Turbine blade online damage state evaluation system based on ultrasonic detection

By adopting an online damage assessment system combining ultrasonic detection and machine learning on the turbine blades, the accuracy of blade damage detection in complex environments is solved, high-precision detection and preventive maintenance are achieved, and economic losses are reduced.

CN118961883APending Publication Date: 2024-11-15安徽新力电业科技有限责任公司
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411263838.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The operating environment of the turbine is complex, and the interference of factors such as temperature, humidity, vibration and other factors lead to inaccurate measurement values, making it difficult to accurately detect the damage status of the blade.

Method used

Adopt an online damage status assessment system based on ultrasonic detection, including ultrasonic detection module, signal processing module, damage assessment module, data management module, environmental compensation module and fault warning and diagnosis module, through phased array ultrasonic detection, signal processing, machine learning Model and environmental compensation algorithms for real-time monitoring and evaluation of blade damage.

Benefits of technology

Improve the accuracy of measurement values, adapt to complex environments, reduce detection time, realize preventive maintenance, and reduce downtime and maintenance costs due to blade damage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118961883A_ABST
    Figure CN118961883A_ABST
Patent Text Reader

Abstract

The invention discloses a steam turbine blade online damage state evaluation system based on ultrasonic detection, and the system comprises an ultrasonic detection module which is responsible for capturing the tiny damage or crack of the internal structure of a blade by exciting the resonance mode of the blade and monitoring the tiny change of the resonance frequency and the vibration mode; the signal processing module is responsible for processing the signal acquired by the ultrasonic detection module and extracting key information related to the damage state of the blade; the damage evaluation module is responsible for judging the damage degree and type of the blade and giving an evaluation result; the data management module is responsible for data collection, storage, processing and transmission; the environment compensation module is responsible for correcting and compensating detection data of the ultrasonic detection module by applying an environment compensation algorithm according to environment parameters monitored in real time; in the invention, according to the environmental parameters monitored in real time, an environmental compensation algorithm is used for correction and compensation, and the influence of environmental factors on a measurement result is eliminated.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of steam turbine blades, and in particular to an online damage state assessment system for steam turbine blades based on ultrasonic detection. Background Art

[0002] Steam turbine blades are the core components of steam turbines, and their working performance directly affects the overall operating efficiency and safety of the steam turbine. The blades in the steam turbine undertake the important task of converting the thermal energy of steam into mechanical energy, so their working condition and performance have a vital impact on the overall performance of the steam turbine. By evaluating the damage status of steam turbine blades, it is possible to timely discover the safety hazards of the blades, and take corresponding maintenance or replacement measures to avoid the impact of blade failure on the overall operation of the steam turbine. This can not only improve the operating safety of the steam turbine, but also reduce the economic losses caused by downtime due to failures. Phased array ultrasonic testing technology can penetrate the blade material and detect tiny internal defects, and can adjust the detection angle and depth as needed to achieve comprehensive inspection of the blades.

[0003] In the prior art, the operating environment of a steam turbine is complex and changeable, including interference from multiple factors such as temperature, humidity, and vibration. These factors may cause inaccurate measurements. For example, in a high temperature environment, the elastic modulus of the material will change, resulting in a shift in the resonance frequency. In situations where vibration is strong, the resonance signal may be overwhelmed by noise, making the detection results difficult to identify. Therefore, an online damage status assessment system for steam turbine blades based on ultrasonic detection is proposed. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art that the turbine operating environment is complex and changeable, including interference from multiple factors such as temperature, humidity, vibration, etc., which may cause inaccurate measurement values, and to propose an online damage status assessment system for turbine blades based on ultrasonic detection.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An online damage status assessment system for steam turbine blades based on ultrasonic testing includes:

[0007] Ultrasonic testing module: responsible for using phased array ultrasonic testing to capture tiny damage or cracks in the internal structure of the blade by exciting the resonant mode of the blade and monitoring the slight changes in the resonant frequency and vibration mode;

[0008] Signal processing module: responsible for filtering, denoising, feature extraction and other processing operations on the signals collected by the ultrasonic detection module to extract key information related to the blade damage status;

