Generator set upper rack loosening fault diagnosis method and system based on frequency division algorithm

Abstract

The invention relates to the technical field of equipment fault diagnosis, in particular to a generator set upper rack loosening fault diagnosis method and system based on a frequency division algorithm, and the method comprises the following steps: adjusting the sampling rate of vibration data according to the equipment operation state based on the real-time operation data of a generator, x-direction and Y-direction horizontal vibration data of a rack on the generator set are collected and recorded, and a real-time vibration data set is obtained. According to the method, the vibration data sampling rate is dynamically adjusted through the real-time operation data, storage of irrelevant data is reduced, it is ensured that fault signals are captured at the critical moment, the fault diagnosis efficiency is improved, a key frequency area is recognized through spectral analysis, the data collection frequency band is adjusted, the capacity of capturing fault related vibration characteristics is enhanced, and the fault diagnosis accuracy is improved. The relation between vibration data and the unit state is analyzed by combining the guide vane opening degree and the working water head data, the accuracy of fault source recognition is improved, and accurate data support is provided for equipment maintenance through abnormal vibration mode analysis and looseness degree quantification.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment fault diagnosis, and in particular to a method and system for diagnosing loose faults of upper frames of generator sets based on a frequency division algorithm. Background Art

[0002] The field of equipment fault diagnosis technology focuses on identifying and solving problems that arise in the operation of mechanical equipment through a variety of monitoring technologies and analysis methods. By discovering equipment failures at an early stage, it reduces economic losses caused by major damage and production downtime. Fault diagnosis technologies include vibration analysis, sound analysis, temperature monitoring, and electrical parameter analysis. Combined with data analysis, it detects and identifies abnormal conditions of equipment, uses smart sensors, the Internet of Things, and artificial intelligence to improve the efficiency and accuracy of fault diagnosis, achieve real-time monitoring of equipment status, automatically analyze the cause of the failure, and improve the predictability and response speed of maintenance work.

[0003] Among them, the fault diagnosis method for loose upper frame of generator set is used to detect and diagnose whether there is looseness in the upper frame of generator set during operation. Loose upper frame will cause unstable operation of generator, and long-term existence will cause mechanical failure and safety accidents. The diagnosis method analyzes the vibration data of the frame and uses signal processing algorithm to identify abnormal vibration patterns, judge the looseness of the frame, realize early intervention and repair, ensure the stability of generator set, prevent production interruption and economic losses caused by equipment failure, and ensure the reliability and safety of power generation facilities.

[0004] In actual operation, the traditional fault diagnosis method for loose frame on generator sets lacks the ability to dynamically adjust the data acquisition frequency, resulting in a large amount of redundant data when the equipment is running stably, wasting storage space and processing resources, and missing key fault signals at critical moments due to insufficient sampling rate. The fixed spectrum analysis method is used in the analysis of vibration data, which lacks accurate identification and tracking of key frequency areas, limits the sensitivity and accuracy of fault diagnosis, fails to fully consider the relationship between mechanical and hydraulic conditions, and is difficult to provide a comprehensive fault analysis perspective, reducing the effectiveness of fault prediction and early intervention, leading to the neglect of potential serious faults, and increasing the risk of equipment operation and maintenance costs. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a generator set upper frame loose fault diagnosis method and system based on a frequency division algorithm.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme, a method for diagnosing loose faults of upper frame of generator set based on frequency division algorithm, comprising the following steps: S1: Based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction of the upper frame of the generator set is collected and recorded to obtain a real-time vibration data set; S2: Based on the real-time vibration data set, by performing spectrum analysis on the vibration data, analyzing the amplitudes of multiple frequency components in the spectrum, identifying key frequency areas, and adjusting the frequency band of data acquisition to capture fault-related vibration characteristics, generating a vibration data analysis record; S3: Analyze the relationship between the vibration data and the hydraulic state according to the vibration data analysis record, combined with the real-time guide vane opening and working water head data, and obtain the state correlation analysis result; S4: According to the state association analysis result, by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with the standard vibration mode library, abnormal vibration modes are identified and abnormal vibration mode data is generated; S5: Based on the abnormal vibration mode data, analyzing the frequency, amplitude and phase of the abnormal vibration, calculating key features of the abnormal mode, quantifying the degree of looseness, and generating looseness degree quantification data; S6: Using the looseness quantification data, by comparing the looseness with a preset threshold, evaluating the severity level of the rack looseness, and predicting the fault development trend, to obtain a looseness fault diagnosis result.

[0007] As a further solution of the present invention, the real-time vibration data set includes sampling rate information, storage timestamp information, and vibration amplitude data in the X and Y directions; the vibration data analysis record includes the energy of key vibration frequencies, environmental noise energy, and adjusted acquisition frequency band information; the state association analysis result includes the guide vane opening value, the working water head value, and the correlation coefficient between the vibration data and the hydraulic state; the abnormal vibration mode data includes the periodic analysis results of the abnormal vibration, the non-periodic analysis results, and the standard vibration mode comparison difference data; the looseness degree quantification data includes the peak frequency information of the abnormal vibration, the energy spectrum density information, and the vibration amplitude and phase analysis results; the looseness fault diagnosis result includes the looseness degree quantification value, the frame looseness severity level, and the fault development trend prediction result.

