Hydraulic power plant water pump operation fault diagnosis method and system
By combining acoustic sensors and support vector machine models, accurate fault monitoring and identification of water pump operation status is achieved, solving the problems of insufficient accuracy and high false alarm rate of traditional diagnostic methods, improving the accuracy and reliability of diagnosis, and providing real-time early warning and cost-effectiveness.
Patent Information
- Application Number
- CN202511044212.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-28
- Publication Date
- 2025-11-18
AI Technical Summary
Traditional water pump fault diagnosis methods are not accurate enough and have a high false alarm rate, and cannot achieve non-invasive, highly sensitive, information-rich, and real-time fault diagnosis.
Acoustic sensors are used to collect acoustic signals from water pump operation. Through principal component analysis and support vector machine models, a fault prediction model is constructed to achieve accurate monitoring of water pump operation status and fault type identification.
It significantly improves the accuracy and reliability of fault diagnosis, has robustness and noise resistance, can monitor the water pump status in real time and provide timely warnings, reduces maintenance costs and extends equipment life.
Smart Images
Figure CN120969159A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of hydropower plant fault monitoring technology, and in particular to a method and system for diagnosing faults in hydropower plant pump operation. Background Technology
[0002] Hydropower plant drainage pumps typically handle drainage during unit maintenance and powerhouse leakage. Due to their complex operating environment, heavy loads, and long operating times, critical components such as pumps are prone to failure. Timely and accurate diagnosis of these faults is crucial for ensuring the safe and stable operation of the hydropower station. Common failure types of large pumps include: wear (material depletion due to long-term operation); cracks (cracks caused by fatigue or stress concentration); poor lubrication (increased friction due to lubrication system malfunctions); and loosening (parts loosening due to improper installation or vibration). Furthermore, traditional fault diagnosis methods are relatively outdated, primarily relying on vibration analysis, temperature monitoring, or oil analysis, requiring periodic shutdowns for manual inspection and maintenance. Summary of the Invention
[0003] This application aims to at least partially address one of the technical problems in the related art.
[0004] Therefore, the first objective of this application is to propose a fault diagnosis method for water pumps in hydropower plants, which solves the problems of insufficient accuracy and high false alarm rate in traditional water pump fault diagnosis methods, and realizes non-invasive, highly sensitive, information-rich, and real-time fault diagnosis.
[0005] The second objective of this application is to propose a fault diagnosis system for water pumps in hydropower plants.
[0006] To achieve the above objectives, the first aspect of this application proposes a method for diagnosing pump operation faults in hydropower plants, comprising: collecting acoustic signature signals of the pump during operation using an acoustic sensor; constructing training and testing sets by selecting acoustic signature signals under normal operating conditions and various fault conditions; extracting fault-related time-frequency domain features from the training and testing sets, and reducing the dimensionality of the extracted features through principal component analysis; selecting a kernel function, setting initial parameters, and constructing a support vector machine model; training and testing the support vector machine model using the training and testing sets and the corresponding time-frequency domain features, and using the support vector machine model with the best prediction accuracy as the pump operation fault prediction model; and using the pump operation fault prediction model to diagnose faults in the hydropower plant pumps.
[0007] To achieve the above objectives, a second aspect of the present invention proposes a fault diagnosis system for hydropower plant pumps, comprising: a data acquisition module for acquiring acoustic signature signals of the pumps during operation using an acoustic sensor; a data selection module for selecting acoustic signature signals under normal operating conditions and various fault conditions to construct training and testing sets; a feature extraction module for extracting fault-related time-frequency domain features from the training and testing sets, and reducing the dimensionality of the extracted features through principal component analysis; a model construction module for selecting a kernel function, setting initial parameters, and constructing a support vector machine model; a model training module for training and testing the support vector machine model using the training and testing sets and corresponding time-frequency domain features, and using the support vector machine model with the best prediction accuracy as the pump operation fault prediction model; and a fault diagnosis module for using the pump operation fault prediction model to diagnose faults in hydropower plant pumps.
[0008] The hydropower plant pump operation fault diagnosis method and system of this application monitors the acoustic signature signal of the pump using an acoustic sensor and combines it with a support vector machine model to achieve accurate fault monitoring and fault type identification of the pump's operating status. This embodiment significantly improves the accuracy and reliability of fault diagnosis, has strong robustness and noise resistance, can monitor the pump's operating status in real time, and immediately issue early warnings when an anomaly is detected, ensuring timely maintenance and preventing further deterioration of the fault. Furthermore, through preventative maintenance and optimized maintenance plans, maintenance costs are effectively reduced and equipment lifespan is extended.