[0009] Damage assessment module: responsible for determining the damage degree and type of the blade and providing the assessment results;

[0010] Data management module: responsible for data collection, storage, processing and transmission, including original detection data, processed signal characteristics, evaluation results, etc.;

[0011] Environmental compensation module: responsible for correcting and compensating the detection data of the ultrasonic detection module using the environmental compensation algorithm according to the real-time monitored environmental parameters (such as temperature, humidity, vibration, etc.) to eliminate the influence of environmental factors on the measurement results;

[0012] Fault warning and diagnosis module: When the system detects an abnormality or fault, it automatically sends out a warning signal and provides fault diagnosis information;

[0013] The ultrasonic detection module transmits the captured original reflected wave signal to the signal processing module, the signal processing module transmits the processed signal characteristics to the damage assessment module, the damage assessment module transmits the assessment results to the data management module and the fault warning and diagnosis module, the data management module receives the data transmitted by the ultrasonic detection module, the signal processing module and the damage assessment module, the environmental compensation module transmits the corrected data to the ultrasonic detection module, and the fault warning and diagnosis module receives the data or abnormal information transmitted by the damage assessment module and the data management module.

[0014] The above technical solution further includes:

[0015] Furthermore, the ultrasonic detection module includes an ultrasonic transmitting unit, an ultrasonic receiving unit, a signal excitation and acquisition circuit, and a resonance analysis unit. The ultrasonic transmitting unit is responsible for transmitting ultrasonic signals to the blades, the ultrasonic receiving unit is responsible for receiving vibration signals generated by the blades in a resonant state, the signal excitation and acquisition circuit is responsible for generating an excitation signal to excite the resonance of the blades, and collecting the vibration signals of the blades in a resonant state. After preliminary processing, the vibration signals are converted into digital signals. The resonance analysis unit receives the digital signals from the signal excitation and acquisition circuit, analyzes the resonance frequency and vibration mode, and evaluates the structural changes and damage of the blades by comparing the resonance characteristics at different time points. The ultrasonic transmitting unit, the ultrasonic receiving unit, and the signal excitation and acquisition circuit interact through electrical signals to realize the transmission of ultrasonic waves and the collection of vibration signals. The signal excitation and acquisition circuit transmits the collected vibration signals to the resonance analysis unit for processing.

[0016] Furthermore, the signal processing module uses a digital signal processor to perform filtering, denoising, feature extraction and other processing operations. The signal processing module will first filter the received signal. The purpose of the filtering process is to remove high-frequency noise and interference components in the signal and retain low-frequency useful information related to blade damage. On the basis of filtering, the signal processing module adopts denoising technology to reduce noise interference in the signal. After filtering and denoising, the signal processing module will extract features of the signal and extract characteristic parameters that can characterize the blade damage state from the signal, such as the amplitude, phase, frequency, waveform, etc. of the signal. The characteristic parameters will provide important data support for subsequent damage assessment.

[0017] Furthermore, the damage assessment module includes a machine learning model, a database and an interface unit. The machine learning model is the core component of the damage assessment module and is responsible for performing damage assessment based on the input signal characteristics. The database is used to store and call the machine learning model and its training data. The database needs to have good data management and access capabilities to ensure that the model can complete the assessment task quickly and accurately. The interface unit is responsible for data interaction between the damage assessment module and the signal processing module. The interface unit defines a clear data format and communication protocol to ensure that data and instructions can be smoothly transmitted between the two modules. The signal processing module transmits the extracted signal characteristics to the damage assessment module through the interface unit, and the interface unit inputs the processed signal characteristics into the machine learning model. The machine learning model needs to access the training data stored in the database during the training process. At the same time, the evaluation results of the machine learning model are stored in the database.

[0018] Furthermore, the data management module includes a database management system, a data storage device, a data transmission interface and a data processing unit. The database management system is responsible for the organization, storage, retrieval and maintenance of data. The database management system provides a data access interface so that other modules can easily access data. The data storage device includes hard disks, SSDs, etc., which are used to actually store data. The data transmission interface includes an Ethernet interface, a serial port, etc., which are used to realize data communication between different modules. The data transmission interface follows communication protocols and standards to ensure the correct transmission and reception of data. The data processing unit is an independent processor or DSP (digital signal processor) responsible for executing data processing algorithms to extract signal features.