[0008] As a further solution of the present invention, based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction horizontal of the frame on the generator set are collected and recorded. The steps of obtaining the real-time vibration data set are specifically as follows: S101: Based on the real-time operation data of the generator, the operation parameters of the equipment are monitored in real time, including power, rotation speed and temperature, and the sampling frequency of the vibration data is adjusted according to the equipment status to generate a sampling frequency adjustment record; S102: Based on the sampling frequency adjustment record, collect and record the horizontal vibration data of the frame on the generator set in the X direction and the Y direction in real time to generate real-time frame vibration data; S103: According to the real-time frame vibration data, the time information of vibration data collection is recorded, and the data is formatted and standardized to generate a real-time vibration data set.

[0009] As a further solution of the present invention, based on the real-time vibration data set, by performing spectrum analysis on the vibration data, analyzing the amplitudes of multiple frequency components in the spectrum, identifying key frequency areas, and adjusting the frequency band of data acquisition to capture fault-related vibration characteristics, the steps of generating vibration data analysis records are specifically as follows: S201: Based on the real-time vibration data set, perform a baseline spectrum analysis of the vibration data, record the amplitudes of multiple frequency components, identify key frequency components of the vibration, and generate a spectrum analysis result; S202: Based on the spectrum analysis result, by comparing the energy of multiple key frequencies and environmental background noise, identifying the key frequency area, and generating a key frequency identification result; S203: Based on the critical frequency identification result, the frequency band setting of data collection is adjusted according to the identified critical frequency area, the capture of fault characteristics is optimized, and a vibration data analysis record is generated.

[0010] As a further solution of the present invention, according to the vibration data analysis record, combined with the real-time guide vane opening and working head data, the relationship between the vibration data and the hydraulic state is analyzed to obtain the state correlation analysis result in the following steps: S301: collecting the guide vane opening and working water head data of the generator set according to the vibration data analysis record, and generating a generator set data set; S302: Analyze the relationship between the vibration frequency data and the generator set operating parameters according to the generator set data set, calculate the correlation coefficient, and generate a correlation analysis result; S303: Based on the correlation analysis result, the relationship between the vibration data change and the hydraulic state change is evaluated to identify the vibration anomaly caused by the hydraulic state change and generate a state correlation analysis result.

[0011] As a further solution of the present invention, the specific formula for calculating the correlation coefficient is:

[0012] in, Represents the correlation coefficient between the vibration frequency data and the generator set operating parameters, indicating the strength and direction of the linear relationship between the two. Represents the vibration frequency value of a single measurement, Represents the average value of all measured vibration frequency values, Represents the corresponding single-measured generator operating parameter value, represents the average value of all measured generator operating parameters, An index used to identify and access multiple elements in a data set.

[0013] As a further solution of the present invention, according to the state association analysis result, by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with the standard vibration mode library, the abnormal vibration mode is identified, and the steps of generating abnormal vibration mode data are specifically as follows: S401: Analyze the vibration data based on the state association analysis result, identify periodic and non-periodic vibration modes, and record the characteristics of multiple modes, including vibration period and amplitude, to generate a vibration mode characteristic record; S402: According to the vibration pattern feature record, by comparing the identified vibration pattern with a preset standard vibration pattern library, identifying an abnormal vibration pattern, and generating abnormal vibration identification information; S403: Based on the abnormal vibration identification information, analyze the abnormal vibration pattern, calculate the frequency, amplitude and phase, evaluate the persistence and impact range of the abnormal pattern, and generate abnormal vibration pattern data.

[0014] As a further solution of the present invention, based on the abnormal vibration mode data, the frequency, amplitude and phase of the abnormal vibration are analyzed, the key features of the abnormal mode are calculated, the degree of looseness is quantified, and the steps of generating the quantified data of the degree of looseness are specifically as follows: S501: Based on the abnormal vibration mode data, analyzing the frequency of the abnormal vibration mode, and recording the vibration amplitude value, generating a frequency amplitude data record; S502: Based on the frequency amplitude data record, perform energy spectrum density calculation on multiple frequency components, evaluate total vibration energy, and generate vibration energy analysis results; S503: Based on the vibration energy analysis result, the calculated total vibration energy is compared with a preset energy threshold to quantify the degree of looseness of the frame and generate looseness quantification data.

[0015] As a further solution of the present invention, the steps of using the looseness quantification data to compare the looseness with a preset threshold value, evaluate the severity level of the rack looseness, and predict the fault development trend to obtain the looseness fault diagnosis result are specifically as follows: S601: using the looseness quantification data, by comparing the looseness value with a preset safety threshold, assessing the severity of the rack looseness, and generating a looseness risk score; S602: Based on the looseness risk score, analyze the change trend of the rack looseness risk, combine the equipment operation log and maintenance record, analyze and predict the development direction of the risk, and generate a fault development trend prediction; S603: Based on the fault development trend prediction, according to the looseness risk and fault trend, the equipment health status is evaluated and the fault is responded to, the safety and operation risk of the equipment operation are optimized, and the looseness fault diagnosis result is generated.

