Thermal power fan vibration quantitative monitoring and fault early warning method and related device

Abstract

The invention discloses a thermal power fan vibration quantitative monitoring and fault early warning method and a related device. The method comprises the following steps: acquiring vibration and operation parameters of a thermal power fan; preprocessing the vibration and operation parameters of the thermal power fan to obtain processed data; performing feature extraction on the preprocessed data, screening the extracted features, and constructing a feature matrix by using the screened features; the feature matrix is input into a trained random forest quantitative model, fault early warning is carried out according to the output result of the random forest quantitative model, the method and the related device can meet the high-precision monitoring requirement of the thermal power generating unit, and meanwhile the method and the related device have the advantages of being high in sensitivity and accurate in fault early warning.

Description

Technical Field

[0001] This invention belongs to the field of thermal power equipment condition monitoring technology, and relates to a method and related device for quantitative monitoring and fault early warning of thermal power fan vibration. Background Technology

[0002] Thermal power plant fans (including induced draft fans, forced draft fans, primary air fans, etc.) are key auxiliary equipment in thermal power plants, and their operational stability directly affects the safe and economical operation of the unit. Vibration is a core indicator reflecting the operating status of fans. During long-term operation, fans are prone to abnormal vibration due to problems such as rotor imbalance, coupling misalignment, bearing wear, and blade dust accumulation and corrosion. If these issues are not monitored and warned in a timely manner, they may lead to serious consequences such as equipment shutdown, component damage, or even unplanned unit outages.

[0003] Existing methods for monitoring vibration in thermal power wind turbines mainly include traditional spectrum analysis, threshold judgment methods, and simple machine learning methods: 1. Traditional spectrum analysis methods (such as FFT analysis) identify faults by extracting the frequency characteristics of vibration signals, but they are greatly affected by fluctuations in the operating conditions of the wind turbine (such as load changes and fluctuations in medium parameters), making it difficult to quantitatively assess the severity of the fault, and they are not sensitive enough to identify early minor faults. 2. The threshold judgment method triggers alarms based on preset vibration amplitude thresholds, which can only achieve qualitative judgment and cannot reflect the fault development trend. In addition, the threshold setting depends on experience and has poor generalization ability. 3. Existing simple machine learning methods (such as single decision trees and support vector machines) attempt quantitative analysis, but they suffer from problems such as inaccurate feature selection, weak model anti-interference ability, and low efficiency in processing high-dimensional monitoring data. The mean square error (MSE) of quantitative prediction is large, and the coefficient of determination (R²) is mostly below 0.85, which makes it difficult to meet the high-precision monitoring requirements of thermal power wind turbines. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and related device for quantitative monitoring and fault early warning of thermal power fan vibration. This method and related device can meet the high-precision monitoring requirements of thermal power units and have the characteristics of high sensitivity and accurate fault early warning.

[0005] To achieve the above objectives, this invention discloses a method for quantitative monitoring and fault early warning of vibration in thermal power wind turbines, comprising: Obtain the vibration and operating parameters of thermal power fans; The vibration and operating parameters of the thermal power fan are preprocessed to obtain the processed data; Feature extraction is performed on the preprocessed data, the extracted features are filtered, and a feature matrix is ​​constructed using the filtered features. The feature matrix is ​​input into the trained random forest quantitative model, and fault warning is performed based on the output of the random forest quantitative model.

[0006] Furthermore, the vibration and operating parameters of the thermal power fan include the acceleration signals of the front and rear bearings in the bearing housing, the radial vibration velocity signal of the coupling, the circumferential vibration displacement signal at the outlet of the casing, and the fan speed, unit load, inlet and outlet air pressure and air volume, medium temperature and bearing temperature.

[0007] Furthermore, the process of preprocessing the vibration and operating parameters of the thermal power fan is as follows: The vibration and operating parameters of the thermal power fan are subjected to baseline correction, noise removal and normalization.

