A nuclear power steam turbine low-pressure last-stage blade online monitoring and diagnosing method, device, system, medium and product based on blade tip timing

Abstract

The present application belongs to the technical field of nuclear power turbine generator set blade monitoring and diagnosis, and particularly relates to a nuclear power turbine low-pressure last-stage blade online monitoring and diagnosis method, device, equipment, medium and product based on blade tip timing. A rotating speed synchronous sensor is arranged, a monitoring probe of the rotating speed synchronous sensor is arranged in a direction facing a rotating blade disc, a groove is arranged on an end surface disc rim of the rotating blade disc at a position corresponding to the monitoring probe of the rotating speed synchronous sensor, and a plurality of blade tip timing sensors are arranged on a cylinder guide ring at positions corresponding to the rotating blade disc in a radial direction. The beneficial effect is that the time when a rotating blade in a nuclear power turbine reaches a blade tip timing sensor is measured based on a blade tip timing principle, a measured time pulse signal is converted into a blade tip amplitude signal, a vibration parameter of the rotating blade of the nuclear power turbine is extracted, and a basis is provided for vibration evaluation, life assessment and fault diagnosis of the rotating blade in the nuclear power turbine.

Description

Technical Field

[0001] This invention belongs to the field of nuclear power turbine generator set blade monitoring and diagnosis technology, specifically relating to a method, device, equipment, medium and product for online monitoring and diagnosis of low-pressure last-stage blades of nuclear power turbines based on tip timing. Background Technology

[0002] The low-pressure last-stage blades are critical and sensitive components of nuclear power turbines, being slender and flexible structural parts. During high-speed rotation, they are subjected to a more complex and severe wet steam environment than thermal power units, enduring a strong coupling of static, alternating, and random loads, including centrifugal force, thermal stress, steam flow impact force, and bending torsional stress. In addition, due to salt accumulation, scaling, and manufacturing and installation errors, low-pressure last-stage blades are highly susceptible to various failures, directly impacting the safety of nuclear power turbines.

[0003] According to research statistics, when a low-pressure blade malfunctions, nuclear power plants lose an average of 2.38 million megawatt-hours. The production losses due to unplanned shutdowns are equivalent to approximately $100 million annually for nuclear power plants, and these risk costs do not include maintenance expenses. Furthermore, domestic turbine accident statistics show that blade accidents account for over 30% of all turbine accidents. Damage to low-pressure last-stage blades mainly includes fatigue damage, surface pitting, water erosion, corrosion, stress corrosion, and erosion of the blade's exhaust edge back arc, with fatigue damage being the primary cause of blade failure.

[0004] Fatigue damage is directly related to vibration and unit operating parameters. Accumulated fatigue damage will eventually lead to blade fracture. Causes of fatigue include improper blade structure design, poor assembly quality, inappropriate operating parameters, pitting corrosion, water erosion, processing and material defects, and poor installation quality. Although the design life of turbine blades is typically hundreds of thousands of operating hours, the above factors often begin with localized material property loss or structural changes, gradually altering the overall mechanical properties of the blade and even the impeller. Over time, microcracks develop, and from the initiation of these cracks, they propagate rapidly, leading to blade fracture. This significantly shortens the blade's service life and seriously threatens the safe and reliable operation of the unit. Therefore, reducing the loss of the terminal blades and ensuring their safety is of great safety and economic importance. Early detection of defects is key to solving these problems. Summary of the Invention

[0005] The purpose of this invention is to provide a method, device, system, medium, and product for online monitoring and diagnosis of the low-pressure last-stage blades of nuclear power turbines based on tip timing, which can realize accurate and reliable online monitoring and fault diagnosis of the last-stage blades of nuclear power turbines.

[0006] The technical solution of the present invention is as follows: an online monitoring and diagnostic device for the low-pressure last stage blade of a nuclear power turbine based on blade tip timing, wherein a speed synchronization sensor is provided, the monitoring probe of the speed synchronization sensor is set facing the rotating blade disk; and a groove is opened on the end face edge of the rotating blade disk at the position corresponding to the monitoring probe of the speed synchronization sensor; at the same time, several blade tip timing sensors are arranged on the cylinder guide ring at the circumferential position corresponding to the radial direction of the rotating blade disk.

