Hydropower station inspection fault data analysis method, electronic equipment and storage medium

Abstract

The invention provides a hydropower station inspection fault data analysis method, electronic equipment and a storage medium, and belongs to the technical field of hydropower engineering. Comprises: establishing a hydropower station inspection fault data prediction model; acquiring historical fault-free data and current operation data; training and testing the prediction model by using historical fault-free data; calculating a similarity distance between a model prediction value and an actual operation value; modeling a similarity distance probability density distribution state between a model prediction value and an actual operation value, and establishing an operation condition evaluation index; the real-time health index of hydropower station operation is obtained by utilizing hydropower station inspection actual measurement monitoring data, and the hydropower station operation influence degree is obtained by analyzing fault data based on the health index. The technical problem of insufficient analysis of hydropower station inspection fault data is solved. According to the invention, comprehensive monitoring and real-time analysis of the state of the hydropower station equipment are realized, the accuracy and response speed of fault diagnosis are improved, the labor cost is reduced, and the operation and maintenance efficiency is improved.

Description

Technical Field

[0001] The present application relates to a fault data analysis method, and in particular to a hydropower station inspection fault data analysis method, electronic equipment and storage medium, and belongs to the technical field of hydropower engineering. Background Art

[0002] With the rapid development of social economy, people's living standards are constantly improving, and the demand for electricity is gradually increasing, which puts higher requirements on the load bearing capacity of the power network. Hydropower station units are an important part of the power network. Effective treatment measures should be taken in time after a fault occurs. The reliable and stable operation must be guaranteed to make the power network safer. The development of hydropower projects is very fast, the scale and number of hydropower stations are increasing, and the level of automation is constantly improving. During the operation of hydropower station units, various types of faults are prone to occur, and the manifestations are also very complex. Targeted treatment measures need to be taken to ensure their normal operation. In addition, the prevention and maintenance of hydropower station units must be done well, so as to effectively reduce the probability of failures and ensure the continuity and stability of power transmission.

[0003] In the process of development of the hydropower industry, the number of hydropower stations, installed capacity and power generation have greatly increased. If there is a lack of analysis and treatment of common faults in the operation of hydropower station units, the safety hazards during the operation of the hydropower station will be greatly increased, and huge property losses will be caused. Common operating faults of hydropower station units mainly include overload, vibration and loss of step, forced excitation, single-phase grounding of bus voltage, single-point grounding of rotor, fire of slip ring and excitation machine carbon brush, loss of control of secondary circuit, shear pin shearing, cooling water interruption and high bearing temperature. If the above faults cannot be discovered and handled in the first time, it will bring huge accidents to the hydropower station, which will not only reduce economic benefits, but also damage power generation facilities, cause casualties, and lead to paralysis of the entire power system. Therefore, it is of great significance to do a good job in the analysis and treatment of common faults in the operation of hydropower station units.

[0004] Therefore, it is particularly important to propose a method for analyzing hydropower station inspection fault data. Summary of the invention

[0005] A brief overview of the present invention is provided below in order to provide a basic understanding of certain aspects of the present invention. It should be understood that this overview is not an exhaustive overview of the present invention. It is not intended to identify key or important parts of the present invention, nor is it intended to limit the scope of the present invention. Its purpose is merely to present certain concepts in a simplified form as a prelude to a more detailed description discussed later.

[0006] In view of this, in order to solve the technical problem of insufficient analysis of hydropower station inspection fault data in the prior art, the present invention provides a method for analyzing hydropower station inspection fault data. The impact of hydropower station unit failure includes the impact at a certain moment and the impact over time. Therefore, the present invention can determine the degree of influence of each parameter on the operation of the unit by establishing the relationship between fault data and fault-free data, thereby guiding the maintenance design, and providing a guiding basis for the maintenance decision-making of maintenance personnel, so as to realize the analysis of the impact of hydropower station unit equipment and other information on the operation of the unit and the rate of change of faults. It can be used to guide unit maintenance, reasonably reduce the time and distribution range of unit operation failures, and improve the unit operation efficiency.

[0007] Solution 1: A method for analyzing fault data of a hydropower station inspection, comprising the following steps:

[0008] Establish a hydropower station inspection fault data prediction model;

[0009] Obtain historical fault-free data and current operation data;

[0010] Use historical fault-free data to train and test the prediction model;

[0011] Calculate the similarity distance between the model prediction value and the actual running value;

[0012] Model the probability density distribution state of the similarity distance between the model prediction value and the actual operation value, and establish the operation status evaluation index;

[0013] The real-time health index of the hydropower station operation is obtained by using the actual monitoring data of the hydropower station inspection. The degree of impact on the operation of the hydropower station is obtained by analyzing the fault data based on the health index.