[0008] Furthermore, the process of extracting features from the preprocessed data, filtering the extracted features, and constructing a feature matrix using the filtered features is as follows: Feature extraction is performed on the preprocessed data. The extracted features are then screened using the feature importance evaluation mechanism of the random forest algorithm, and a feature matrix is ​​constructed using the screened features.

[0009] Furthermore, the extracted features include time-domain features, frequency-domain features, and operating parameter features.

[0010] Furthermore, the output of the random forest quantitative model includes quantitative values ​​of vibration intensity, probability of fault type, and coefficient of fault severity.

[0011] Furthermore, the process of performing fault early warning based on the output results of the random forest quantitative model is as follows: If the vibration intensity quantitative value is ≥4, or the fault severity coefficient is 0.3-0.5, a prompt alarm will be triggered and pushed to the operation and maintenance terminal. If the vibration intensity quantitative value is ≥6, or the fault severity coefficient is 0.5-0.8, an important alarm will be triggered, and the power plant's DCS system will be linked to display fault information. If the vibration intensity quantitative value is ≥8, or the fault severity coefficient is ≥0.8, an emergency alarm will be triggered, and it is recommended that the unit reduce its load or be shut down for maintenance.

[0012] This invention discloses a quantitative monitoring and fault early warning system for vibration of thermal power wind turbines, comprising: The acquisition module is used to acquire the vibration and operating parameters of thermal power fans; The preprocessing module is used to preprocess the vibration and operating parameters of the thermal power fan to obtain the processed data. The extraction module is used to extract features from the preprocessed data, filter the extracted features, and construct a feature matrix using the filtered features. The early warning module is used to input the feature matrix into the trained random forest quantitative model and to provide fault warnings based on the output of the random forest quantitative model.

[0013] This invention discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method for quantitative monitoring and fault early warning of thermal power wind turbine vibration.

[0014] The present invention discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method for quantitative monitoring and fault early warning of thermal power wind turbine vibration.

[0015] The present invention has the following beneficial effects: In specific operation, the vibration quantitative monitoring and fault early warning method and related device of the thermal power wind turbine described in this invention preprocesses the vibration and operating parameters of the thermal power wind turbine to obtain processed data. Feature extraction is performed on the preprocessed data, the extracted features are screened, and a feature matrix is ​​constructed using the screened features to meet the high-precision monitoring requirements of thermal power units. In addition, this invention is based on a random forest quantitative model for early warning, which has the characteristics of high sensitivity and accurate fault early warning, and is extremely practical. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] 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, not all, of the embodiments of the present invention. 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.

[0019] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0020] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0021] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0022] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0023] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0024] 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, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0026] Example 1 refer to Figure 1 The method for quantitative monitoring and fault early warning of thermal power fan vibration according to the present invention includes the following steps: 1) Obtain the vibration and operating parameters of thermal power fans; The specific process of step 1) is as follows: Sensors are deployed at key monitoring points of thermal power wind turbines; Piezoelectric accelerometers are installed in the bearing housings (front and rear bearings) to collect vibration acceleration signals in the X, Y, and Z directions. The sampling frequency is set to 50kHz to ensure coverage of the fan rotor's fundamental frequency and harmonic frequencies. A vibration velocity sensor is installed at the coupling to collect radial vibration velocity signals; A vibration displacement sensor is installed at the outlet of the casing to collect axial vibration displacement signals; Operating parameters, including fan speed, unit load, inlet and outlet air pressure and air volume, medium temperature and bearing temperature, are collected synchronously through the PLC interface of the fan control system.

[0027] All sensor data is transmitted to the monitoring host via the data acquisition card, forming a multi-dimensional raw monitoring dataset. The data sampling interval is set to 2 seconds to ensure the real-time nature and integrity of the data.