[0007] A method for online monitoring and diagnosis of low-pressure last-stage blades in nuclear power steam turbines based on tip timing includes the following steps: Step 1: Measure the arrival time t of two adjacent key phase pulses using a speed synchronization sensor. n and t n+1 Determine the time T for the rotating blade disk to complete one revolution. n =t n+1 -t n And the rotational speed of the nuclear power turbine impeller ω=60 / T n ; Step 2: Determine the order in which each blade passes the tip timing sensor based on the relative position of each blade on the rotating blade disk and the rotational speed synchronization sensor, and determine the number of each blade. ， Assumption t n At what time, the blade numbered by the tip timing sensor is b The blade numbers subsequently detected by the tip timing sensor are as follows: b、b+ 1. b +2..., the number of the blades that have already been passed is... b- 1 、b- 2. b -3...; Step 3: Based on the rotational speed of the nuclear power turbine impeller and the determined blade numbering... b Based on the key phase pulse of the speed synchronization sensor, the theoretical arrival time of each blade when it is not vibrating and passes through the blade tip timing sensor is determined. And the actual arrival time of each blade through the blade tip timing sensor during vibration. The theoretical arrival time of the blade tip timing sensor when the blade is not vibrating. Calculations were performed, assuming the bond phase pulse time was... t n If the impeller has N blades per revolution, then the time interval between adjacent blades passing the blade tip timing sensor when the blades are not vibrating is (t... n+1 -t n ) / N, b The theoretical arrival time of blade number 1 after passing the tip timing sensor is: t nb, b The theoretical arrival time of blade +1 passing the tip timing sensor is: t nb +(t n+1 -t n ) / N, the actual arrival time of the blade tip timing sensor during blade vibration. This is obtained through data collection; Step 4: Based on the theoretical arrival time The actual arrival time The time required for the rotating blade disk to rotate one revolution The vibration displacement of each blade is determined by combining the distance from the blade tip to the center of the rotating blade disk.

[0008] In step 4, based on the theoretical arrival time... The actual arrival time The time required for the rotating blade disk to rotate one revolution Based on the distance from the blade tip to the center of the rotating blade disk, the vibration displacement of each blade is calculated as follows: , In the formula, r This refers to the distance from the tip of the blade to the center of the rotating blade disk. This refers to the actual arrival time of the blade; This refers to the theoretical arrival time of the blade. This refers to the time it takes for the impeller to rotate once. This represents the blade vibration displacement; subscript n Indicates the current lap number; subscript b Number the blades; subscript s This is the serial number of the leaf tip timing sensor.

[0009] An online monitoring and diagnostic system for the low-pressure last stage blade of a nuclear power steam turbine based on tip timing includes a data acquisition module, a data validity judgment module, a data cleaning module, a speed calculation module, a vibration calculation module, a vibration analysis module, a condition assessment module, and a condition early warning module. The data acquisition module is used to acquire the arrival time of the impeller blades and the arrival time of the key phase; The data validity assessment module determines the validity of the blade arrival time data. The data cleaning module cleans the arrival time data of the blades that were mistakenly triggered. The rotational speed calculation module solves the rotational speed measurement matrix constructed by the arrival time of the key phase, and then calculates the rotational speed of the rotor blades; The vibration calculation module calculates the blade vibration displacement based on the blade arrival time, the key phase arrival time, the blade rotation speed, the blade tip rotation radius, and the angle between the blade and the key phase. The vibration analysis module analyzes and processes the blade vibration displacement, extracting the steady-state and transient displacements of the blade. The state assessment module reconstructs the transient and steady-state displacements of the blade into the blade stress, and performs blade state assessment by comparing the blade stress with stress thresholds for different states. The status warning module is used to send feedback information based on the blade status assessment results. When the blade is in a healthy state, the status warning module does not send warning information. When the blade is in a sub-healthy state, the status warning module issues a warning. When the blade is in a faulty state, the status warning module issues an alarm.

[0010] The data acquisition module installs five blade vibration sensors in the diaphragm above the last-stage blades of the turbine, simultaneously collecting vibration data from the five sensors. A blade tip timing sensor is arranged on the exhaust guide ring of the last-stage turbine to measure the time t when the blade reaches the blade tip timing sensor. b,n A key phase is installed on the turbine rotor, and a key phase timing sensor is placed on top of the key phase to measure the time t when the key phase arrives at the key phase timing sensor. o,n The subscript b represents the blade number, the subscript o represents the bond phase, and the subscript n represents the number of revolutions.

[0011] The data validity assessment module's assessment process includes: The period of one rotation of the rotor is: T=t o,n+1 -t o,n (1), The arrival time window width for each blade during this cycle is: t w =T / n b (2), n b The number of blades; The range of blade arrival times is: t o,n +(b-1)t w ≤t b,n ≤t o,n +bt w (3), In the formula, t o,n Let t be the arrival time of the bond phase in the nth rotational cycle. o,n+1 Let n be the arrival time of the (n+1)th rotating periodic bond phase. b b is the number of blades on the tested blade disk and b is the blade number; the validity of the blade arrival time data is determined by equation (3).

[0012] The data cleaning module cleans the arrival time data of the blades that were mistakenly triggered, including: If there are multiple arrival time signals or no blade arrival time signal within the range of values ​​of equation (3), this is considered a false triggering of the blade tip timing signal. The method for cleaning false triggering data is as follows: The frequency conversion of the bond phase is: f o =1 / (t o,n+1 -t o,n (4), The actual passing frequency between the arrival time signals of two adjacent blades is: f o act =1 / (t b+1,n -t b,n (5), If f o act >n b f b If there are redundant false triggering signals in the blade arrival time signal, the cleaning method in this case is to directly remove the redundant arrival time signal corresponding to this passing frequency. If f o act <n b f b If the blade arrival time signal contains a missing false trigger signal, the cleaning method in this case is to supplement the missing arrival time signal according to the key phase frequency. The supplementation method is as follows: t b+1,n =t b,n +1 / (n b f o ), t b,n =t b+1,n -1 / (n b f o ), In the formula, t b,n Let t be the arrival time of blade b in the nth rotation cycle. b+1,n This represents the arrival time of blade b+1 in the (n+1)th rotation cycle.