[0016] A generator set upper frame loose fault diagnosis system based on a frequency division algorithm, the generator set upper frame loose fault diagnosis system based on a frequency division algorithm is used to execute the generator set upper frame loose fault diagnosis method based on the frequency division algorithm, the system comprises: The rack status monitoring module monitors the operating status of the generator set in real time based on the real-time operating data of the generator, adjusts the sampling frequency of the vibration data according to the changes in power and speed, optimizes data storage and data integrity, and generates rack vibration collection data; The spectrum analysis module identifies key frequency components in the vibration data through spectrum analysis based on the rack vibration collection data, identifies key frequency areas through energy comparison, and adjusts the data collection frequency band to generate vibration characteristic data; The dependency analysis module uses the vibration characteristic data, combined with the real-time guide vane opening and working water head data, to analyze the relationship between the vibration data and the hydraulic state, and generates correlation analysis data; The abnormal vibration detection module analyzes the periodic and non-periodic change patterns of the vibration data based on the correlation analysis data, and compares them with a preset standard vibration pattern library to identify abnormal vibration patterns and generate abnormal pattern recognition records; The loosening risk assessment module uses the abnormal pattern recognition records to analyze the frequency, amplitude and phase of the abnormal vibration, calculates the key characteristics of the abnormal vibration, quantifies the degree of loosening and compares it with the preset threshold, evaluates the severity level of the loosening and predicts the fault development trend, and generates a loosening fault diagnosis result.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the vibration data sampling rate is dynamically adjusted by utilizing real-time operation data, the storage of irrelevant data is reduced, and it is ensured that fault signals are captured at critical moments, thereby improving the efficiency and accuracy of fault diagnosis. The key frequency areas are identified through spectrum analysis, and the data acquisition frequency band is adjusted to enhance the ability to capture fault-related vibration characteristics. The relationship between vibration data and hydraulic state is analyzed in combination with guide vane opening and working head data, thereby improving the accuracy of fault source identification. By analyzing abnormal vibration patterns and quantifying the degree of looseness, accurate data support is provided for equipment maintenance, thereby avoiding production interruptions and economic losses caused by equipment failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flow chart of S1 of the present invention; Figure 3 This is a detailed flow chart of S2 of the present invention; Figure 4 This is a detailed flow chart of S3 of the present invention; Figure 5 This is a detailed flow chart of S4 of the present invention; Figure 6 This is a detailed flow chart of S5 of the present invention; Figure 7 This is a detailed flow chart of S6 of the present invention; Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0021] See also Figure 1 The present invention provides a technical solution, a method for diagnosing a loose frame fault of a generator set based on a frequency division algorithm, comprising the following steps: S1: Based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction of the upper frame of the generator set is collected and recorded to obtain a real-time vibration data set; S2: Based on the real-time vibration data set, the vibration data is spectrally analyzed to analyze the amplitudes of multiple frequency components in the spectrum, identify the key frequency area, and adjust the frequency band of data acquisition to capture the fault-related vibration characteristics and generate vibration data analysis records; S3: According to the vibration data analysis records, combined with the real-time guide vane opening and working water head data, the relationship between the vibration data and the hydraulic state is analyzed to obtain the state correlation analysis results; S4: According to the state association analysis result, by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with the standard vibration mode library, the abnormal vibration mode is identified and the abnormal vibration mode data is generated; S5: Based on the abnormal vibration mode data, the frequency, amplitude and phase of the abnormal vibration are analyzed, the key features of the abnormal mode are calculated, the degree of looseness is quantified, and the quantitative data of the degree of looseness is generated; S6: Using the looseness quantification data, by comparing the looseness with the preset threshold, the severity level of the rack looseness is evaluated, and the fault development trend is predicted to obtain the looseness fault diagnosis result.

[0022] The real-time vibration data set includes sampling rate information, storage timestamp information, and vibration amplitude data in the X and Y directions. The vibration data analysis record includes the energy of key vibration frequencies, environmental noise energy, and adjusted acquisition frequency band information. The state correlation analysis results include the guide vane opening value, working water head value, and the correlation coefficient between vibration data and hydraulic state. The abnormal vibration mode data includes the periodic analysis results of abnormal vibration, the non-periodic analysis results, and the comparison difference data of standard vibration modes. The looseness degree quantitative data includes the peak frequency information of abnormal vibration, energy spectrum density information, vibration amplitude and phase analysis results. The looseness fault diagnosis results include the looseness degree quantitative value, the severity level of the frame looseness, and the fault development trend prediction results.

[0023] See also Figure 2 Based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction of the frame on the generator set are collected and recorded. The steps to obtain the real-time vibration data set are as follows: S101: Based on the real-time operation data of the generator, the operation parameters of the equipment are monitored in real time, including power, rotation speed and temperature, and the sampling frequency of the vibration data is adjusted according to the equipment status to generate a sampling frequency adjustment record; In sub-step S101, based on the real-time operating data of the generator, the operating parameters of the equipment are monitored in real time, including the detection of multiple key parameters such as power, speed and temperature. The operating parameters are crucial to the overall performance and safe operation of the equipment. By real-time monitoring of the operating parameters, the operating status of the equipment is evaluated in real time. When the operating status of the equipment is abnormal, the sampling frequency of the vibration data is adjusted immediately according to the current operating efficiency and safety requirements of the equipment. The adjusted sampling frequency of the vibration data is used to capture signs of mechanical failure and performance degradation. Without affecting the operation of the equipment, potential problems can be diagnosed early to avoid equipment failure and reduce maintenance costs. The generated sampling frequency adjustment record provides data support for subsequent maintenance and monitoring.

