generated during generator operation. Monitoring the changes in audio characteristics caused by overcurrent or overload during generator operation, or the acoustic field characteristics generated by local high-frequency discharge anomalies, can provide data support for subsequent diagnosis.
[0031] Step S202: Input the visual data and sound field data into the pre-established time-series prediction model to obtain the visual prediction time-series data and sound field prediction time-series data of the generator compartment within the target time period.
[0032] Visual prediction time-series data refers to future visual data predicted using a time-series prediction model, while sound field prediction time-series data refers to future sound field data predicted using a time-series prediction model. Both visual and sound field prediction time-series data are used in subsequent processes to simulate fault conditions using a digital twin model.
[0033] Step S203: Synchronize the visual prediction time series data and the sound field prediction time series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator compartment.
[0034] A digital twin model of a generator can organically combine visual prediction time-series data with acoustic field prediction time-series data. It can simulate generator fault states within a future prediction timeframe. In this process, the visual prediction time-series data provides the model with information reflecting the generator's operating status from a visual perspective, while the acoustic field prediction time-series data provides the model with information about abnormal generator operating sounds from an acoustic perspective. Step S204: Simulate different operating strategies of the generator in the digital twin model of the generator, obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results.
[0035] In a digital twin model of a generator, various operating strategies of the generator can be adjusted, and different fault state optimization results can be simulated based on the adjusted operating strategies. Based on the different fault state optimization results, the operating strategy corresponding to the optimal fault state optimization result can be determined, and the actual generator operating strategy can be adjusted accordingly.
[0036] The generator state prediction method provided in this embodiment obtains visual prediction time-series data and sound field prediction time-series data by predicting future visual and sound field data. Based on these data and a digital twin model, it simulates future generator fault states. Then, it predicts the adjusted fault state by adjusting the operating strategy. The optimal operating strategy can be obtained based on the fault state. Simultaneously, by capturing current visual and sound field data, the current generator fault state can be detected and diagnosed. Both visual and acoustic data are non-contact detection methods, resulting in low efficiency, low risk of missed detections, and relatively simple operation and maintenance.
[0037] This embodiment provides a generator state prediction method. Step S201, which involves collecting visual and acoustic field data from inside the generator compartment, includes the following steps: Step a1: Collect raw visual data and raw sound field data.
[0038] Raw visual data and raw sound field data are data obtained directly from the acquisition equipment without any data processing.
[0039] For example, raw visual data can be acquired using a hyperspectral camera. A high-resolution, wide-spectral-range industrial hyperspectral camera, covering the 400-1000 nm visible and near-infrared bands, is suitable for capturing minute spectral changes on metal and composite material surfaces. A high-precision industrial lens is provided to meet the needs of measuring the minute deformations of target components. A uniformly illuminated multi-band LED light source is configured to cover the working range of the hyperspectral camera, ensuring the spectral quality of the image. The light source must be stable to avoid the influence of light intensity fluctuations on the detection results. The camera is fixed above key components of the generator, and the camera's focal length and angle are adjusted to ensure optimal field-of-view coverage of the target area. The hyperspectral camera is calibrated using a standard reference board to eliminate the influence of instrument and ambient light on the data. This device can capture information in hundreds of bands within the visible, near-infrared, and ultraviolet spectral ranges, accurately capturing minute changes in physical properties, such as surface deformation and texture variations. It is suitable for operating equipment without interfering with its normal operation. It can acquire hyperspectral images of the target surface in real time, recording its spatial and spectral information. Each image consists of hundreds of spectral bands, containing the reflection intensity of the target surface in each band. Hyperspectral data of the target component is acquired periodically to construct time-series images, recording the dynamic changes in deformation over time. A stable sampling frequency is maintained, acquiring a hyperspectral image once per second to capture characteristics of rapid vibrations or slow deformations.
[0040] Raw sound field data can be acquired by placing high-sensitivity microphones or acoustic sensors near the generator, with monitoring points selected close to critical components such as windings or bearings. Ensure the sensor installation location avoids mechanical noise interference and covers areas prone to acoustic anomalies. The sensors acquire real-time sound signals from the generator during operation, including normal operating audio and potential abnormal audio. A high sampling rate of 48kHz ensures accurate capture of high-frequency signals from partial discharge.
[0041] Step a2: Denoise the original visual data and the original sound field data respectively to obtain visual data and sound field data.
[0042] Since some unwanted interference data is collected during the process of acquiring raw visual and acoustic data, denoising the data and removing interference data can yield cleaner and higher-quality visual and acoustic data.
[0043] For example, denoising the original sound field data can be performed using beamforming noise suppression algorithms to remove background noise, and dynamic range compression techniques can be used to improve signal discernibility. Fast Fourier Transform can be used to analyze the frequency distribution of the sound signal, and after extracting abnormal spectral features, key features such as time-frequency characteristics, harmonic distortion rate, and frequency band energy ratio can be extracted from the sound signal.