[0013] The speed calculation module solves the speed measurement matrix constructed based on the key phase arrival time, and then calculates the speed of the rotor blades. The method is as follows: f=1 / (t o,n+1 -t o,n (7), The rotational speed calculated by equation (7) is the average speed of one revolution of the blade. The calculation method for the rotational speed under all operating conditions is as follows: The rotational speed of the turbine rotor blades is expressed as: f(t) = f0 + Σ1 c a c t c (8), In the formula, f0 is the initial rotational speed of the blade when the key phase reaches the key phase timing sensor, Σ1 c a c t c Σ1 represents the change in rotational speed over time. c a c t c =0 indicates constant speed operation, c=1 indicates linear speed change operation, and c>1 indicates nonlinear speed change operation. The rotational speed measurement matrix is ​​constructed based on the arrival time of the key phase as follows: (9), Equation (9) can be written in matrix form as follows: C=MF (10) In the formula, C is a matrix related to the number of blade rotations, and the value of c is related to the rate of change of rotational speed. In this invention, it is taken as 4. M is a matrix related to the arrival time of the bond phase, and F is the rotational speed matrix, which is obtained by the least squares method. F=(M T M) -1 M T C (11), After obtaining the rotational speed matrix, the blade rotational speed is calculated according to equation (8).

[0014] The vibration calculation module calculates the blade vibration displacement based on the blade arrival time, key phase arrival time, blade rotation speed, blade tip rotation radius, and the angle between the blade and the key phase, including: The method for calculating blade vibration displacement is as follows: (12) (13) In the formula, Let R be the angle between blade b and the bond phase, and R be the blade tip rotation radius. The blade vibration displacement calculation method under all operating conditions is as follows: (14) The blade vibration displacement calculated by equation (14) includes the blade tip vibration displacement caused by the blade modal response, as well as the blade tip steady-state displacement caused by axial movement, steam pressure, rotor thermal expansion and corrosion deformation. By reading the blade vibration data, the vibration data of each blade from the first sensor to the fifth sensor are rearranged, and the blade vibration waveform of each blade from the first sensor to the fifth sensor is sorted out. The time-domain vibration waveform of each blade is calculated, and the vibration spectrum of each blade is calculated by Fourier transform. The health status of the last stage blade of the turbine is analyzed by analyzing the shape, amplitude and spectrum of the time-domain waveform of each blade.

[0015] The vibration analysis module analyzes and processes the blade vibration displacement, extracting the steady-state and transient displacements of the blade, including: The blade vibration displacement is analyzed and processed to extract the steady-state and transient displacements of the blade. The blade vibration displacement is expressed as: x=x s +x t (15) Where, x s For steady-state displacement, the extraction method involves using an SG filter to filter the blade vibration displacement; the low-frequency displacement component obtained by the filter is the steady-state displacement of the blade. t The transient displacement is obtained by subtracting the steady-state displacement from the blade vibration displacement.

[0016] The state assessment module reconstructs the blade's transient and steady-state displacements into blade stress. By comparing the blade stress with stress thresholds for different states, it performs blade state assessment, including: By performing blade modal analysis, the displacement-stress transfer function of the vibration mode corresponding to the transient displacement and the displacement-stress reconstruction coefficient of the steady-state displacement are obtained. The transient and steady-state displacements obtained from the analysis are then reconstructed into dynamic and steady-state stresses, and the stress of the blade is calculated as follows: (16) in, The displacement-stress reconstruction coefficient is the steady-state displacement. Let σ be the displacement-stress transfer function of the vibration mode, and σ be the stress value of the blade. The blade's condition is divided into healthy, sub-healthy, and faulty states. The stress thresholds for different blade states are determined by dynamic simulation calculations combined with calibration tests. The stress threshold for the healthy state is σ1, and the stress threshold for the sub-healthy state is σ2. By comparing the relationship between σ and σ1 and σ2, the blade's condition can be assessed. Specifically, when σ < σ1, the blade is in a healthy state; when σ1 < σ < σ2, the blade is in a sub-healthy state; and when σ > σ2, the blade is in a faulty state.

[0017] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it performs online monitoring of the last stage blades of a steam turbine low-pressure cylinder based on the tip timing method described above.

[0018] A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements online monitoring of the last stage blades of a steam turbine low-pressure cylinder based on the tip timing method described above.