[0009] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description
[0010] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:
[0011] Figure 1 This is a flowchart illustrating a method for diagnosing operational faults of a hydropower plant water pump, as provided in Embodiment 1 of this application.
[0012] Figure 2 This is a schematic diagram of a hydropower plant water pump operation fault diagnosis system provided in an embodiment of this application. Detailed Implementation
[0013] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.
[0014] The following describes a method and system for diagnosing pump operation faults in a hydropower plant, based on embodiments of the present application, with reference to the accompanying drawings.
[0015] Figure 1 This is a flowchart illustrating a method for diagnosing operational faults of a hydropower plant water pump, as provided in Embodiment 1 of this application.
[0016] like Figure 1 As shown, the method for diagnosing pump malfunctions at this hydropower plant includes the following steps:
[0017] Step 101: Use an acoustic sensor to collect acoustic signature signals during water pump operation;
[0018] In this embodiment, the acquired voiceprint signals are preprocessed, including denoising (wavelet transform and Fourier transform) to remove noise and improve signal quality. Simultaneously, missing values in the data are deleted or filled to ensure data integrity and consistency. Furthermore, some outlier data are removed to prevent erroneous data from affecting the model.
[0019] Step 102: Select voiceprint signals under normal operating conditions and various fault conditions to construct training and testing sets;
[0020] In this embodiment, collected data such as vibration signal intensity, temperature, current, and pressure are selected to construct a sample set. The sample set is designed to ensure a balanced distribution of samples of different fault types, with each fault type accounting for at least 20% of the samples to avoid affecting prediction accuracy. The constructed sample set is divided into a training set and a test set in a 9:1 ratio. The training and test sets are ensured to be mutually exclusive, containing no duplicate samples, to avoid overestimating the evaluation results.
[0021] Step 103: Extract time-frequency domain features related to the fault from the training set and the dataset, and reduce the dimensionality of the extracted features through principal component analysis;
[0022] In this embodiment, time-domain features (such as mean, variance, peak value, peak-to-peak value, etc.), frequency-domain features (such as frequency spectrum, power spectral density, etc.), and time-frequency-domain features (such as short-time Fourier transform, wavelet transform, etc.) are extracted from the original data. Features highly correlated with the fault state are screened through correlation analysis, and principal component analysis (PCA) is used to reduce dimensionality, retaining the main information and reducing the feature dimensionality.
[0023] Step 104: Select the kernel function, set the initial parameters, and construct the support vector machine model;
[0024] In this embodiment, a suitable kernel function is selected. This embodiment uses the RBF kernel to adapt to different types of fault data. Initial parameters are set, such as C (penalty parameter) and γ (RBF kernel parameter).
[0025] Step 105: Use the training set, test dataset, and corresponding time-frequency domain features to train and test the support vector machine model, and use the support vector machine model with the best prediction accuracy as the pump operation failure prediction model.
[0026] In this embodiment, model training is achieved using the Scikit-learn library in Python or other machine learning toolkits.
[0027] In this embodiment, the support vector machine model is tested using test set data. The fault types predicted by the model are compared with the actual fault types. Error evaluation parameters such as accuracy, recall, F1 score, MSE, RMSE, R correlation coefficient, MAE, and MAPE are calculated, and a confusion matrix is plotted to observe the trend of parameter changes, which serves as a reference for modifying model parameters C and γ.
[0028] In this embodiment, the parameters of the support vector machine model, such as C and γ, are modified according to the evaluation results and confusion matrix, and the training set is used to generate the model and the test set is used to test the model.
[0029] In this embodiment, to avoid local optima, the model parameters need to be modified at least n times (n is the number of training iterations, and should ideally be no less than 10 times) to obtain multiple support vector machine models. During parameter selection, it is recommended that C and γ be chosen within a reasonable range. After training multiple models, the final model is selected based on the evaluation parameters.
[0030] In this embodiment, the evaluation result is based on test accuracy, while recall and precision should be used as auxiliary indicators for parameter modification. The error evaluation parameters such as MSE, RMSE, R correlation coefficient, MAE, and MAPE have basically the same trend. When modifying model parameters, we can focus on the overall numerical trend, without needing to reduce all parameters.
[0031] Step 106: Use the pump operation failure prediction model to diagnose the faults of the hydropower plant's pumps.