[0028] 2) Preprocess the collected data; The raw data contains useless information such as baseline drift, environmental noise, and electromagnetic interference, and requires three preprocessing steps to improve data quality: Baseline correction: A fifth-order polynomial fitting method is used to correct the baseline drift of the vibration signal, eliminating trend interference caused by sensor installation errors and temperature drift. The fitting formula is as follows:

[0029] in, The coefficients are polynomials, which are solved using the least squares method; Noise Removal: The Daubechies-4 wavelet basis function is used to perform a three-level wavelet transform on the corrected signal, which decomposes the signal into low-frequency approximate components and high-frequency detail components. The high-frequency detail components are processed by soft thresholding (the threshold is set to 0.02 times the maximum value of the signal). Then, the signal is reconstructed by inverse wavelet transform to achieve noise removal. Normalization: The Min-Max normalization method is used to scale the vibration signal and operating parameter data to the [0,1] interval to eliminate the dimensional differences between different parameters. The normalization formula is as follows:

[0030] in, x This is the original data. , These are the minimum and maximum values ​​of the dataset, respectively. x' This is the normalized data.

[0031] 3) Construct the feature matrix; Extract multi-dimensional features from the preprocessed data to form an initial feature set: Time-domain features: Extract 12 time-domain statistical features of the vibration signal, including peak value, mean, variance, kurtosis, skewness, peak factor, and impulse factor; Frequency domain features: Perform FFT transformation on the vibration signal to extract eight frequency domain features, including fundamental frequency amplitude, second harmonic amplitude, third harmonic amplitude, and total harmonic distortion rate. Operating parameter characteristics: Six key operating parameters, including fan speed, unit load, and bearing temperature, are selected as characteristics.

[0032] Key features are selected using the feature importance evaluation mechanism of the random forest algorithm. The reduction in mean squared error (MSE) for each feature across all decision tree nodes is calculated using the following formula:

[0033] in, This represents the mean square error of the parent node. These represent the number of samples in the left and right child nodes, respectively.N The number of samples in the parent node. These are the mean square errors of the left and right child nodes, respectively.

[0034] The MSE reduction of all features is normalized, and the top 40% of high-importance features with the highest normalized scores are selected to form the final feature matrix, thereby reducing the computational complexity of the model and improving its generalization ability.

[0035] 4) Train a quantitative random forest model; Data set division: Collect historical normal operation data of wind turbines, monitoring data of different fault types (imbalance, misalignment, bearing wear, blade dust accumulation) and different fault severity, construct a sample library (sample size ≥ 1000 groups), and randomly divide it into training set and test set in a ratio of 8:2; Model building: A quantitative random forest model was built based on the scikit-learn library in Python. The number of decision trees was set to 150, the maximum depth was 20 layers, and the number of features randomly selected when splitting a node was 1 / 2 of the total number of features. The Bootstrap sampling method was used to generate training subsets for each decision tree. Model training and validation: Train the model using the training set and validate its performance using the test set, employing the coefficient of determination (R²). 2 The mean squared error (MSE) and mean relative error (MRE) are used as evaluation metrics to iteratively optimize the hyperparameters until the model meets the following requirements: ≥0.92, MSE ≤0.05, MRE ≤0.15.

[0036] 5) Quantitative analysis and fault early warning.

[0037] Real-time quantitative analysis: The pre-processed monitoring data is input into the trained random forest model, and three quantitative results are output: vibration intensity quantitative value (0-10, corresponding to the vibration level of GB / T6075.3-2019), fault type probability distribution (e.g., unbalanced fault probability 85%), and fault severity coefficient (0-1, 0 for normal and 1 for extreme fault state). Tiered early warning: Preset three-level early warning thresholds: Level 1 warning (minor fault): The vibration intensity quantitative value is ≥4, or the fault severity coefficient is 0.3-0.5, triggering a prompt alarm and pushing it to the operation and maintenance terminal; Level 2 warning (moderate fault): The vibration intensity quantitative value is ≥6, or the fault severity coefficient is 0.5-0.8, triggering an important alarm and linking the power plant's DCS system to display fault information; Level 3 warning (severe fault): The vibration intensity quantitative value is ≥8, or the fault severity coefficient is ≥0.8, triggering an emergency alarm. It is recommended that the unit reduce its load or be shut down for maintenance.