[0024] S102: Based on the sampling frequency adjustment record, the horizontal vibration data of the frame on the generator set in the X direction and the Y direction is collected and recorded in real time to generate real-time frame vibration data; In sub-step S102, based on the sampling frequency adjustment record, the horizontal vibration data of the X-direction and Y-direction of the frame on the generator set is collected and recorded in real time through the Internet of Things vibration sensor. The sensor is used to capture the tiny vibrations of the frame in two directions. The real-time collected vibration data provides a rich source of original data for subsequent analysis. Through the real-time recording of the bomb data, the vibration condition of the equipment is monitored in real time, and any abnormalities or trend changes are discovered in time. The generated real-time frame vibration data includes vibration intensity and frequency information. The specific time of each data collection is marked by a timestamp, which provides a basis for data tracking and historical comparison.

[0025] S103: According to the real-time frame vibration data, record the time information of vibration data collection, and format and standardize the data to generate a real-time vibration data set; In sub-step S103, the time information of vibration data collection is recorded according to the real-time rack vibration data, and the data is formatted and standardized. The formatting process includes converting the vibration data into a unified format for easy storage and analysis. The standardization process is used to ensure that data from different sensors can be compared with each other and eliminate the differences in data between multiple devices. The generated real-time vibration data set has a high degree of consistency and reliability, making the data set suitable for data analysis and vibration diagnosis, and providing data support for equipment maintenance and fault prediction.

[0026] See also Figure 3 , based on the real-time vibration data set, by performing spectrum analysis on the vibration data, analyzing the amplitudes of multiple frequency components in the spectrum, identifying the key frequency area, and adjusting the frequency band of data acquisition to capture the fault-related vibration characteristics, the steps of generating vibration data analysis records are as follows: S201: Based on the real-time vibration data set, perform baseline spectrum analysis on the vibration data, record the amplitudes of multiple frequency components, identify key frequency components of the vibration, and generate spectrum analysis results; In sub-step S201, a baseline spectrum analysis of the vibration data is performed based on the real-time vibration data set. The multiple frequency components in the data set are checked and recorded, including the amplitude of each frequency component. The amplitude recording is used for subsequent vibration analysis. The baseline spectrum analysis is used to reveal the standard vibration mode and abnormal frequency of the equipment during operation. Through this analysis, the key frequency components in the equipment vibration are identified. The key frequency components are used to identify associated mechanical problems, which improves the accuracy of fault diagnosis and provides a scientific basis for equipment maintenance. The generated spectrum analysis results provide a data basis for subsequent analysis and fault prediction.

[0027] S202: Based on the spectrum analysis result, by comparing the energy of multiple key frequencies and environmental background noise, the key frequency area is identified and a key frequency identification result is generated; In the above content, based on the spectrum analysis results, the energy of the key frequency is compared with the ambient background noise energy. According to the formula , calculate the energy ratio of the key frequency to the noise; In the formula, Represents the energy ratio of the key frequency to the noise, Represents the total energy of the key frequency components, Represents the energy of ambient noise; Detailed explanation of the formula and the process of formula calculation and derivation: Assuming that the energies of the key frequency components are 20, 15, 25, 10, and 30 dB respectively, and the ambient noise energy is 60 dB, calculate the energy ratio of the key frequency to the noise:

[0028] The results show that the total energy of the key frequency components is 1.67 times the energy of the ambient noise. The numerical value indicates the significance of the vibration source and is used to confirm the importance of the key frequency area.

[0029] S203: Based on the key frequency identification result, adjusting the frequency band setting of data collection according to the identified key frequency area, optimizing the capture of fault characteristics, and generating a vibration data analysis record; In sub-step S203, based on the key frequency identification results, the frequency band setting of data acquisition is adjusted to improve the accuracy and efficiency of vibration data analysis. The adjusted frequency band setting will target the identified key frequency areas to ensure that the data acquisition process focuses on potential fault signals and reduces interference from irrelevant data. The optimized vibration data is used to optimize subsequent fault analysis and prediction, and the generated vibration data analysis records provide accurate data support for equipment maintenance and fault handling.

[0030] See also Figure 4 According to the vibration data analysis records, combined with the real-time guide vane opening and working head data, the relationship between the vibration data and the hydraulic state is analyzed, and the steps to obtain the state correlation analysis results are as follows: S301: According to the vibration data analysis record, the guide vane opening and working water head data of the generator set are collected to generate a generator set data set; In sub-step S301, the guide vane opening and working head data of the generator set are collected based on the vibration data analysis records. The data are key parameters for the operating efficiency and safe operation of the generator set. The guide vane opening determines the amount of water flowing through the turbine, while the working head directly affects the power generation efficiency. By recording the operating data in real time, a generator set data set is generated. The data set includes vibration data and key operating parameters of power generation equipment, providing a comprehensive perspective for subsequent operating status analysis and fault prediction, ensuring the accuracy and practicality of data analysis, and providing strong data support for equipment management and optimization decisions.