[0032] The hydropower plant pump operation fault diagnosis method of this application monitors the acoustic signature signal of the pump using an acoustic sensor and combines it with a support vector machine model to achieve accurate fault monitoring and fault type identification of the pump's operating status. This embodiment significantly improves the accuracy and reliability of fault diagnosis, has strong robustness and noise resistance, can monitor the pump's operating status in real time, and immediately issue early warnings when an anomaly is detected, ensuring timely maintenance and preventing further deterioration of the fault. Furthermore, through preventative maintenance and optimized maintenance plans, maintenance costs are effectively reduced and equipment lifespan is extended.
[0033] To achieve the above embodiments, this application also proposes a fault diagnosis system for water pump operation in hydropower plants.
[0034] Figure 2 This is a schematic diagram of a hydropower plant water pump operation fault diagnosis system provided in an embodiment of this application.
[0035] like Figure 2 As shown, the hydropower plant's pump operation fault diagnosis system includes:
[0036] The data acquisition module is used to collect acoustic signature signals of the water pump during operation using an acoustic sensor.
[0037] The data selection module is used to select voiceprint signals under normal operating conditions and various fault conditions to construct training and test sets.
[0038] The feature extraction module is used to extract time-frequency domain features related to faults from the training set and dataset, and to reduce the dimensionality of the extracted features through principal component analysis.
[0039] The model building module is used to select the kernel function, set the initial parameters, and build the support vector machine model.
[0040] The model training module is used to train and test the support vector machine model using the training set, test dataset and corresponding time-frequency domain features, and to use the support vector machine model with the best prediction accuracy as the water pump operation fault prediction model.
[0041] The fault diagnosis module is used to diagnose faults in hydropower plant pumps using a pump operation fault prediction model.
[0042] Furthermore, in this embodiment of the application, a data processing module is also included, which is used to perform denoising processing on the collected voiceprint signals and delete missing values and outlier data before constructing training and test sets by selecting voiceprint signals under normal working conditions and various fault conditions. The denoising processing includes wavelet transform and Fourier transform.
[0043] Specifically, in the embodiments of this application, voiceprint signals under normal operating conditions and various fault conditions are selected to construct training sets and test sets, including:
[0044] A sample set was constructed by uniformly selecting voiceprint signals under normal operating conditions and various fault conditions;
[0045] The sample set is divided into a training set and a test set according to a preset ratio. During the division, the training set and the test set are mutually exclusive and do not contain the same samples.
[0046] Furthermore, in the embodiments of this application, the time-frequency domain features include time-domain features and frequency-domain features. The time-domain features include mean, variance, and peak value, while the frequency-domain features include frequency spectrum and power spectral density.
[0047] Specifically, in the embodiments of this application, selecting a kernel function, setting initial parameters, and constructing a support vector machine model include:
[0048] Select the RBF kernel, set the initial parameters including the penalty parameter C and the RBF kernel parameter γ, and construct the support vector machine model;
[0049] The support vector machine model was tested, including:
[0050] At least one of precision, recall, F1 score, MSE, RMSE, R correlation coefficient, MAE, and MAPE is selected as an error evaluation parameter, and a confusion matrix is constructed.
[0051] Training the support vector machine model includes:
[0052] Modify the parameters C and γ of the support vector machine model based on the error evaluation parameters and the confusion matrix;
[0053] Training a support vector machine model also includes:
[0054] Set the number of modifications to n to obtain n trained support vector machine models;
[0055] Determine the pump operation failure prediction model, including:
[0056] Based on the error evaluation parameters and confusion matrix, the support vector machine model with the best prediction accuracy is selected from n trained support vector machine models as the pump operation failure prediction model.
[0057] It should be noted that the foregoing explanation of the embodiment of the method for diagnosing pump operation failures in hydropower plants also applies to the hydropower plant pump operation failure diagnosis system of this embodiment, and will not be repeated here.
[0058] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0059] Furthermore, 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 at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0060] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.
[0061] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.
[0062] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0063] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
[0064] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
[0065] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.
Claims
1. A method for diagnosing a fault in the operation of a water pump of a hydroelectric plant, characterized in that, The method comprises the following steps: Collecting soundprint signals of the water pump in operation by using an acoustic sensor; Selecting soundprint signals under normal working conditions and various fault working conditions to construct a training set and a test set; Extracting time-frequency domain features related to faults from the training set and the data set, and reducing the extracted features by principal component analysis; Selecting a kernel function, setting initial parameters, and constructing a support vector machine model; Training and testing the support vector machine model using the training set, the test data set, and the corresponding time-frequency domain features, and selecting the support vector machine model with the best prediction accuracy as the water pump operation fault prediction model; Using the water pump operation fault prediction model to diagnose the faults of the water pump in the hydropower plant.