[0019] Furthermore, the environmental compensation module includes an environmental parameter monitoring unit and a data processing unit. The environmental parameter monitoring unit includes temperature sensors, humidity sensors, vibration sensors and other equipment, which are responsible for real-time collection of environmental parameter values. The data processing unit is responsible for executing the environmental compensation algorithm and processing the collected environmental parameter values. The data processing unit receives the environmental parameter values ​​and obtains the original measurement data (such as the emission and reception time of the ultrasonic wave, etc.) from the ultrasonic detection module. During the processing, the data processing unit corrects and compensates the original measurement data according to the environmental parameter values ​​and generates the final measurement results.

[0020] Furthermore, the data processing unit uses machine learning to correct and compensate the original measurement data, specifically in the following steps:

[0021] Data collection: Collect the environmental parameter data monitored by the environmental parameter monitoring unit and the corresponding ultrasonic detection module measurement data. These data should cover various environmental conditions, such as different temperatures, humidity and vibration levels;

[0022] Data preprocessing: Cleaning and formatting the collected data to remove noise and outliers, which may include removing invalid data, filling missing values, and standardizing data;

[0023] Feature engineering: Extract useful features from raw data. These features will be used as input to the machine learning model. Features may include environmental parameters such as temperature, humidity, vibration, and measurement data such as ultrasonic propagation time and signal strength.

[0024] Model selection: Select a machine learning model to simulate the relationship between environmental factors and ultrasonic propagation characteristics;

[0025] Model training: Use the collected dataset to train the machine learning model;

[0026] Model validation and testing: Evaluate the performance of the model on independent validation and test sets;

[0027] Model deployment: Deploy the trained model to actual applications for real-time monitoring of environmental parameters and correction of ultrasonic sensor data;

[0028] Continuous learning and optimization: Over time, collect more data and use the collected data to retrain and optimize the model to improve its accuracy and adaptability.

[0029] Furthermore, in the model selection, the data processing unit selects support vector regression to simulate the relationship between environmental factors and ultrasonic wave propagation characteristics.

[0030] Furthermore, during model training, the support vector regression attempts to find the optimal hyperplane so that the prediction error on the data set is minimized;

[0031] The goal of support vector regression is to minimize the following loss function:

[0032]

[0033] At the same time, the constraints are met:

[0034] y i -(w T x i +b)≤ε+ξ i ,ξ i ≥0

[0035]

[0036] Where w is the normal vector of the hyperplane, b is the bias of the hyperplane, ξ i and is the slack variable, C is the regularization parameter, and ε is the tolerance of the loss function.

[0037] Furthermore, the fault warning and diagnosis module includes a fault warning algorithm unit, a fault diagnosis algorithm unit and a user interaction interface unit. The fault warning algorithm unit performs real-time analysis on the processed data according to preset warning rules and thresholds to determine whether the system is in an abnormal or faulty state. Once an abnormality is detected, a warning signal will be triggered. After the warning signal is triggered, the fault diagnosis algorithm unit uses a fault diagnosis algorithm to perform an in-depth analysis of the abnormal data to determine the specific type, location and cause of the fault. The user interaction interface unit includes a display screen, an alarm light, a sound prompt, etc., which are used to display the warning signal, fault diagnosis results and related maintenance suggestions to the operation and maintenance personnel.

[0038] The present invention has the following beneficial effects:

[0039] 1. In the present invention, according to the environmental parameters monitored in real time, the environmental compensation algorithm is used to correct and compensate the signals collected by the ultrasonic detection module to eliminate the influence of environmental factors on the measurement results, which can adapt to the complex and changeable operating environment and improve the accuracy of the measurement value.