[0031] S302: Analyze the relationship between the vibration frequency data and the generator set operating parameters according to the generator set data set, calculate the correlation coefficient, and generate a correlation analysis result; The specific formula for calculating the correlation coefficient is:

[0032] in, Represents the correlation coefficient between the vibration frequency data and the generator set operating parameters, indicating the strength and direction of the linear relationship between the two. Represents the vibration frequency value of a single measurement, Represents the average value of all measured vibration frequency values, Represents the corresponding single-measured generator operating parameter value, represents the average value of all measured generator operating parameters, An index used to identify and access multiple elements in a data set.

[0033] formula:

[0034] Detailed explanation of the formula and the process of formula calculation and derivation: The formula is used to calculate the correlation coefficient between the vibration frequency data and the generator set operating parameters , evaluate the strength of the linear relationship between the two, and understand the interaction between the change in vibration frequency and the change in the generator operating state; Parameter meaning and setting value: Represents the vibration frequency value of a single measurement, assuming that Hz, is the average value of the vibration frequency, calculated as Hz, Indicates the corresponding single-measured generator operating parameter value, assuming it is power output, and the value is kW, is the average value of the generator power output, calculated as kW; Substitute the parameters into the formula for calculation:

[0035]

[0036]

[0037]

[0038] result It shows that there is a positive correlation between vibration frequency and generator power output, indicating that an increase in generator power output is usually accompanied by an increase in vibration frequency, which is used for monitoring equipment health and early fault detection.

[0039] S303: Based on the correlation analysis result, by evaluating the relationship between the vibration data change and the hydraulic state change, the vibration anomaly caused by the hydraulic state change is identified, and the state correlation analysis result is generated; In sub-step S303, based on the correlation analysis results, the relationship between the vibration data changes and the hydraulic state changes is evaluated, and the vibration anomalies caused by the hydraulic state changes are identified. The changes in the hydraulic state indicate potential equipment problems or changes in operating efficiency. The identified correlation is used to discover and solve the problems that lead to equipment failures at an early stage. The generated state correlation analysis results reflect the direct correlation between the hydraulic state changes and the vibration data anomalies. The analysis results are of great significance for formulating more effective maintenance strategies and optimizing equipment operations.

[0040] See also Figure 5 According to the state association analysis results, the abnormal vibration mode is identified by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with the standard vibration mode library. The specific steps for generating abnormal vibration mode data are as follows: S401: Analyze the vibration data based on the state association analysis result, identify periodic and non-periodic vibration modes, and record the characteristics of multiple modes, including vibration period and amplitude, to generate a vibration mode characteristic record; In the above content, the vibration data is analyzed for periodic and non-periodic patterns according to the formula and Calculate vibration amplitude and Phase ; In the formula, and Represent the peak values ​​of periodic and non-periodic vibrations, respectively; Detailed explanation of the formula and the process of formula calculation and derivation: Assuming the peak value of periodic vibration mm / s², non-periodic vibration peak value mm / s², substitute the value into the formula to calculate the vibration amplitude: mm / s² Calculate the phase: radian The results show that by synthesizing the characteristics of periodic and non-periodic vibration modes, the dynamic characteristics of vibration can be accurately described, providing a basis for subsequent abnormal vibration mode recognition.

[0041] S402: According to the vibration pattern feature record, by comparing the identified vibration pattern with a preset standard vibration pattern library, identifying the abnormal vibration pattern, and generating abnormal vibration identification information; In sub-step S402, based on the vibration pattern feature record, the identified vibration pattern is compared with a preset standard vibration pattern library, which includes the vibration characteristics that the equipment should have under various operating conditions. Abnormal vibration patterns that are inconsistent with the norm are identified. Vibration records that deviate significantly from the standard pattern during the comparison process are considered abnormal, and the causes of the abnormality are analyzed, including equipment wear, installation errors, or external influences. By discovering and recording abnormal vibration patterns, abnormal vibration identification information is generated for formulating maintenance and adjustment strategies for equipment operation.

[0042] S403: Based on the abnormal vibration identification information, analyze the abnormal vibration pattern, calculate the frequency, amplitude and phase, evaluate the persistence and impact range of the abnormal pattern, and generate abnormal vibration pattern data; In sub-step S403, based on the abnormal vibration identification information, the abnormal vibration pattern is analyzed, which involves calculating the specific frequency, amplitude and phase of each identified pattern. The frequency measurement is used to reveal the speed of the vibration, the amplitude indicates the intensity of the vibration, and the phase provides the position information of the vibration waveform relative to the reference point. The measurement results are crucial to understanding the nature of the abnormal vibration. According to the change of the vibration pattern over time, the persistence of the abnormal pattern is evaluated, that is, whether it is a temporary phenomenon or has a trend of continuous deterioration. The assessment of the scope of impact focuses on determining the equipment areas that may be affected by the abnormal vibration and the possible system-level consequences, and how to affect the overall operating efficiency and safety of the equipment. The generated abnormal vibration pattern data records the analysis results of the frequency, amplitude, phase, persistence and scope of impact of each abnormal pattern, providing an in-depth understanding of the current status of the equipment and providing a scientific basis for formulating targeted maintenance strategies and immediate intervention measures.