[0044] The generator state prediction method provided in this embodiment includes a data denoising process. The denoising operation improves the data quality, thereby enhancing the accuracy of subsequent calculation results.
[0045] This embodiment provides a generator state prediction method. Figure 3 This is a flowchart of a generator state prediction method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps: Step S301: Collect visual and acoustic data from inside the generator compartment.
[0046] Specifically, step S301 includes: Step S3011: Construct spectral reflectance curves for different pixels based on the wavelengths of each pixel in different time frames.
[0047] The spectral reflectance curve of a pixel is a continuous function curve describing the proportion of reflected energy to incident energy for a single pixel in a digital image under electromagnetic wave illumination of different wavelengths. It is used to establish the correspondence between "wavelength" and "reflectance".
[0048] Step S3012: Decompose the spectral reflectance curve of each pixel into low-frequency data and high-frequency data.
[0049] Low-frequency data refers to the slowly changing, smooth, and continuous portion of the corresponding curve, which can retain the main structural information; high-frequency data refers to the rapidly fluctuating portion of the corresponding curve, which is mostly noise components. By separating low-frequency and high-frequency data, denoising can be achieved more accurately, providing data support for subsequent calculations.
[0050] For example, multi-scale wavelet decomposition is used to decompose low-frequency and high-frequency data. The following method can be used for calculation, with time-series hyperspectral data as input: ,in, For pixel coordinates, For wavelength, This is a time frame. It will be used later. Spectral reflectance curve for each pixel Or, the image grayscale signal is subjected to multi-scale wavelet decomposition, where, Represents the approximation coefficients of the Jth layer; This represents the approximation coefficient for the j-th layer.
[0051] Step S3013: Use a pre-established threshold function to denoise the high-frequency data to obtain high-frequency denoised data.
[0052] For example, the following threshold function can be used to denoise high-frequency data:
[0053] in The VisuShrink soft thresholding algorithm is used. This is the high-frequency data after noise reduction.
[0054] Step S3014: Combine low-frequency data and high-frequency denoised data to obtain the spectral reflectance curves of different pixels at different time frames.
[0055] After decomposing the spectral reflectance curve into low-frequency and high-frequency data, the noisy high-frequency data is denoised, and then the denoised high-frequency data is combined with the low-frequency data to obtain the spectral reflectance curves of different pixels at different time frames.
[0056] For example, high-frequency data can be recombinated with low-frequency data using the following formula:
[0057] Retained low-frequency data, The high-frequency data after denoising is shown. This is the spectral reflectance curve after noise reduction.
[0058] Step S3015: Determine visual data based on the spectral reflectance curves of different pixels.
[0059] Data from the spectral reflectance curves of different pixels are extracted and converted into visual data that can be calculated in subsequent processes.
[0060] For example, in addition to wavelet denoising algorithms, the denoising process can also use whiteboard correction or reference spectra to standardize the image, eliminating errors caused by uneven light sources or differences in sensor sensitivity. Feature information on surface reflection changes is extracted, and spectral changes at different time points are compared to identify spectral shifts or intensity fluctuations caused by deformation.
[0061] Step S302: Input the visual data and sound field data into the pre-established temporal prediction model to obtain the visual and sound field prediction time-series data of the generator nacelle within the target time period. For details, please refer to... Figure 2 Step S202 of the illustrated embodiment will not be described again here.
[0062] Step S303: Synchronize the visual prediction time-series data and the sound field prediction time-series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator nacelle. For details, please refer to... Figure 2 Step S203 of the illustrated embodiment will not be described again here.
[0063] Step S304: Simulate different operating strategies of the generator in the generator digital twin model to obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results. For details, please refer to [link to details]. Figure 2 Step S204 of the illustrated embodiment will not be described again here.
[0064] The generator state prediction method provided in this embodiment offers a method for denoising visual data, thereby improving data accuracy and the accuracy of subsequent calculation results.
[0065] This embodiment provides a generator state prediction method. Figure 4 This is a flowchart of a generator state prediction method according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps: Step S401: Acquire visual and acoustic data from inside the generator nacelle. For details, please refer to [link to relevant documentation]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.
[0066] Step S402: Input the visual data and sound field data into the pre-established temporal prediction model to obtain the visual and sound field prediction time-series data of the generator nacelle within the target time period. For details, please refer to... Figure 2Step S202 of the illustrated embodiment will not be described again here.
[0067] Step S403: Synchronize the visual prediction time series data and the sound field prediction time series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator compartment.
[0068] Specifically, step S403 includes: Step S4031: Perform temporal consistency processing and spatial consistency processing on the visual prediction time series data and the sound field prediction time series data.