[0040] 2. In the present invention, phased array technology can provide high-precision detection because it can flexibly focus on different areas on the blade, thereby improving the accuracy of detection and quickly scanning the entire blade, greatly reducing the detection time and improving production efficiency. At the same time, through real-time evaluation, preventive maintenance can be achieved to avoid downtime and maintenance costs caused by blade damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a system block diagram of an online damage status assessment system for steam turbine blades based on ultrasonic testing proposed by the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] See also Figure 1 As shown, the present invention is an online damage status assessment system for steam turbine blades based on ultrasonic detection, comprising:

[0044] Ultrasonic testing module: responsible for using phased array ultrasonic testing to capture tiny damage or cracks in the internal structure of the blade by exciting the resonant mode of the blade and monitoring the slight changes in the resonant frequency and vibration mode;

[0045] Signal processing module: responsible for filtering, denoising, feature extraction and other processing operations on the signals collected by the ultrasonic detection module to extract key information related to the blade damage status;

[0046] Damage assessment module: responsible for determining the damage degree and type of the blade and providing the assessment results;

[0047] Data management module: responsible for data collection, storage, processing and transmission, including original detection data, processed signal characteristics, evaluation results, etc.;

[0048] Environmental compensation module: responsible for correcting and compensating the detection data of the ultrasonic detection module using the environmental compensation algorithm according to the real-time monitored environmental parameters (such as temperature, humidity, vibration, etc.) to eliminate the influence of environmental factors on the measurement results;

[0049] Fault warning and diagnosis module: When the system detects an abnormality or fault, it automatically sends out a warning signal and provides fault diagnosis information;

[0050] The ultrasonic detection module transmits the captured original reflected wave signal to the signal processing module, the signal processing module transmits the processed signal characteristics to the damage assessment module, the damage assessment module transmits the assessment results to the data management module and the fault warning and diagnosis module, the data management module receives the data transmitted by the ultrasonic detection module, the signal processing module and the damage assessment module, the environmental compensation module transmits the corrected data to the ultrasonic detection module, and the fault warning and diagnosis module receives the data or abnormal information transmitted by the damage assessment module and the data management module.

[0051] The working principle of the online damage status assessment system for turbine blades based on ultrasonic detection proposed in the present invention is that the ultrasonic detection module starts working, transmits ultrasonic signals and receives vibration signals generated by the blades in a resonant state, the signal processing module filters, denoises, extracts features and other processing on the collected vibration signals, the damage assessment module uses a machine learning model to perform damage assessment based on signal characteristics, the data management module is responsible for data collection, storage and transmission, the environmental compensation module compensates for ultrasonic propagation characteristics based on environmental parameters, and the fault warning and diagnosis module monitors the system status. Once an abnormality is detected, a warning signal is immediately issued and fault diagnosis information is provided.

[0052] In one embodiment, for the above-mentioned ultrasonic detection module, the ultrasonic detection module includes an ultrasonic transmitting unit, an ultrasonic receiving unit, a signal excitation and acquisition circuit, and a resonance analysis unit. The ultrasonic transmitting unit is responsible for transmitting ultrasonic signals to the blades, the ultrasonic receiving unit is responsible for receiving vibration signals generated by the blades in a resonant state, the signal excitation and acquisition circuit is responsible for generating an excitation signal to excite the resonance of the blades, and collecting the vibration signals of the blades in a resonant state. After preliminary processing, the vibration signals are converted into digital signals. The resonance analysis unit receives the digital signals from the signal excitation and acquisition circuit, analyzes the resonance frequency and vibration mode, and evaluates the structural changes and damage of the blades by comparing the resonance characteristics at different time points. The ultrasonic transmitting unit, the ultrasonic receiving unit, and the signal excitation and acquisition circuit interact through electrical signals to realize the transmission of ultrasonic waves and the collection of vibration signals. The signal excitation and acquisition circuit transmits the collected vibration signals to the resonance analysis unit for processing.

[0053] In one embodiment, for the above-mentioned signal processing module, the signal processing module uses a digital signal processor to perform filtering, denoising, feature extraction and other processing operations. The signal processing module will first filter the received signal. The purpose of the filtering process is to remove high-frequency noise and interference components in the signal and retain low-frequency useful information related to blade damage. On the basis of filtering, the signal processing module adopts denoising technology to reduce noise interference in the signal. After filtering and denoising, the signal processing module will extract features of the signal and extract characteristic parameters that can characterize the blade damage state from the signal, such as the amplitude, phase, frequency, waveform, etc. of the signal. The characteristic parameters will provide important data support for subsequent damage assessment.