[0043] See also Figure 6 Based on the abnormal vibration mode data, the frequency, amplitude and phase of the abnormal vibration are analyzed, the key features of the abnormal mode are calculated, and the degree of looseness is quantified. The specific steps for generating the quantitative data of the degree of looseness are as follows: S501: Based on the abnormal vibration mode data, the frequency of the abnormal vibration mode is analyzed, and the vibration amplitude value is recorded to generate a frequency amplitude data record; In sub-step S501, the frequency of the abnormal vibration mode is analyzed based on the abnormal vibration mode data. The analysis focuses on identifying and recording the frequency characteristics and corresponding vibration amplitude values ​​of each abnormal vibration mode. The speed and periodic characteristics of the vibration are determined by accurately measuring the frequency. The vibration amplitude value is recorded to measure the intensity of the vibration in the maximum fluctuation. The frequency and amplitude data records are used to analyze the impact of the vibration mode and potential mechanical problems. The analysis helps to determine the frequencies associated with mechanical failures. The generated frequency amplitude data records will contain detailed amplitude values ​​for each identified frequency, providing basic data for energy analysis.

[0044] S502: Based on the frequency amplitude data records, energy spectrum density calculation is performed on multiple frequency components, and total vibration energy is evaluated to generate vibration energy analysis results; In sub-step S502, based on the frequency amplitude data records, the energy spectrum density of multiple frequency components is calculated, including the precise quantification of the vibration energy of each frequency component. The energy spectrum density calculation is used to evaluate the energy contribution of each frequency component and reveal the total energy layout of the vibration. By calculating and comparing the energy of different frequency components, the vibration sources that have a significant impact on the operation of the equipment are identified. Evaluating the total vibration energy is an important step in diagnosing the overall health status of the system and is used to indicate the degree of fatigue or damage to the equipment. The generated vibration energy analysis results describe the energy density and overall energy distribution of each frequency component, providing key technical support for formulating mitigation measures and making maintenance decisions.

[0045] S503: Based on the vibration energy analysis result, the calculated total vibration energy is compared with a preset energy threshold, the degree of looseness of the frame is quantified, and looseness quantification data is generated; In the above content, based on the vibration energy analysis results, the calculated total vibration energy is compared with the preset energy threshold, according to the formula , calculate the quantitative value of the looseness of the rack; In the formula, Represents the percentage of looseness. represents the calculated total vibration energy, represents the preset energy threshold; Detailed explanation of the formula and the process of formula calculation and derivation: Assuming that the total vibration energy obtained through energy spectrum density calculation is 500 units, and the preset energy threshold is 400 units, the calculation process is as follows:

[0046] The results show that the total vibration energy exceeds the preset threshold by 25%. The value indicates that the rack may have a serious loose problem and requires further inspection and maintenance to prevent potential mechanical failure. The looseness quantification data provides an important basis for evaluating the current status of the equipment and taking appropriate maintenance measures.

[0047] See also Figure 7 , using the looseness quantitative data, by comparing the looseness with the preset threshold, evaluating the severity level of the rack looseness, and predicting the fault development trend, the steps to obtain the looseness fault diagnosis result are as follows: S601: using the looseness quantification data, by comparing the looseness value with a preset safety threshold, assessing the severity of the rack looseness, and generating a looseness risk score; In sub-step S601, the looseness degree is quantified using data obtained by accurately measuring the fixed state and position deviation of each component of the equipment. The severity of the rack looseness is evaluated by comparing the looseness degree value with the preset safety threshold. The evaluation is critical to maintaining the structural integrity and operational safety of the equipment. The safety threshold is set based on historical data and engineering standards. Readings exceeding this threshold are used to indicate potential structural problems and impending failures. The generated looseness risk score will be quantified based on the difference between the looseness degree and the safety threshold, providing an intuitive risk indicator for the equipment maintenance team.

[0048] S602: Based on the looseness risk score, analyze the change trend of the rack looseness risk, combine the equipment operation log and maintenance record, analyze and predict the development direction of the risk, and generate a fault development trend prediction; In sub-step S602, based on the looseness risk score, the changing trend of the rack looseness risk is analyzed. The analysis is combined with the equipment operation log and maintenance record to evaluate how the looseness risk develops over time and its possible future direction. The operation log records the operating status of the equipment and any abnormal conditions, and the maintenance record records the maintenance activities and component replacements performed in the past. Historical data is the key to analyzing the development trend of risks. It is used to reveal repeated problems and long-term unresolved problems. Predicting the development direction of faults helps plan maintenance and preventive measures in advance, optimize resource allocation and maintenance plans, and the generated fault development trend prediction will provide strategic recommendations for equipment management to ensure that the equipment can operate continuously and stably.