[0019] A computer program product includes a computer program that, when executed by a processor, implements online monitoring and fault diagnosis analysis of the last stage blades of a steam turbine low-pressure cylinder based on the tip timing method described above.

[0020] The beneficial effects of this invention are as follows: The non-contact vibration measurement device for nuclear power turbine rotating blades of this invention involves setting up a speed synchronization sensor with its monitoring probe facing the rotating blade disk; creating grooves on the end face edge of the rotating blade disk corresponding to the monitoring probes of the speed synchronization sensor; and simultaneously arranging several blade tip timing sensors on the cylinder guide ring at circumferential positions corresponding to the radial direction of the rotating blade disk. This invention achieves the measurement of the moment when the rotating blades in a nuclear power turbine reach the blade tip timing sensor based on the blade tip timing principle, converts the measured time pulse signal into a blade tip amplitude signal, and further extracts the vibration parameters of the nuclear power turbine rotating blades through a blade tip timing signal parameter identification method. This provides a basis for vibration assessment, life evaluation, fault diagnosis, and improved design of the rotating blades within the entire turbine. This invention enables the effective acquisition of speed synchronization signals in the overall environment of a steam turbine. By combining this with the reliable placement, reasonable installation, and rapid disassembly of blade tip timing sensors, it achieves non-contact vibration measurement of all levels of rotating blades within the steam turbine. This ensures that the acquired non-contact vibration measurement signals of the rotating blades have a good signal-to-noise ratio, thus providing a simple, convenient, and effective online monitoring method for the accurate identification of the desired steam turbine vibration modes. Attached Figure Description

[0021] Figure 1 A flowchart of an online monitoring and diagnostic method for the low-pressure last-stage blades of a nuclear power turbine based on tip timing, provided by the present invention; Figure 2 This is a schematic diagram of sensor installation.

[0022] In the diagram: 1. Baffle plate, 2. Moving blade, 3. Blade tip timing sensor, 4. Guide ring. Detailed Implementation

[0023] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] This invention provides a method, device, system, medium, and product for online monitoring and diagnosis of low-pressure last-stage blades in nuclear power turbines based on tip timing. By real-time monitoring of blade vibration amplitude, frequency, tilt angle, and other data, the invention couples and correlates the blade health status, refers to historical operating data and numerical calculation of fault response databases, rationally formulates unit operation procedures, and conducts targeted blade maintenance. This solves common problems of low-pressure last-stage blades in nuclear power units, shortens maintenance time, and improves operation and maintenance efficiency.

[0025] This system constructs a complete blade tip signal acquisition and analysis chain by installing a high-precision, high-frequency response fine-tuning signal sensor on the turbine cylinder head and configuring a blade vibration signal amplifier, data acquisition instrument, data signal processor, and remote human-machine interface host on site. During turbine operation, the signal acquisition system acquires the blade tip arrival signal in real time and binds it with spatiotemporal coordinate information; the signal analysis system extracts key dynamic parameters such as blade amplitude, vibration frequency, and torsional amplitude based on the blade tip timing signal.

[0026] The fault diagnosis system combines the above parameters with historical turbine operating data and finite element simulation data, inputting them into a pre-set defect analysis model. Through two levels of comprehensive indicators, it intelligently diagnoses the blade condition, achieving quantitative identification of defect type, defect level, abnormal area location, and geometric deformation. The system further integrates with a structural mechanics defect response database to perform matching analysis and root cause inference on the diagnostic results, generating target fault cause data. The early warning system outputs a blade health assessment report based on the fault type and level, abnormal area, and deformation. The report includes a defect location map and corresponding graded maintenance strategies.

[0027] This invention achieves intelligent monitoring of the operating status of nuclear power turbine blades by fusing real-time tip timing data with historical / simulated data. This significantly improves the accuracy, reliability, and efficiency of blade fault diagnosis, enabling the understanding of blade health trends and early warning of potential faults. It can effectively ensure the safe and stable operation of nuclear power plant turbine units and avoid accidental shutdowns.

[0028] An online monitoring device for the low-pressure last-stage blades of a nuclear power steam turbine based on the tip timing method was developed in four levels: theoretical level, monitoring level, diagnostic level and system level.

[0029] The theoretical level focuses on breakthroughs in key technologies for blade fault monitoring and diagnosis, researching and revealing the typical fault mechanisms, dynamic characteristics and multi-source information characterization methods of high-speed blades, and combining multi-source information such as vibration amplitude, frequency and tilt angle measurement to identify weak signs of blade faults and long blade life prediction and maintenance technology.

[0030] The monitoring layer focuses on breakthroughs in advanced sensor technology for blade faults under high speed and high humidity conditions, completes the demonstration application and verification of online fault monitoring and diagnosis technology for high-speed blades in nuclear power turbine scenarios, solves the problem of high-precision multi-source status monitoring of blades in extreme environments, and comprehensively optimizes sensor performance, tooling, monitoring systems and construction schemes.