[0054] In one embodiment, for the above-mentioned damage assessment module, the damage assessment module includes a machine learning model, a database and an interface unit. The machine learning model is the core component of the damage assessment module and is responsible for performing damage assessment based on the input signal characteristics. The database is used to store and call the machine learning model and its training data. The database needs to have good data management and access capabilities to ensure that the model can complete the assessment task quickly and accurately. The interface unit is responsible for data interaction between the damage assessment module and the signal processing module. The interface unit defines a clear data format and communication protocol to ensure that data and instructions can be smoothly transmitted between the two modules. The signal processing module transmits the extracted signal characteristics to the damage assessment module through the interface unit, and the interface unit inputs the processed signal characteristics into the machine learning model. The machine learning model needs to access the training data stored in the database during the training process. At the same time, the evaluation results of the machine learning model are stored in the database.

[0055] In one embodiment, for the above-mentioned data management module, the data management module includes a database management system, a data storage device, a data transmission interface and a data processing unit. The database management system is responsible for the organization, storage, retrieval and maintenance of data. The database management system provides a data access interface so that other modules can easily access data. The data storage device includes hard disks, SSDs, etc., which are used to actually store data. The data transmission interface includes an Ethernet interface, a serial port, etc., which are used to realize data communication between different modules. The data transmission interface follows communication protocols and standards to ensure the correct transmission and reception of data. The data processing unit is an independent processor or DSP (digital signal processor) responsible for executing data processing algorithms to extract signal features.

[0056] In one embodiment, for the above-mentioned environmental compensation module, the environmental compensation module includes an environmental parameter monitoring unit and a data processing unit. The environmental parameter monitoring unit includes temperature sensors, humidity sensors, vibration sensors and other equipment, which are responsible for real-time collection of environmental parameter values. The data processing unit is responsible for executing the environmental compensation algorithm and processing the collected environmental parameter values. The data processing unit receives the environmental parameter values. The data processing unit obtains the original measurement data (such as the emission and reception time of the ultrasonic wave, etc.) from the ultrasonic detection module. During the processing, the data processing unit corrects and compensates the original measurement data according to the environmental parameter values ​​and generates the final measurement results.

[0057] In one embodiment, for the above data processing unit, the data processing unit uses machine learning to correct and compensate the original measurement data, specifically in the following steps:

[0058] Data collection: Collect the environmental parameter data monitored by the environmental parameter monitoring unit and the corresponding ultrasonic detection module measurement data. These data should cover various environmental conditions, such as different temperatures, humidity and vibration levels;

[0059] Data preprocessing: Cleaning and formatting the collected data to remove noise and outliers, which may include removing invalid data, filling missing values, and standardizing data;

[0060] Feature engineering: Extract useful features from raw data. These features will be used as input to the machine learning model. Features may include environmental parameters such as temperature, humidity, vibration, and measurement data such as ultrasonic propagation time and signal strength.

[0061] Model selection: Select a machine learning model to simulate the relationship between environmental factors and ultrasonic propagation characteristics;

[0062] Model training: Use the collected dataset to train the machine learning model;

[0063] Model validation and testing: Evaluate the performance of the model on independent validation and test sets;

[0064] Model deployment: Deploy the trained model to actual applications for real-time monitoring of environmental parameters and correction of ultrasonic sensor data;

[0065] Continuous learning and optimization: Over time, collect more data and use the collected data to retrain and optimize the model to improve its accuracy and adaptability.

[0066] In one embodiment, for model selection, in model selection, the data processing unit selects support vector regression to simulate the relationship between environmental factors and ultrasonic wave propagation characteristics.