[0049] S603: Based on the prediction of fault development trend, the equipment health status is evaluated and the fault is responded to according to the looseness risk and fault trend, the safety and operation risk of the equipment operation are optimized, and the looseness fault diagnosis result is generated; In the above content, based on the prediction of fault development trend, combined with looseness risk and fault trend, the health status of the equipment is evaluated and the fault response is performed according to the formula , calculate the health index of the device; In the formula, Represents the health index of the device. represents the loose risk score, represents the failure tendency score; Detailed explanation of the formula and the process of formula calculation and derivation: Assuming a loosening risk score of 20% and a failure tendency score of 30%, the calculation process is as follows:

[0050] This shows that the equipment's health index is 50%, indicating a medium risk level. Timely maintenance and parts replacement are required to prevent equipment damage. The generated loose fault diagnosis results will illustrate the current health status of the equipment and the required maintenance measures, which is crucial for the maintenance team to develop effective maintenance plans and respond to faults.

[0051] See also Figure 8 , A generator set upper frame loose fault diagnosis system based on a frequency division algorithm, a generator set upper frame loose fault diagnosis system based on a frequency division algorithm is used to execute the generator set upper frame loose fault diagnosis method based on a frequency division algorithm, and the system includes: The rack status monitoring module monitors the operating status of the generator set in real time based on the real-time operating data of the generator, adjusts the sampling frequency of the vibration data according to the changes in power and speed, optimizes data storage and data integrity, and generates rack vibration collection data; The spectrum analysis module is based on the rack vibration data collection, identifies the key frequency components in the vibration data through spectrum analysis, identifies the key frequency area through energy comparison, and adjusts the data collection frequency band to generate vibration characteristic data; The dependency analysis module uses vibration characteristic data, combined with real-time guide vane opening and working head data, to analyze the relationship between vibration data and hydraulic state and generate correlation analysis data; The abnormal vibration detection module analyzes the periodic and non-periodic change patterns of vibration data based on correlation analysis data, and compares it with the preset standard vibration pattern library to identify abnormal vibration patterns and generate abnormal pattern recognition records; The loosening risk assessment module uses abnormal pattern recognition records to analyze the frequency, amplitude and phase of abnormal vibration, calculate the key characteristics of abnormal vibration, quantify the degree of loosening and compare it with the preset threshold, evaluate the severity level of loosening and predict the development trend of faults, and generate loosening fault diagnosis results.

[0052] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A fault diagnosis method for loose frame of generator set based on frequency division algorithm, characterized in that: The following steps are involved: Based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction of the upper frame of the generator set is collected and recorded to obtain a real-time vibration data set; Based on the real-time vibration data set, by performing spectrum analysis on the vibration data, analyzing the amplitudes of multiple frequency components in the spectrum, identifying key frequency areas, and adjusting the frequency band of data acquisition to capture fault-related vibration characteristics, a vibration data analysis record is generated; According to the vibration data analysis record, combined with the real-time guide vane opening and working water head data, the relationship between the vibration data and the hydraulic state is analyzed to obtain the state correlation analysis result; According to the state association analysis result, by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with a standard vibration mode library, abnormal vibration modes are identified and abnormal vibration mode data is generated; Based on the abnormal vibration mode data, the frequency, amplitude and phase of the abnormal vibration are analyzed, key features of the abnormal mode are calculated, the degree of looseness is quantified, and looseness degree quantification data is generated; By using the looseness quantification data, the severity level of the rack looseness is evaluated by comparing the looseness degree with a preset threshold, and the fault development trend is predicted to obtain a looseness fault diagnosis result.

2. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: The real-time vibration data set includes sampling rate information, storage timestamp information, and vibration amplitude data in the X and Y directions; the vibration data analysis record includes the energy of key vibration frequencies, environmental noise energy, and adjusted acquisition frequency band information; the state association analysis result includes the guide vane opening value, the working water head value, and the correlation coefficient between the vibration data and the hydraulic state; the abnormal vibration mode data includes the periodic analysis results of the abnormal vibration, the non-periodic analysis results, and the standard vibration mode comparison difference data; the looseness degree quantitative data includes the peak frequency information of the abnormal vibration, the energy spectrum density information, and the vibration amplitude and phase analysis results; the looseness fault diagnosis result includes the looseness degree quantitative value, the frame looseness severity level, and the fault development trend prediction result.

3. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: Based on the real-time operation data of the generator, the sampling rate of the vibration data is adjusted according to the operation status of the equipment, and the vibration data of the X-direction and Y-direction of the frame on the generator set are collected and recorded. The specific steps to obtain the real-time vibration data set are as follows: Based on the real-time operation data of the generator, the operating parameters of the equipment, including power, speed and temperature, are monitored in real time, and the sampling frequency of the vibration data is adjusted according to the equipment status to generate a sampling frequency adjustment record; Based on the sampling frequency adjustment record, the horizontal vibration data of the frame on the generator set in the X direction and the Y direction are collected and recorded in real time to generate real-time frame vibration data; According to the real-time frame vibration data, the time information of vibration data collection is recorded, and the data is formatted and standardized to generate a real-time vibration data set.

4. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: Based on the real-time vibration data set, by performing spectrum analysis on the vibration data, analyzing the amplitudes of multiple frequency components in the spectrum, identifying key frequency areas, and adjusting the frequency band of data acquisition to capture fault-related vibration characteristics, the steps of generating vibration data analysis records are specifically as follows: Based on the real-time vibration data set, a baseline spectrum analysis of the vibration data is performed, the amplitudes of multiple frequency components are recorded, key frequency components of the vibration are identified, and a spectrum analysis result is generated; Based on the spectrum analysis result, by comparing the energy of multiple key frequencies and environmental background noise, the key frequency area is identified and a key frequency identification result is generated; Based on the critical frequency identification result, the frequency band setting of data acquisition is adjusted according to the identified critical frequency area, the capture of fault characteristics is optimized, and the vibration data analysis record is generated.

5. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: According to the vibration data analysis record, combined with the real-time guide vane opening and working water head data, the relationship between the vibration data and the hydraulic state is analyzed to obtain the state correlation analysis result in the following steps: According to the vibration data analysis record, the guide vane opening and working water head data of the generator set are collected to generate a generator set data set; According to the generator set data set, analyzing the relationship between the vibration frequency data and the generator set operating parameters, calculating the correlation coefficient, and generating a correlation analysis result; Based on the correlation analysis results, the relationship between the vibration data changes and the hydraulic state changes is evaluated to identify vibration anomalies caused by hydraulic state changes and generate state correlation analysis results.

6. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 5 is characterized in that: The specific formula for calculating the correlation coefficient is: in, Represents the correlation coefficient between the vibration frequency data and the generator set operating parameters, indicating the strength and direction of the linear relationship between the two. Represents the vibration frequency value of a single measurement, Represents the average value of all measured vibration frequency values, Represents the corresponding single-measured generator operating parameter value, represents the average value of all measured generator operating parameters, An index used to identify and access multiple elements in a data set.

7. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: According to the state association analysis result, by analyzing the periodic and non-periodic vibration modes of the frame and comparing them with the standard vibration mode library, the abnormal vibration mode is identified and the steps of generating abnormal vibration mode data are specifically as follows: Based on the state association analysis results, analyzing the vibration data, identifying periodic and non-periodic vibration modes, and recording characteristics of multiple modes, including vibration period and amplitude, to generate a vibration mode characteristic record; According to the vibration pattern feature record, by comparing the identified vibration pattern with a preset standard vibration pattern library, identifying the abnormal vibration pattern, and generating abnormal vibration identification information; Based on the abnormal vibration identification information, the abnormal vibration pattern is analyzed, the frequency, amplitude and phase are calculated, the persistence and impact range of the abnormal pattern are evaluated, and the abnormal vibration pattern data is generated.

8. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1 is characterized in that: Based on the abnormal vibration mode data, the frequency, amplitude and phase of the abnormal vibration are analyzed, the key features of the abnormal mode are calculated, and the degree of looseness is quantified. The steps of generating the quantitative data of the degree of looseness are specifically as follows: Based on the abnormal vibration mode data, analyzing the frequency of the abnormal vibration mode, and recording the vibration amplitude value, generating a frequency amplitude data record; Based on the frequency amplitude data record, energy spectrum density calculation is performed on multiple frequency components, and total vibration energy is evaluated to generate vibration energy analysis results; Based on the vibration energy analysis result, the calculated total vibration energy is compared with a preset energy threshold to quantify the degree of looseness of the frame and generate looseness quantification data.

9. The generator set upper frame loose fault diagnosis method based on frequency division algorithm according to claim 1, characterized in that: By using the looseness quantification data, by comparing the looseness with a preset threshold, evaluating the severity level of the rack looseness, and predicting the fault development trend, the steps of obtaining the looseness fault diagnosis result are specifically as follows: Using the looseness quantification data, by comparing the looseness value with a preset safety threshold, the severity of the rack looseness is evaluated to generate a looseness risk score; Based on the looseness risk score, analyze the changing trend of the rack looseness risk, combine the equipment operation log and maintenance record, analyze and predict the development direction of the risk, and generate a fault development trend prediction; Based on the fault development trend prediction, the equipment health status is evaluated and the fault is responded to according to the looseness risk and fault trend, the safety and operation risk of the equipment operation are optimized, and the looseness fault diagnosis result is generated.

10. The generator set upper frame loose fault diagnosis system based on frequency division algorithm is characterized by: According to the method for diagnosing loose faults on a generator set based on a frequency division algorithm according to any one of claims 1 to 9, the system comprises: The rack status monitoring module monitors the operating status of the generator set in real time based on the real-time operating data of the generator, adjusts the sampling frequency of the vibration data according to the changes in power and speed, optimizes data storage and data integrity, and generates rack vibration collection data; The spectrum analysis module identifies key frequency components in the vibration data through spectrum analysis based on the rack vibration collection data, identifies key frequency areas through energy comparison, and adjusts the data collection frequency band to generate vibration characteristic data; The dependency analysis module uses the vibration characteristic data, combined with the real-time guide vane opening and working water head data, to analyze the relationship between the vibration data and the hydraulic state, and generates correlation analysis data; The abnormal vibration detection module analyzes the periodic and non-periodic change patterns of the vibration data based on the correlation analysis data, and compares them with a preset standard vibration pattern library to identify abnormal vibration patterns and generate abnormal pattern recognition records; The loosening risk assessment module uses the abnormal pattern recognition records to analyze the frequency, amplitude and phase of the abnormal vibration, calculates the key characteristics of the abnormal vibration, quantifies the degree of loosening and compares it with the preset threshold, evaluates the severity level of the loosening and predicts the fault development trend, and generates a loosening fault diagnosis result.

Images