[0031] The diagnostic layer focuses on breakthroughs in the detection technology of complex defects in blades, comprehensively optimizes virtual simulation models, establishes a database of common defects and defect sensitivity analysis, improves the fault factor vector matching algorithm, realizes online monitoring and diagnosis of blade health status, and establishes a model library and expert diagnostic library through multi-objective analysis.

[0032] The system layer focuses on building system functions and system engineering installation schemes based on user experience and feedback optimization, developing blade online intelligent operation and maintenance analysis systems, electronic equipment and storage solutions, and realizing the productization and standardization of the developed system.

[0033] like Figure 2 As shown, an online monitoring and diagnostic device for the low-pressure last stage blades of a nuclear power turbine based on tip timing is provided. A speed synchronization sensor is set up with its monitoring probe facing the rotating blade disk. A groove is made on the end face edge of the rotating blade disk at the position corresponding to the monitoring probe of the speed synchronization sensor. At the same time, several tip timing sensors are arranged on the cylinder guide ring at circumferential positions corresponding to the radial direction of the rotating blade disk.

[0034] The present invention provides an online monitoring and diagnostic method for the low-pressure last-stage blades of a nuclear power turbine based on tip timing, comprising the following steps: Step 1: Measure the arrival time t of two adjacent key phase pulses using a speed synchronization sensor. n and t n+1 Determine the time T for the rotating blade disk to complete one revolution. n =t n+1 -t n And the rotational speed of the nuclear power turbine impeller ω=60 / T n Unit: revolutions per minute (rpm); Step 2: Determine the order in which each blade passes the tip timing sensor based on the relative position of each blade on the rotating blade disk and the rotational speed synchronization sensor, thereby determining the number of each blade. 。 Assumption t n At what time, the blade numbered by the tip timing sensor is b The blade numbers subsequently detected by the tip timing sensor are as follows: b、b+ 1. b +2..., the number of the blades that have already been passed is... b- 1、b- 2. b -3...; Step 3: Based on the rotational speed of the nuclear power turbine impeller and the determined blade numbering... b Based on the key phase pulse of the speed synchronization sensor, the theoretical arrival time of each blade when it is not vibrating and passes through the blade tip timing sensor is determined. And the actual arrival time of each blade through the blade tip timing sensor during vibration. The theoretical arrival time of the blade tip timing sensor when the blade is not vibrating. Calculations were performed, assuming the bond phase pulse time was... t n If the impeller has N blades per revolution, then the time interval between adjacent blades passing the blade tip timing sensor when the blades are not vibrating is (t... n+1 -t n ) / N, b The theoretical arrival time of blade number 1 after passing the tip timing sensor is: t nb , b The theoretical arrival time of blade +1 passing the tip timing sensor is: t nb +(t n+1 -t n ) / N. The actual arrival time of the blade tip timing sensor during blade vibration. This is obtained through data collection; Step 4: Based on the theoretical arrival time The actual arrival time The time required for the rotating blade disk to rotate one revolution The vibration displacement of each blade is determined by combining the distance from the blade tip to the center of the rotating blade disk. According to the theoretical arrival time The actual arrival time The time required for the rotating blade disk to rotate one revolution Based on the distance from the blade tip to the center of the rotating blade disk, the vibration displacement of each blade is calculated as follows: , In the formula, r This refers to the distance from the tip of the blade to the center of the rotating blade disk. This refers to the actual arrival time of the blade; This refers to the theoretical arrival time of the blade. This refers to the time it takes for the impeller to rotate once. This represents the blade vibration displacement; subscript n Indicates the current lap number; subscript b Number the blades; subscripts This is the serial number of the leaf tip timing sensor.

[0035] This invention provides an online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing, comprising: a data acquisition module, a data validity judgment module, a data cleaning module, a speed calculation module, a vibration calculation module, a vibration analysis module, a condition assessment module, and a condition early warning module.

[0036] The data acquisition module is used to acquire the arrival time of the impeller blades and the arrival time of the key phase. The data acquisition module adopts the tip-time measurement principle, installing five blade vibration sensors in the diaphragm above the last-stage turbine blades. The acquisition system uses 86MHz high-speed synchronous sampling to simultaneously acquire vibration data from the five blade vibration sensors, and the acquired blade vibration data is stored in the system database.

[0037] A blade tip timing sensor is installed on the exhaust guide ring of the last stage of the steam turbine to measure the time t when the blade reaches the blade tip timing sensor. b,n A key phase is installed on the turbine rotor, and a key phase timing sensor is placed on top of the key phase to measure the time t when the key phase arrives at the key phase timing sensor. o,n The subscript 'b' represents the blade number, the subscript 'o' represents the bond phase, and the subscript 'n' represents the number of revolutions.

[0038] In some embodiments, the data validity determination module determines the validity of the blade arrival time data, including: The period of one rotation of the rotor is: T=t o,n+1 -t o,n (1), The arrival time window width for each blade during this cycle is: t w =T / n b (2), n b The number of blades; The range of blade arrival times is: t o,n +(b-1)t w ≤t b,n ≤t o,n +bt w (3), In the formula, t o,n Let t be the arrival time of the bond phase in the nth rotational cycle. o,n+1 Let n be the arrival time of the (n+1)th rotating periodic bond phase. b 'b' represents the number of blades on the tested blade disk, and 'b' represents the blade number. The validity of the blade arrival time data is determined by equation (3).