[0067] In one embodiment, for model training, during model training, the support vector regression attempts to find an optimal hyperplane so that the prediction error on the data set is minimized;

[0068] The goal of support vector regression is to minimize the following loss function:

[0069]

[0070] At the same time, the constraints are met:

[0071] y i -(w T x i +b)≤ε+ξ i ,ξ i ≥0

[0072]

[0073] Where w is the normal vector of the hyperplane, b is the bias of the hyperplane, ξ i and is the slack variable, C is the regularization parameter, and ε is the tolerance of the loss function.

[0074] In one embodiment, for the above-mentioned fault warning and diagnosis module, the fault warning and diagnosis module includes a fault warning algorithm unit, a fault diagnosis algorithm unit and a user interaction interface unit. The fault warning algorithm unit performs real-time analysis on the processed data according to preset warning rules and thresholds to determine whether the system is in an abnormal or faulty state. Once an abnormality is detected, a warning signal will be triggered. After the warning signal is triggered, the fault diagnosis algorithm unit uses a fault diagnosis algorithm to perform an in-depth analysis of the abnormal data to determine the specific type, location and cause of the fault. The user interaction interface unit includes a display screen, an alarm light, a sound prompt, etc., which are used to display the warning signal, fault diagnosis results and related maintenance suggestions to the operation and maintenance personnel.

[0075] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An online damage status assessment system for steam turbine blades based on ultrasonic testing, characterized in that: include: Ultrasonic detection module: responsible for capturing tiny damage or cracks in the internal structure of the blade by exciting the resonant mode of the blade and monitoring the slight changes in the resonant frequency and vibration mode; Signal processing module: responsible for processing the signals collected by the ultrasonic detection module and extracting key information related to the blade damage status; Damage assessment module: responsible for determining the damage degree and type of the blade and providing the assessment results; Data management module: responsible for data collection, storage, processing and transmission; Environmental compensation module: responsible for correcting and compensating the detection data of the ultrasonic detection module using the environmental compensation algorithm according to the real-time monitored environmental parameters; Fault warning and diagnosis module: When the system detects an abnormality or fault, it automatically sends out a warning signal and provides fault diagnosis information; The ultrasonic detection module transmits the captured original reflected wave signal to the signal processing module, the signal processing module transmits the processed signal characteristics to the damage assessment module, the damage assessment module transmits the assessment results to the data management module and the fault warning and diagnosis module, the data management module receives the data transmitted by the ultrasonic detection module, the signal processing module and the damage assessment module, the environmental compensation module transmits the corrected data to the ultrasonic detection module, and the fault warning and diagnosis module receives the data or abnormal information transmitted by the damage assessment module and the data management module.

2. The online damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1 is characterized in that: The ultrasonic detection module includes an ultrasonic transmitting unit, an ultrasonic receiving unit, a signal excitation and acquisition circuit, and a resonance analysis unit. The ultrasonic transmitting unit is responsible for transmitting ultrasonic signals to the blades, the ultrasonic receiving unit is responsible for receiving vibration signals generated by the blades in a resonant state, the signal excitation and acquisition circuit is responsible for generating an excitation signal to excite the resonance of the blades, and collecting the vibration signals of the blades in a resonant state. After preliminary processing, the vibration signals are converted into digital signals. The resonance analysis unit receives the digital signals from the signal excitation and acquisition circuit, analyzes the resonance frequency and vibration mode, and evaluates the structural changes and damage of the blades by comparing the resonance characteristics at different time points. The ultrasonic transmitting unit, the ultrasonic receiving unit, and the signal excitation and acquisition circuit interact through electrical signals to realize the emission of ultrasonic waves and the collection of vibration signals. The signal excitation and acquisition circuit transmits the collected vibration signals to the resonance analysis unit for processing.

3. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1 is characterized in that: The signal processing module uses a digital signal processor to perform processing operations. The signal processing module will first filter the received signal. The purpose of the filtering process is to remove high-frequency noise and interference components in the signal and retain low-frequency useful information related to blade damage. On the basis of filtering, the signal processing module uses denoising technology to reduce noise interference in the signal. After filtering and denoising, the signal processing module will extract features from the signal and extract characteristic parameters that can characterize the blade damage status.

4. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1 is characterized in that: The damage assessment module includes a machine learning model, a database and an interface unit. The machine learning model is responsible for performing damage assessment based on input signal features. The database is used to store and call the machine learning model and its training data. The interface unit is responsible for data interaction between the damage assessment module and the signal processing module. The interface unit defines the data format and communication protocol to ensure that data and instructions can be transmitted between the two modules. The signal processing module transmits the extracted signal features to the damage assessment module through the interface unit. The interface unit inputs the processed signal features into the machine learning model. The machine learning model needs to access the training data stored in the database during the training process. At the same time, the evaluation results of the machine learning model are stored in the database.

5. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1 is characterized in that: The data management module includes a database management system, a data storage device, a data transmission interface and a data processing unit. The database management system is responsible for the organization, storage, retrieval and maintenance of data. The database management system provides a data access interface. The data storage device is used to actually store data. The data transmission interface is used to realize data communication between different modules. The data transmission interface complies with communication protocols and standards. The data processing unit is an independent processor or DSP, which is responsible for executing data processing algorithms to extract signal features.

6. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1 is characterized in that: The environmental compensation module includes an environmental parameter monitoring unit and a data processing unit. The environmental parameter monitoring unit is responsible for collecting environmental parameter values ​​in real time. The data processing unit is responsible for executing the environmental compensation algorithm and processing the collected environmental parameter values. The data processing unit receives the environmental parameter values ​​and obtains the original measurement data from the ultrasonic detection module. During the processing, the data processing unit corrects and compensates the original measurement data according to the environmental parameter values ​​and generates a final measurement result.

7. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 6 is characterized in that: The data processing unit uses machine learning to correct and compensate the original measurement data, specifically in the following steps: Data collection: Collect environmental parameter data monitored by the environmental parameter monitoring unit and corresponding ultrasonic detection module measurement data; Data preprocessing: cleaning and formatting the collected data; Feature engineering: extracting useful features from raw data, which will serve as input to the machine learning model; Model selection: Select a machine learning model to simulate the relationship between environmental factors and ultrasonic propagation characteristics; Model training: Use the collected dataset to train the machine learning model; Model validation and testing: Evaluate the performance of the model on independent validation and test sets; Model deployment: Deploy the trained model to actual applications for real-time monitoring of environmental parameters and correction of ultrasonic sensor data; Continuous learning and optimization: Collect more data over time and use the collected data to retrain and optimize the model.

8. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 7 is characterized in that: In the model selection, the data processing unit selects support vector regression to simulate the relationship between environmental factors and ultrasonic wave propagation characteristics.

9. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 8 is characterized in that: During model training, the support vector regression attempts to find the optimal hyperplane so that the prediction error on the data set is minimized; The goal of support vector regression is to minimize the following loss function: At the same time, the constraints are met: y i -(w T x i +b)≤ε+ξ i ,x i ≥0 Where w is the normal vector of the hyperplane, b is the bias of the hyperplane, ξ i and is the slack variable, C is the regularization parameter, and ε is the tolerance of the loss function.

10. The on-line damage status assessment system for steam turbine blades based on ultrasonic testing according to claim 1, characterized in that: The fault warning and diagnosis module includes a fault warning algorithm unit, a fault diagnosis algorithm unit and a user interaction interface unit. The fault warning algorithm unit performs real-time analysis on the processed data according to preset warning rules and thresholds to determine whether the system is in an abnormal or faulty state. Once an abnormality is detected, a warning signal will be triggered. After the warning signal is triggered, the fault diagnosis algorithm unit uses a fault diagnosis algorithm to perform an in-depth analysis of the abnormal data to determine the specific type, location and cause of the fault. The user interaction interface unit is used to display the warning signal, fault diagnosis results and related maintenance suggestions to the operation and maintenance personnel.

Citation Information

Cited By

  • Forklift framework safety detection system based on ultrasonic waves

    CN119959385A

  • Turbofan engine compressor blade crack multi-mode monitoring system

    CN120044195A

  • Multimodal Monitoring System for Compressor Blade Cracks in Turbofan Engines

    CN120044195B

  • Whole-process management method for transparent detection of power grid materials based on intelligent internet of things

    CN121189347A