[0014] Preferably, the input data of the prediction model is hydropower station inspection data, and the output data is hydropower station inspection fault data.

[0015] Preferably, the inspection data includes: equipment pressure data, equipment temperature data, equipment flow data; equipment released gas concentration data, equipment switch status data, equipment audio spectrum data, equipment vibration frequency data, equipment displacement data, equipment speed data, equipment wear data, equipment corrosion data and equipment sealing data.

[0016] Preferably, the inspection data collection frequency is 10s to 20min.

[0017] Preferably, the historical fault-free data is full-operation condition operation data.

[0018] Preferably, the prediction model adopts a multivariate nonlinear regression model, a BP neural network model, a LSTM neural network model or an autoencoder model.

[0019] Preferably, the method for calculating the similarity distance between the model prediction value and the actual running value is:

[0020]

[0021] in, is the similarity distance, is the model prediction vector, x is the actual operation vector, is the Euclidean distance, is the cosine of the angle, represents the inner product of the predicted vector and the actual running vector, and || || represents the binary norm of the vector.

[0022] Preferably, the probability density distribution state is in the form of:

[0023]

[0024] Among them, ω k is the weight coefficient, and φ(x|θ k ) is the single Gaussian probability density function of the kth sub-model, where μ k That is, the mean, is the variance of the Gaussian probability density, calculated as follows:

[0025]

[0026] The operating status evaluation index is represented by a health index, which is defined as:

[0027]

[0028] HI t =(1-α)HI t-1 +GPD t

[0029] Where: GPD at time t t Moment likelihood probability; HI t is the health index of the hydropower station at time t; α is the balance coefficient, and its value range is 0.05~0.30.

[0030] Solution 2: An electronic device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a hydropower station inspection fault data analysis method described in Solution 1 are implemented.

[0031] Solution three: A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a method for analyzing hydropower station inspection fault data as described in Solution one.

[0032] The beneficial effects of the present invention are as follows: The present invention utilizes the operating mechanism of hydropower station equipment, combined with multivariate nonlinear, neural network and other big data intelligent analysis algorithms, to propose a prediction model, which can determine the benchmark output (i.e., predicted output) of the equipment under different operating conditions. A similarity distance is proposed to measure the similarity information and trend information between the predicted output and the actual operating value, and a corresponding probability density analysis is performed to analyze the equipment operating performance degradation and abnormal conditions in advance, reminding maintenance personnel to focus on them, so as to perform intelligent diagnosis and careful inspection and maintenance of hydropower station equipment as soon as possible, avoid serious faults, and ensure the safety and economy of hydropower station units. The present invention realizes comprehensive monitoring and real-time analysis of the status of hydropower station equipment, improves the accuracy and response speed of fault diagnosis, reduces labor costs, and improves operation and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0034] Figure 1 The present invention is a flow chart of a method for analyzing fault data of a hydropower station inspection. DETAILED DESCRIPTION

[0035] In order to make the technical solutions and advantages in the embodiments of the present application more clearly understood, the exemplary embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0036] Example 1, reference Figure 1 The present embodiment describes a method for analyzing fault data of a hydropower station inspection, comprising the following steps:

[0037] S1. Establish a hydropower station inspection fault data prediction model;

[0038] The input data of the prediction model is the inspection data of the hydropower station, and the output data is the inspection fault data of the hydropower station.

[0039] The inspection data include: equipment pressure data, equipment temperature data, equipment flow data; equipment released gas concentration data, equipment switch status data, equipment audio spectrum data, equipment vibration frequency data, equipment displacement data, equipment speed data, equipment wear data, equipment corrosion data and equipment sealing data.

[0040] Equipment includes turbines, generators, pressure pipes, gates, water inlets, water outlets, transformers, switch stations, etc. that need to be monitored for faults.

[0041] For example, in the water pipeline pressure input value prediction model that transports water to the turbine, the output water pipeline pressure value is abnormal. When the water pipeline pressure value range is not within the normal value range, the output data is faulty.

[0042] The inspection data collection frequency is 10s to 20min.

[0043] The prediction model uses a multivariate nonlinear regression model, a BP neural network model, a LSTM neural network model or an autoencoder model. Its expression is as follows:

[0044]

[0045] Where: y is the pressure value of the water pipeline; i represents the i-th group of data; f(x1,x2,…,x j ,k1,k2,…,k p ) is a multivariate nonlinear function, representing the deterministic part; x1,x2,…,x j are independent variables; k1, k2, …, k p is the unknown model parameter of the multivariate nonlinear function; εσ i is the random part, ε is a random variable that obeys N(0,1) distribution; σ i is the standard deviation of the random distribution of the i-th group of data.