2. The method of claim 1, wherein, Before selecting soundprint signals under normal working conditions and various fault working conditions to construct a training set and a test set, the method further comprises the following steps: Performing denoising processing on the collected soundprint signals, and deleting missing values and outlier data, wherein the denoising processing comprises wavelet transform and Fourier transform.
3. The method of claim 1, wherein, The step of selecting soundprint signals under normal working conditions and various fault working conditions to construct a training set and a test set comprises the following steps: Uniformly selecting soundprint signals under normal working conditions and various fault working conditions to construct a sample set; Dividing the sample set into a training set and a test set according to a predetermined ratio, and ensuring that the training set and the test set are mutually exclusive and do not contain the same samples.
4. The method of claim 1, wherein, The time-frequency domain features include time domain features and frequency domain features, the time domain features include mean, variance, and peak value, and the frequency domain features include frequency spectrum and power spectral density.
5. The method of claim 1, wherein, The step of selecting a kernel function, setting initial parameters, and constructing a support vector machine model comprises the following steps: Selecting an RBF kernel, setting initial parameters including a penalty parameter C and an RBF kernel parameter γ, and constructing a support vector machine model; The step of testing the support vector machine model comprises the following steps: Selecting at least one of accuracy, recall rate, F1 score, MSE, RMSE, R correlation coefficient, MAE, and MAPE as an error evaluation parameter, and constructing a confusion matrix; The step of training the support vector machine model comprises the following steps: Modifying the parameters C and γ of the support vector machine model according to the error evaluation parameter and the confusion matrix; The step of training the support vector machine model further comprises the following steps: Setting the number of modifications to n times to obtain n trained support vector machine models; The step of determining the water pump operation fault prediction model comprises the following steps: Selecting the support vector machine model with the best prediction accuracy from the n trained support vector machine models as the water pump operation fault prediction model according to the error evaluation parameter and the confusion matrix.
6. A hydroelectric plant water pump operation failure diagnosis system characterized by, The method comprises the following steps: A data collection module is configured to collect soundprint signals of the water pump in operation by using an acoustic sensor; A data selection module is configured to select soundprint signals under normal working conditions and various fault working conditions to construct a training set and a test set; A feature extraction module is configured to extract time-frequency domain features related to faults from the training set and the data set, and reduce the extracted features by principal component analysis; A model construction module is configured to select a kernel function, set initial parameters, and construct a support vector machine model; A model training module is configured to train and test the support vector machine model using the training set, the test data set, and the corresponding time-frequency domain features, and select the support vector machine model with the best prediction accuracy as the water pump operation fault prediction model; and A fault diagnosis module is configured to perform fault diagnosis on the water pump of the hydropower plant using the water pump operation fault prediction model.
7. The system of claim 6, wherein, The data processing module is further configured to perform denoising processing on the collected voiceprint signals and delete missing values and outlier data before selecting the voiceprint signals under normal working conditions and various fault working conditions to construct a training set and a test set, wherein the denoising processing includes wavelet transform and Fourier transform.
8. The system of claim 6, wherein, The selecting the voiceprint signals under normal working conditions and various fault working conditions to construct the training set and the test set includes: uniformly selecting the voiceprint signals under normal working conditions and various fault working conditions to construct a sample set; dividing the sample set into the training set and the test set according to a preset ratio, and making the training set and the test set mutually exclusive and not containing the same sample during the dividing.
9. The system of claim 6, wherein, The time-frequency domain features include time domain features and frequency domain features, the time domain features include mean, variance and peak value, and the frequency domain features include frequency spectrum and power spectral density.
10. The system of claim 6, wherein, The selecting the kernel function and setting the initial parameters includes: selecting an RBF kernel and setting initial parameters including a penalty parameter C and an RBF kernel parameter gamma to construct a support vector machine model; The testing the support vector machine model includes: selecting at least one of accuracy, recall, F1 score, MSE, RMSE, R correlation coefficient, MAE and MAPE as an error evaluation parameter and constructing a confusion matrix; The training the support vector machine model includes: modifying the parameters C and gamma of the support vector machine model according to the error evaluation parameter and the confusion matrix; The training the support vector machine model further includes: setting the number of modifications to n times to obtain n trained support vector machine models; The determining the water pump operation fault prediction model includes: selecting a support vector machine model with the best prediction accuracy from the n trained support vector machine models as the water pump operation fault prediction model according to the error evaluation parameter and the confusion matrix.