[0038] Example 2 Taking the induced draft fan of a thermal power unit as the monitoring object, this paper details the training, validation, and field testing of a quantitative random forest model. The specific steps are as follows: Sample library construction: Historical monitoring data of the induced draft fan for the past three years were collected to construct a sample library containing 1000 valid samples, ensuring that the samples cover all operating states of the fan. This includes 300 sets of normal operation data, collected under fault-free and stable load conditions; and 700 sets of fault data, covering four typical fault types: 200 sets of rotor imbalance faults, subdivided into 100 mild, 60 moderate, and 40 severe faults, based on vibration acceleration amplitude and fault impact; 200 sets of coupling misalignment faults; 150 sets of bearing wear faults; and 150 sets of blade dust accumulation faults. All samples are labeled with corresponding "vibration intensity level," "fault type code," and "fault severity coefficient," which serve as target variables for model training.

[0039] Data preprocessing: The raw data of 1000 samples were preprocessed according to step 2 in the claims—baseline drift was corrected by 5th order polynomial fitting to eliminate the effects of temperature drift and installation error; then, 3-level wavelet transform was performed using Daubechies-4 wavelet basis functions, and the high-frequency noise components were processed with a soft threshold (0.02 times the maximum signal value) to reconstruct the signal; finally, all data were scaled to the [0,1] interval by Min-Max normalization to eliminate dimensional differences.

[0040] Feature extraction and selection: 26 initial features were extracted from the preprocessed data, including 12 time-domain features (peak value, mean, variance, kurtosis, skewness, etc.), 8 frequency-domain features (fundamental frequency amplitude, second harmonic amplitude, third harmonic amplitude, total harmonic distortion, etc., obtained through FFT transformation), and 6 operating parameter features (wind turbine speed, unit load, bearing temperature, etc.). A feature importance evaluation mechanism using the random forest algorithm (calculating and normalizing the reduction in mean squared error of each feature when splitting across all decision tree nodes) was employed to select the top 40% of key features with the highest normalized scores (including peak vibration acceleration, fundamental frequency amplitude, second harmonic amplitude, wind turbine speed, bearing temperature, etc.), forming the final feature matrix. This approach reduces model computational complexity while retaining core discriminative information.

[0041] Dataset splitting: The 1000 samples were randomly divided into a training set (800 sets) and a test set (200 sets) in an 8:2 ratio to ensure that the distribution of fault types and severity was consistent between the two sets of data.

[0042] Hyperparameter configuration: A random forest quantitative model was built based on the scikit-learn library. The grid search method was used to optimize the hyperparameters. The final settings were: 150 decision trees, 20 layers at maximum depth, half of the total number of features randomly selected when splitting a node, and 500 iterations of the training set. The Bootstrap sampling method was used to generate an independent training subset for each decision tree to avoid overfitting.

[0043] Model Validation: After training the model using the training set, its performance was validated using the test set. The coefficient of determination (R²), mean squared error (MSE), and mean relative error (MRE) were used as evaluation metrics. The validation results are as follows: R² = 0.94, MSE = 0.032, and MRE = 0.12, all meeting the preset standards (R² ≥ 0.92, MSE ≤ 0.05, MRE ≤ 0.15), indicating that the model possesses high-precision quantitative analysis capabilities.

[0044] On-site testing: The trained model was deployed to the monitoring host and connected to the induced draft fan for real-time data monitoring. During one unit operation, the model captured a vibration acceleration of 0.8g in the X direction of the front bearing housing (a characteristic of early rotor imbalance faults). After preprocessing and feature extraction, the model was input, and the output fault type was "rotor imbalance" (probability 89%), fault severity coefficient 0.35, and vibration intensity quantitative value 4.2, triggering a level one warning (consistent with the threshold range of minor faults in the claims). Based on the warning information and accompanying handling suggestions, maintenance personnel shut down the unit for dynamic balancing experiments, successfully eliminating the fault and preventing damage to the fan bearings and unplanned unit shutdowns caused by the fault escalating.