[0039] In some embodiments, the data cleaning module cleans the arrival time data of falsely triggered blades, including: If there are multiple arrival time signals or no blade arrival time signal within the range of values ​​of equation (3), this is considered a false triggering of the blade tip timing signal. The method for cleaning false triggering data is as follows: The frequency conversion of the bond phase is: f o =1 / (t o,n+1 -t o,n (4), The actual passing frequency between the arrival time signals of two adjacent blades is: f o act =1 / (t b+1,n -t b,n (5), If f o act >n b f b If there are redundant false triggering signals in the blade arrival time signal, the cleaning method in this case is to directly remove the redundant arrival time signal corresponding to this passing frequency. If f o act <n b f b If the blade arrival time signal contains a missing false trigger signal, the cleaning method in this case is to supplement the missing arrival time signal according to the key phase frequency. The supplementation method is as follows: t b+1,n =t b,n +1 / (n b f o ), t b,n =t b+1,n -1 / (n b f o ), In the formula, t b,n Let t be the arrival time of blade b in the nth rotation cycle. b+1,n This represents the arrival time of blade b+1 in the (n+1)th rotation cycle.

[0040] In some embodiments, the speed calculation module solves for the speed measurement matrix constructed at the key phase arrival time, and then calculates the speed of the rotor blades. The method is as follows: f=1 / (t o,n+1 -t o,n (7), The rotational speed calculated by equation (7) is the average speed of one revolution of the blade, which is applicable to constant speed conditions but not to variable speed conditions. The present invention proposes the following method for calculating the rotational speed under all operating conditions: The rotational speed of the turbine rotor blades can be expressed as: f(t) = f0 + Σ1 c a c t c (8), In the formula, f0 is the initial rotational speed of the blade when the key phase reaches the key phase timing sensor, Σ1 c a c t c Σ1 represents the change in rotational speed over time. c a c t c =0 indicates constant speed operation, c=1 indicates linear speed change operation, and c>1 indicates nonlinear speed change operation. Since the transient and steady-state displacements of the rotor blades can cause deviations in the blade arrival time, using the blade arrival time data to calculate the rotational speed would introduce calculation errors. Therefore, the rotational speed measurement matrix is ​​constructed based on the key phase arrival time as follows: (9), Equation (9) can be written in matrix form as follows: C=MF (10) In the formula, C is a matrix related to the number of blade rotations, and the value of c is related to the rate of change of rotational speed. In this invention, it is taken as 4. M is a matrix related to the arrival time of the bond phase, and F is the rotational speed matrix, which can be obtained by the least squares method. F=(M T M) -1 M T C (11), After obtaining the rotational speed matrix, the blade rotational speed can be calculated according to equation (8).

[0041] In some embodiments, the vibration calculation module calculates the blade vibration displacement based on the blade arrival time, key phase arrival time, blade rotation speed, blade tip rotation radius, and the angle between the blade and the key phase, including: In tip timing technology, the method for calculating blade vibration displacement is as follows: (12) (13) In the formula, Let R be the angle between blade b and the key phase, and R be the blade tip rotation radius. Since the rotational speed in equation (12) is the average speed of one revolution of the blade, the blade vibration displacement calculated by equation (12) is the vibration displacement under constant speed conditions, and is not applicable to variable speed conditions. The present invention proposes the following method for calculating blade vibration displacement under all operating conditions: (14) Equation (14) can be used to calculate the blade vibration displacement, which includes the blade tip vibration displacement caused by the blade modal response, as well as the blade tip steady-state displacement caused by axial movement, steam pressure, rotor thermal expansion and corrosion deformation.

[0042] The turbine last-stage blade vibration waveform calculation module reads blade vibration data from the system database, rearranges the vibration data of each blade from the first sensor to the fifth sensor, organizes the blade vibration waveform of each blade from the first sensor to the fifth sensor, calculates the time-domain vibration waveform of each blade, and calculates the vibration spectrum of each blade through Fourier transform. By analyzing the shape, amplitude and spectrum of the time-domain waveform of each blade, the health status of the turbine last-stage blade is analyzed.

[0043] In some embodiments, the vibration analysis module analyzes and processes the blade vibration displacement to extract the steady-state and transient displacements of the blade, including: The blade vibration displacement is analyzed and processed to extract the steady-state and transient displacements. The blade vibration displacement can be expressed as: x=x s +x t (15) Where, x s For steady-state displacement, the extraction method involves using an SG filter to filter the blade vibration displacement; the low-frequency displacement component obtained by the filter is the steady-state displacement of the blade. t The transient displacement can be obtained by subtracting the steady-state displacement from the blade vibration displacement.