[0069] Time consistency processing addresses the inconsistency in the temporal dimension between visual prediction time-series data and sound field prediction time-series data. Through standardization, alignment, and other operations, it ensures data consistency at the temporal granularity. Time consistency processing allows for the identification of visual and sound field prediction time-series data at any point within the measurement time range. For example, time consistency processing can be performed on visual and sound field prediction time-series data using timestamps.
[0070] Spatial consistency processing solves the problem of inconsistent spatial dimensions between visual prediction time series data and sound field prediction time series data. Through spatial consistency processing, the visual prediction time series data and sound field prediction time series data at specific locations can be obtained.
[0071] For example, after time-consistency processing and spatial consistency processing, a mapping relationship between sound signals and structural stress / vibration can be established. During generator operation, mechanical stress and deformation cause structural vibration; structural vibration excites sound waves, which are captured by sensors after propagating in the air. Therefore, a mapping relationship between mechanical stress and deformation and sound field data can be obtained.
[0072] Step S4032: Synchronize the processed visual prediction time series data and sound field prediction time series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator compartment.
[0073] The processed visual prediction time series data and acoustic field prediction time series data are input into the finite element model or multiphysics model of the digital twin to simulate the fault state of the generator nacelle.
[0074] Step S404: Simulate different operating strategies of the generator in the generator digital twin model to obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results. For details, please refer to [link to details]. Figure 2 Step S204 of the illustrated embodiment will not be described again here.
[0075] The generator state prediction method provided in this embodiment first aligns the visual prediction time series data and the sound field prediction time series data in terms of time and space dimensions, and then synchronizes the aligned data to the generator digital twin model for fault simulation, thereby locating abnormal locations and making the fault state simulation results of the digital twin model more accurate.
[0076] This embodiment provides a generator state prediction method. In step S204, different operating strategies of the generator are simulated in the generator digital twin model to obtain the fault state optimization results corresponding to different strategies. The operating strategies are adjusted according to the fault state optimization results. The operating strategies include one or more of adjusting the speed, reducing the load, or dynamically adjusting the stress distribution of components.
[0077] Wind turbines achieve safe, efficient, and long-life operation by dynamically adjusting operating parameters or structural states through real-time monitoring of operating conditions such as wind speed, rotational speed, load, and component stress. Specific strategies can include adjusting rotational speed, reducing load, or dynamically adjusting component stress distribution, or a combination of these strategies. Adjusting rotational speed refers to regulating the generator rotor speed to match the real-time wind speed. Reducing load refers to decreasing the electrical or mechanical load on the generator, alleviating the load on the equipment, and preventing component fatigue damage. Dynamically adjusting component stress distribution involves adjusting the attitude or stress state of key wind turbine components to ensure uniform stress distribution within the components or the entire structure, avoiding premature damage caused by localized stress concentration.
[0078] The generator state prediction method provided in this embodiment gives possible adjustment methods in the operation strategy. Different operation strategies are obtained according to different adjustment methods. By bringing different operation strategies into the generator digital twin model, the fault state results under different adjustment methods can be obtained.
[0079] This embodiment provides a generator state prediction method. When performing step S204 above, the method further includes the following steps: Step b1: Determine the mapping relationship between each operating parameter in the operating strategy and the generator equipment.
[0080] Operating parameters refer to quantitative indicators that describe the operating status, working conditions, equipment load, and structural stress of the wind turbine. Mapping relationships describe the direct correlation between various operating parameters and the various components within the generator. They can reflect which parameters can be adjusted to control the physical structure and function of the equipment, and can also directly reflect the working status of the equipment components through changes in the values of operating parameters.
[0081] Step b2: Adjust the generator equipment according to the operating strategy and mapping relationship.
[0082] The generator equipment can be adjusted according to the mapping relationship between the operating parameters and the generator equipment in the above steps.
[0083] The adjustment process can begin by clarifying the operational goals to be achieved, then by determining the core parameters that need to be adjusted through operational strategies, and finally by driving specific equipment components to perform actions based on the mapping relationship between operational parameters and generator equipment, thereby ultimately regulating the generator's operating state to the optimal range.
[0084] For example, reverse decision-making adjustments can also be made through the simulation and optimization module, and the operating strategy can be adjusted: if a high risk is predicted, the speed or power is reduced; if stress concentration is identified, the rotor balance or the stiffness of the support structure is adjusted; digital twin feedback: simulate the energy distribution and stress response under different adjustment strategies; compare the vibration amplitude, sound energy density and temperature rise of the equipment under different strategies to select the optimal adjustment strategy.
[0085] This embodiment provides a generator state prediction method that outlines the mapping relationship between operating parameters and generator equipment, and details the process of adjusting the generator equipment based on this mapping relationship. This achieves a precise adjustment process for the operating state of the generator equipment.