[0046] S2. Obtain historical fault-free data and current operation data;

[0047] The historical fault-free data is the full-condition operation data. The historical fault-free data can be collected and filtered from the unit database; the current operation data is obtained from the relevant collection devices arranged on the equipment to be monitored;

[0048] The collection device may be a vibration sensor, a gas sensor, a collection robot, an image collection device, or a drone, etc., which can collect data related to the equipment to be monitored;

[0049] The collection device is arranged at a fixed point on the equipment that needs to be monitored.

[0050] S3. Use historical fault-free data to train and test the prediction model;

[0051] Divide the historical fault-free data into training data and test data to train and test the prediction model;

[0052] Specifically: preprocess the historical accessibility data, including data cleaning, missing value filling, and normalization; eliminate the influence of different dimensions through preprocessing.

[0053] The preprocessed data is divided into a training set and a test set. The training set accounts for 70% of the total data; the test set accounts for 30% of the total data. This division ensures that the model has enough data to learn during training, while also retaining a portion of the data for the final model evaluation.

[0054] Optimize the model by adjusting model parameters and hyperparameters;

[0055] Use the test set to evaluate the performance of the model using accuracy, recall, and precision metrics;

[0056] Accuracy: Accuracy is the most intuitive evaluation indicator, which indicates the proportion of samples predicted correctly by the model to the total number of samples. The calculation formula for accuracy is:

[0057]

[0058] Recall: Recall, also known as the true positive rate (TPR), measures the proportion of samples that are correctly predicted to be positive among all samples that are actually positive. The calculation formula for recall is

[0059]

[0060] Precision: Precision, also known as the precision rate, refers to the samples predicted to be positive. It means the proportion of samples predicted to be positive that are actually positive. The calculation formula for precision is:

[0061]

[0062] S4. Calculate the similarity distance between the model prediction value and the actual operation value;

[0063] The method to calculate the similarity distance between the model prediction value and the actual running value is:

[0064]

[0065] in, is the similarity distance, is the model prediction vector, x is the actual operation vector, is the Euclidean distance, is the cosine of the angle, represents the inner product of the predicted vector and the actual running vector, and || || represents the binary norm of the vector.

[0066] S5. Model the probability density distribution state of the similarity distance between the model prediction value and the actual operation value, and establish the operation status evaluation index;

[0067] The probability density distribution state form is:

[0068]

[0069] Among them, ω k is the weight coefficient, and ω k ≥0, φ(x|θ k ) is the single Gaussian probability density function of the kth sub-model, where μ k That is, the mean, is the variance of the Gaussian probability density, calculated as follows:

[0070]

[0071] The operating status evaluation index is represented by a health index, which is defined as:

[0072]

[0073] HI t =(1-α)HI t-1 +GPD t

[0074] Where: GPD at time t t Moment likelihood probability; HI t is the health index of the hydropower station at time t; α is the balance coefficient, and its value range is 0.05~0.30.

[0075] S6. Utilize the actual monitoring data of the hydropower station inspection to obtain the real-time health index of the hydropower station operation, and analyze the fault data based on the health index to obtain the degree of impact on the operation of the hydropower station.

[0076] The healthy threshold is set by selecting a confidence level of 99.9% and using the kernel density estimation method to obtain the healthy threshold. In this embodiment, the healthy threshold is 2.9603. The kernel density estimation expression is:

[0077]

[0078] Where: is the estimated probability density value, n is the number of samples, h is the window width, and K(·) is the kernel function.

[0079] The method of analyzing fault data based on health index to obtain the degree of impact on the operation of the hydropower station is: according to the normal operating status of the equipment and historical fault data, the impact threshold is set to quantify the health status of the equipment. The impact threshold range is set according to different equipment conditions. When the health index exceeds a certain threshold, an early warning signal is triggered to prompt maintenance personnel to conduct relevant maintenance inspections;

[0080] The specific method is: establish a health index model, which can predict the health status of the equipment based on the input health index.

[0081] Building a health index model can use machine learning algorithms or statistical methods;

[0082] Machine learning algorithms include support vector machines, random forests, or neural networks;

[0083] Statistical methods include principal component analysis or cluster analysis;

[0084] Training and validation: Use historical data to train the model and evaluate the model's performance through cross-validation or setting aside a portion of the data as a test set. Adjust the model parameters until satisfactory accuracy and generalization ability are achieved.