[0045] This invention employs a dual-channel early warning output method combining local visualization and remote push notifications to ensure that early warning information is promptly viewed by operations and maintenance personnel. The specific implementation is as follows: Local output: The monitoring host display screen dynamically displays three core pieces of information in real time: quantitative value of vibration intensity (0-10, corresponding to vibration level in GB / T6075.3-2019), fault type (such as "rotor imbalance" or "coupling misalignment"), and fault severity coefficient (0-1). It is also equipped with four levels of status indicator lights: green indicates normal operation (vibration intensity <4 and severity coefficient <0.3), yellow corresponds to level one warning (minor fault), orange corresponds to level two warning (moderate fault), and red corresponds to level three warning (severe fault). On-site maintenance personnel can quickly judge the equipment status through the indicator lights without having to consult detailed data.

[0046] Remote output: Early warning information is synchronously pushed to the power plant's DCS system and maintenance personnel's mobile terminals (mobile APP) via Ethernet interface and Modbus TCP protocol. The pushed content includes: equipment number, fault occurrence time, quantitative value of vibration intensity, fault type and probability, fault severity coefficient, and targeted handling suggestions (such as "shutdown and dynamic balancing inspection" for level one warning, and "load reduction and bearing maintenance" for level two warnings), ensuring that remote maintenance personnel accurately grasp the equipment status and promptly formulate maintenance plans.

[0047] The beneficial effects of this embodiment are as follows: high quantitative accuracy, with the model achieving a quantitative prediction R² of 0.94 and an MSE of only 0.032 for vibration intensity and fault severity, reducing the error by more than 45% compared to traditional methods, and accurately identifying early minor faults (such as rotor imbalance faults with vibration acceleration of 0.8g); strong anti-interference capability: through wavelet transform noise reduction and random forest ensemble learning characteristics, it can effectively detect unit load fluctuations. Under 25% of the operating conditions, the model prediction error fluctuates. 5% accuracy, unaffected by environmental noise and changes in operating conditions; good early warning timeliness: early warning of faults 3-7 days in advance, more than 60% earlier than traditional monitoring methods, allowing maintenance personnel sufficient time to handle the situation; strong practicality: universal and easily accessible hardware configuration, standardized model training process, directly adaptable to various thermal power fans such as induced draft fans, forced draft fans, and primary air fans, without the need for repeated modeling for individual devices.

[0048] Example 3 The thermal power fan vibration quantitative monitoring and fault early warning system of the present invention includes: The acquisition module is used to acquire the vibration and operating parameters of thermal power fans; The preprocessing module is used to preprocess the vibration and operating parameters of the thermal power fan to obtain the processed data. The extraction module is used to extract features from the preprocessed data, filter the extracted features, and construct a feature matrix using the filtered features. The early warning module is used to input the feature matrix into the trained random forest quantitative model and to provide fault warnings based on the output of the random forest quantitative model.