[0086] This embodiment also provides a generator state prediction device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.
[0087] This embodiment provides a generator state prediction device, such as... Figure 5 As shown, it includes: The data acquisition module 501 is used to acquire visual and acoustic data inside the generator compartment.
[0088] The data prediction module 502 is used to input the visual data and sound field data into a pre-established time-series prediction model to obtain the visual prediction time-series data and sound field prediction time-series data of the generator compartment within the target time period.
[0089] The fault simulation module 503 is used to synchronize the visual prediction time series data and the sound field prediction time series data to the generator digital twin model, and use the generator digital twin model to simulate the fault state of the generator compartment.
[0090] The strategy adjustment module 504 is used to simulate different operating strategies of the generator in the generator digital twin model, obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results.
[0091] The generator state prediction device provided in this embodiment of the invention can execute the generator state prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.
[0092] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.
[0093] The following is a detailed reference. Figure 6 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from memory 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the electronic device. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface *05 is also connected to the bus *04.
[0094] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.
[0095] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a memory 608, or installed from a ROM 602. When the computer program is executed by the processor 601, it performs the functions defined in the generator state prediction method of the embodiments of the present invention.
[0096] Figure 6The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.
[0097] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the generator state prediction method shown in the above embodiments is implemented.
[0098] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.
[0099] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.
Claims
1. A generator state prediction method, characterized in that, The method includes: Collect visual and acoustic data from inside the generator compartment; The visual data and sound field data are input into a pre-established time-series prediction model to obtain the visual prediction time-series data and sound field prediction time-series data of the generator compartment within the target time period. The visual prediction time series data and the sound field prediction time series data are synchronized to the generator digital twin model. The generator digital twin model is used to simulate the fault state of the generator compartment. The generator digital twin model is constructed based on the generator basic structure data. Different operating strategies of the generator are simulated in the digital twin model of the generator to obtain the fault state optimization results corresponding to different strategies, and the operating strategy is adjusted according to the fault state optimization results.
2. The method according to claim 1, characterized in that, The acquisition of visual and acoustic data within the generator compartment includes: Collect raw visual data and raw sound field data; The original visual data and original sound field data are denoised separately to obtain the visual data and sound field data.
3. The method according to claim 2, characterized in that, The original visual data includes the wavelengths of each pixel at different time frames. Noise removal processing is performed on the original visual data to obtain the final visual data, which includes: Construct spectral reflectance curves for different pixels based on the wavelength of each pixel in different time frames; The spectral reflectance curve of each pixel is decomposed into low-frequency data and high-frequency data; The high-frequency data is denoised using a pre-established threshold function to obtain high-frequency denoised data. By combining the low-frequency data and the high-frequency denoised data, the spectral reflectance curves of different pixels are constructed using the wavelengths of each pixel at different time frames. The visual data is determined based on the spectral reflectance curves of the different pixels.
4. The method according to claim 1, characterized in that, The visual prediction time-series data and sound field prediction time-series data are synchronized to the generator digital twin model, and the generator digital twin model is used to simulate the fault state of the generator compartment, including: The visual prediction time series data and the sound field prediction time series data are subjected to temporal consistency processing and spatial consistency processing. The processed visual prediction time-series data and sound field prediction time-series data are synchronized to the generator digital twin model, and the generator digital twin model is used to simulate the fault state of the generator compartment.
5. The method according to claim 1, characterized in that, The generator is simulated in the digital twin model of the generator to obtain the fault state optimization results corresponding to the different strategies, and the operating strategy is adjusted according to the fault state optimization results. The operating strategy includes one or more of adjusting the speed, reducing the load, or dynamically adjusting the stress distribution of the components.
6. The method according to claim 5, characterized in that, The method further includes: Determine the mapping relationship between each operating parameter in the operating strategy and the generator equipment; The generator equipment is adjusted according to the operating strategy and the mapping relationship.
7. A generator state prediction method and apparatus, characterized in that, The device includes: The data acquisition module is used to collect visual and acoustic data inside the generator compartment; The timing prediction module is used to input the visual data and sound field data into a pre-established timing prediction model to obtain the visual prediction timing data and sound field prediction timing data of the generator nacelle within the target time period. The fault simulation module is used to synchronize the visual prediction time series data and the sound field prediction time series data to the generator digital twin model, and to simulate the fault state of the generator compartment using the generator digital twin model, which is constructed based on the generator basic structure data. The strategy optimization module is used to simulate different operating strategies of the generator in the generator digital twin model, obtain the fault state optimization results corresponding to different strategies, and adjust the operating strategy according to the fault state optimization results.
8. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform a generator state prediction method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute a generator state prediction method according to any one of claims 1 to 6.
10. A computer program product, characterized in that, Includes computer instructions for causing a computer to execute a generator state prediction method according to any one of claims 1 to 6.