[0044] In some embodiments, the state assessment module reconstructs the transient and steady-state displacements of the blade into blade stress, and performs blade state assessment by comparing the blade stress with stress thresholds for different states, including: By performing blade modal analysis, the displacement-stress transfer function of the vibration mode corresponding to the transient displacement and the displacement-stress reconstruction coefficient of the steady-state displacement are obtained. The transient and steady-state displacements obtained from the analysis are then reconstructed into dynamic and steady-state stresses, and the stress of the blade is calculated as follows: (16) in, The displacement-stress reconstruction coefficient is the steady-state displacement. Let σ be the displacement-stress transfer function of the vibration mode, and σ be the stress value of the blade. Then, the blade's condition is divided into healthy, sub-healthy, and faulty states. The stress thresholds for different blade states are determined by dynamic simulation calculations combined with calibration tests. The stress threshold for the healthy state is σ1, and the stress threshold for the sub-healthy state is σ2. By comparing the relationship between σ and σ1 and σ2, the blade's condition is assessed. Specifically, when σ < σ1, the blade is in a healthy state; when σ1 < σ < σ2, the blade is in a sub-healthy state; and when σ > σ2, the blade is in a faulty state.

[0045] The status warning module is used to send feedback information based on the blade status assessment results. When the blade is in a healthy state, the status warning module does not send warning information. When the blade is in a sub-healthy state, the status warning module issues a warning. When the blade is in a faulty state, the status warning module issues an alarm.

[0046] As a further improvement of the present invention, the method further includes; Based on the blade tip amplitude curve, the vibration parameters of each blade are obtained through single-degree-of-freedom fitting and circumferential Fourier transform, including: vibration frequency, vibration amplitude, resonant rotation speed, phase, excitation order, modal order, and modal damping.

[0047] Based on the time it takes for the blade to reach the sensor and the angle between the sensors, the vibration velocity of the blade tip in the nth revolution can be calculated using the angle between the kth pair of sensors and their acquired arrival times. ,in, R is the turbine radius. Let the angle be the smaller angle of the k-th pair of sensors. The angle of the k-th sensor pair with the larger angle; turbine speed. Constant.

[0048] Record the sequence of blade vibration velocity points V N×1 and the corresponding time sequence T N×1 N is the number of sampling points; According to the Nyquist sampling theorem, determine the sequence length N of the reconstructed signal. res ,satisfy f max f is the highest frequency of the vibration mode. Ω Rotational speed; Record the positions of the undersampled signal data points in the reconstructed signal sequence, and record the position matrix q. N×1 Then add an N res ×N res The q corresponding to the identity matrixN×1 Extract as observation matrix ; Establish N res Discrete dictionary matrix of dimension 1 Then establish the coefficient transformation matrix. ; Establish a coefficient model for the undersampled signal. The solution obtained by interior point method is λ is the regularization parameter; ; right The vibration modal frequencies are obtained by performing a Fourier transform. Based on the frequencies of different modal vibrations and the arrival times of different sensors, a sensor design matrix for the nth cycle is constructed. , Where f m Let t be the vibration frequency of the m-th mode, where m is the mode number, and t is the frequency of the m-th mode. k,n Let be the average blade arrival time collected by the k-th pair of sensors on the nth lap, where , Let be the time acquired during the nth cycle for the sensor with the smaller angle in the k-th pair of sensors. The time is the time collected in the nth cycle by the sensor with the larger angle in the k-th pair of sensors. Based on the obtained blade tip vibration velocity, the blade tip timing vibration parameters for different frequency bands are obtained by solving the least squares method; the vibration parameters are obtained by the following formula: ,in ,in The vibration velocity value obtained by the k-th pair of sensors in the nth cycle; , where a is the vibration parameter matrix; Based on the conversion relationship between the measured blade tip vibration displacement and velocity, the blade vibration displacement parameters are finally obtained; Vibration velocity amplitude under different blade vibration modes Vibration velocity phase ; Vibration displacement amplitude under different blade vibration modes Vibration velocity phase As a further improvement of the present invention, the method of the state assessment module further includes: The vibration stress distribution and the relationship between tip amplitude and vibration stress of each blade under the excitation order of interest are calculated to obtain the vibration stress at the maximum stress point of each blade under each excitation order.

[0049] The calculation method for the vibration stress distribution and the relationship between blade tip amplitude and vibration stress, in some embodiments, includes the following steps: Step S101: Collect information on the nuclear power turbine blade impeller based on the preset physical entity model to obtain the corresponding entity information set; Step S102: Construct a finite element model using the entity information set to obtain the corresponding target virtual entity model; Step S103: Perform finite element numerical simulation calculations using the target virtual entity model to obtain the corresponding set of simulated working condition data; In some specific embodiments, the process of performing step S101 may specifically include the following steps: (1) Extract the identifiers from the target blade and impeller to obtain the identifier information corresponding to the target blade and impeller; (2) Analyze the storage location of the pre-set physical entity model using the identification information to generate the corresponding information storage address; (3) Collect information based on the information storage address to obtain the corresponding entity information set.