[0085] Embodiment 2 is different from Embodiment 1 in that it also includes S7. Visualizing the real-time monitoring data of the hydropower station, the real-time health index of the hydropower station operation, and the degree of impact on the operation of the hydropower station.

[0086] Use line charts or bar charts to display real-time monitoring data to intuitively show the changing trends of various indicators. Use color coding or marking to distinguish data from different devices or measurement points, which helps to quickly locate problem areas.

[0087] Display the health index on a chart in the form of numerical values ​​or color levels to intuitively reflect the health status of the device. Heat maps can be used to represent the spatial distribution of the health index to help identify potential problem areas.

[0088] Combined with historical data, draw a trend chart of the health index to analyze its changing patterns and potential risks. By comparing the health indexes of different devices or measurement points, evaluate their relative health status.

[0089] Display fault data in the form of scatter plots or pie charts; combine with the time axis to draw a time series diagram of fault occurrence to reveal the regularity and periodicity of fault occurrence.

[0090] Use association rules or network diagrams to analyze the associations between faults and provide clues for fault diagnosis. Use visualization to display the fault handling process and results and evaluate the efficiency and effectiveness of fault handling.

[0091] Build a comprehensive visualization platform that integrates real-time monitoring, health assessment, and fault diagnosis to achieve centralized data management and analysis.

[0092] Embodiment 3: The computer device of the present invention may be a device including a processor and a memory, such as a single chip microcomputer including a central processing unit, etc. Moreover, the processor is used to implement the steps of the above-mentioned hydropower station inspection fault data analysis method when executing the computer program stored in the memory.

[0093] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0094] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0095] Embodiment 4, computer readable storage medium embodiment

[0096] The computer-readable storage medium of the present invention can be any form of storage medium that can be read by a processor of a computer device, including but not limited to non-volatile memory, volatile memory, ferroelectric memory, etc. A computer program is stored on the computer-readable storage medium. When the processor of the computer device reads and executes the computer program stored in the memory, the steps of the above-mentioned method for analyzing hydropower station inspection fault data can be implemented.

[0097] The computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer readable media do not include electric carrier signals and telecommunication signals.

[0098] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or defining the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative, not restrictive, with respect to the scope of the present invention, which is defined by the appended claims.

Claims

1. A method for analyzing fault data of a hydropower station inspection, characterized in that: The following steps are involved: Establish a hydropower station inspection fault data prediction model; Obtain historical fault-free data and current operation data; Use historical fault-free data to train and test the prediction model; Calculate the similarity distance between the model prediction value and the actual running value; Model the probability density distribution state of the similarity distance between the model prediction value and the actual operation value, and establish the operation status evaluation index; The real-time health index of the hydropower station operation is obtained by using the actual monitoring data of the hydropower station inspection. The degree of impact on the operation of the hydropower station is obtained by analyzing the fault data based on the health index.

2. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The input data of the prediction model is the inspection data of the hydropower station, and the output data is the inspection fault data of the hydropower station.

3. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The inspection data include: equipment pressure data, equipment temperature data, equipment flow data; equipment released gas concentration data, equipment switch status data, equipment audio spectrum data, equipment vibration frequency data, equipment displacement data, equipment speed data, equipment wear data, equipment corrosion data and equipment sealing data.

4. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The inspection data collection frequency is 10s to 20min.

5. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The historical fault-free data refers to the full-operation condition operation data.

6. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The prediction model adopts a multivariate nonlinear regression model, a BP neural network model, a LSTM neural network model or an autoencoder model.

7. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The method to calculate the similarity distance between the model prediction value and the actual running value is: in, is the similarity distance, is the model prediction vector, x is the actual operation vector, is the Euclidean distance, is the cosine of the angle, represents the inner product of the predicted vector and the actual running vector, and || || represents the binary norm of the vector.

8. A method for analyzing fault data of a hydropower station inspection according to claim 1, characterized in that: The probability density distribution state form is: Among them, ω k is the weight coefficient, and φ(x|θ k ) is the single Gaussian probability density function of the kth sub-model, where μ k That is, the mean, is the variance of the Gaussian probability density, calculated as follows: The operating status evaluation index is expressed as a health index, which is defined as: HI t =(1-α)HI t-1 +GPD t Where: GPD at time t t Moment likelihood probability; HI t is the health index of the hydropower station at time t; α is the balance coefficient, and its value range is 0.05~0.

30.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a hydropower station inspection fault data analysis method as described in any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, a method for analyzing inspection fault data of a hydropower station as described in any one of claims 1 to 8 is implemented.

Images