[0049] The module division in this embodiment is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in each embodiment of this application can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0050] Example 4 A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of a method for quantitative monitoring and fault early warning of vibration in thermal power wind turbines. For example, the method includes: acquiring vibration and operating parameters of the thermal power wind turbine; preprocessing the vibration and operating parameters to obtain processed data; extracting features from the preprocessed data; filtering the extracted features and constructing a feature matrix using the filtered features; inputting the feature matrix into a trained random forest quantitative model; and providing fault early warning based on the output of the random forest quantitative model. The memory may include main memory, such as high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device. The processor, network interface, and memory are interconnected via an internal bus, which may be an industry standard architecture bus, a peripheral component interconnection standard bus, an extended industry standard architecture bus, etc. The bus may be classified as an address bus, data bus, control bus, etc. The memory stores the program; specifically, the program may include program code, which includes computer operation instructions. The memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0051] Example 5 A computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of a method for quantitative monitoring and fault early warning of vibration in thermal power wind turbines. For example, the method includes: acquiring vibration and operating parameters of the thermal power wind turbine; preprocessing the vibration and operating parameters to obtain processed data; extracting features from the preprocessed data, filtering the extracted features, and constructing a feature matrix using the filtered features; inputting the feature matrix into a trained random forest quantitative model, and providing fault early warning based on the output of the random forest quantitative model. Specifically, the computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. The volatile memory may include random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include read-only memory (ROM), hard disk, flash memory, optical disk, magnetic disk, etc.

[0052] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0054] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0055] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0056] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and disclosure of the invention. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0057] It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

[0058] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for quantitative monitoring and fault early warning of vibration in thermal power wind turbines, characterized in that, include: Obtain vibration and operating parameters of thermal power fan; The vibration and operating parameters of the thermal power fan are preprocessed to obtain the processed data; Feature extraction is performed on the preprocessed data, the extracted features are filtered, and a feature matrix is ​​constructed using the filtered features. The feature matrix is ​​input into the trained random forest quantitative model, and fault warning is performed based on the output of the random forest quantitative model.

2. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 1, characterized in that, The vibration and operating parameters of the thermal power fan include the acceleration signals of the front and rear bearings in the bearing housing, the radial vibration velocity signal of the coupling, the circumferential vibration displacement signal at the outlet of the casing, and the fan speed, unit load, inlet and outlet air pressure and air volume, medium temperature and bearing temperature.

3. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 1, characterized in that, The process of preprocessing the vibration and operating parameters of the thermal power fan is as follows: The vibration and operating parameters of the thermal power fan are subjected to baseline correction, noise removal and normalization.

4. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 1, characterized in that, The process of extracting features from the preprocessed data, filtering the extracted features, and constructing a feature matrix using the filtered features is as follows: Feature extraction is performed on the preprocessed data. The extracted features are then screened using the feature importance evaluation mechanism of the random forest algorithm, and a feature matrix is ​​constructed using the screened features.

5. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 1, characterized in that, The extracted features include time-domain features, frequency-domain features, and operating parameter features.

6. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 1, characterized in that, The output of the random forest quantitative model includes quantitative values ​​of vibration intensity, probability of fault type, and coefficient of fault severity.

7. The method for quantitative monitoring and fault early warning of thermal power fan vibration according to claim 6, characterized in that, The process of performing fault early warning based on the output of the random forest quantitative model is as follows: If the vibration intensity quantitative value is ≥4, or the fault severity coefficient is 0.3-0.5, a prompt alarm will be triggered and pushed to the operation and maintenance terminal. If the vibration intensity quantitative value is ≥6, or the fault severity coefficient is 0.5-0.8, an important alarm will be triggered, and the power plant's DCS system will be linked to display fault information. If the vibration intensity quantitative value is ≥8, or the fault severity coefficient is ≥0.8, an emergency alarm will be triggered, and it is recommended that the unit reduce its load or be shut down for maintenance.

8. A quantitative monitoring and fault early warning system for vibration of thermal power wind turbines, characterized in that, include: The acquisition module is used to acquire the vibration and operating parameters of thermal power fans; The preprocessing module is used to preprocess the vibration and operating parameters of the thermal power fan to obtain the processed data. The extraction module is used to extract features from the preprocessed data, filter the extracted features, and construct a feature matrix using the filtered features. The early warning module is used to input the feature matrix into the trained random forest quantitative model and to provide fault warnings based on the output of the random forest quantitative model.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for quantitative monitoring and fault early warning of thermal power wind turbine vibration as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for quantitative monitoring and fault early warning of thermal power wind turbine vibration as described in any one of claims 1-7.

Images