[0050] Specifically, the server extracts identifiers from the target blades and impellers. The server pre-obtains identifier extraction functions corresponding to the target blades and impellers, and uses these functions to obtain identifier information. Further, the server matches the identifier information to the information storage address and determines the corresponding address description protocol based on preset address resolution rules, generating the corresponding information storage address. Finally, the server collects information based on the information storage address to obtain the corresponding entity information set.

[0051] In some specific embodiments, the process of performing step S102 may specifically include the following steps: (1) Extract features from the entity information set to obtain the corresponding entity feature information; (2) Perform parameter matching based on entity feature information to determine the corresponding entity feature parameters; (3) Based on the entity feature parameters and entity feature information, a finite element model is constructed to obtain the corresponding target virtual entity model.

[0052] In this invention, the server takes the physical entity of the target blade impeller as the object, collects data on its geometric dimensions and assembly mode, and obtains the corresponding entity feature information. Based on the construction of entity feature information, the blade rotor structure is simplified into the boundary conditions of the target blade impeller to obtain entity feature parameters. In this scheme, the server takes the blade impeller as the core research object and constructs it as a cyclic symmetric model in the finite element model to obtain the corresponding target virtual entity model.

[0053] In some specific embodiments, after performing step S102 and before performing step S103, the method further includes: S201: Perform simulation calculations based on a single variable parameter on the target virtual entity model to obtain multiple corresponding performance evaluation parameters; S202: Perform fluctuation range analysis on multiple performance evaluation parameters to obtain the fluctuation range corresponding to each performance evaluation parameter; S203: Perform weight analysis on multiple performance evaluation parameters to obtain the parameter weight corresponding to each performance evaluation parameter.

[0054] It should be noted that when the server performs simulation calculations based on a single variable parameter to obtain multiple corresponding performance evaluation parameters for the target virtual entity model, since the defect type, crack location, crack size, and crack depth are different in actual working conditions, in the solution of this invention, the server pre-sets the calculation working point for each parameter, and then calculates the stress, strain, and vibration frequency based on the fluctuation range, and outputs a set of data such as the maximum equivalent stress, maximum equivalent strain, and modal vibration frequency.

[0055] In one specific embodiment, the process of executing step S103 may specifically include the following steps: S301: Perform unit information matching on the target virtual entity model to determine the corresponding unit information set, wherein the unit information set includes unit shape and unit type; S302: Based on the unit set information, the target virtual entity model is matched with the working environment through a preset mesh division algorithm to obtain the corresponding working environment set; S303: By combining the working environment with the target virtual entity model, a working condition model simulation calculation is performed to obtain the corresponding simulated working condition data set.

[0056] Specifically, after assigning parameter weights to the target blade impeller, this step determines the parameters that have a significant impact on blade performance and performs multi-parameter analysis. Before the analysis, since nuclear power turbines typically do not involve peak-shaving operations, their operating parameters such as temperature, pressure, and speed are very stable except during startup and shutdown. Therefore, the influence of operating temperature, pressure, and speed on the turbine's condition is not considered in the multi-parameter analysis. Based on the application purpose of the nuclear power turbine blade condition monitoring system provided in this invention patent, the goal is to monitor and determine the operational health status of the nuclear power turbine. Potential defects in the blade body, root, and bosses are taken as the research objects, including but not limited to cracks, corrosion, and fractures.

[0057] First, the server performs element information matching on the target virtual entity model to determine the corresponding element information set. This set includes element shape and type. To improve the analysis accuracy of the finite element model, the server controls the specific mesh density, element shape, element type, and mesh generation technique for different elements in the finite element model to determine the corresponding element information set. Simultaneously, due to the blade's motion characteristics, in contact analysis, the server determines the rotating components and reference points corresponding to the target virtual entity model and defines and sets the contact attributes, contact relationships, and contact surfaces. Then, based on the element set information, the server performs working environment matching on the target virtual entity model using a preset mesh generation algorithm to obtain the corresponding working environment set.

[0058] It should be noted that, based on single-variable analysis, by setting defect parameters and defect types in the finite element model, experimental points are designed to simulate various unhealthy operating states that the blade may encounter during actual operation, performing operating condition simulation calculations to obtain the corresponding simulated operating condition data set. In this invention, the server uses Python to automatically output the calculation results of blade-impeller amplitude, frequency, and stress, thereby obtaining a sufficient set of simulated operating condition data.

[0059] The server performs simulation calculations based on parameter settings to obtain multiple sets of operating condition data, which are then used as a set of simulated operating condition data. In this solution, the calculation results given by the finite element model are clear, intuitive, and highly reliable, and can obtain data information that cannot be obtained by general monitoring methods. This comprehensively shows the operating status of the blade under different operating conditions, ensuring the accuracy of fault diagnosis when diagnosing blade performance in the future.

[0060] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it realizes online monitoring of the last stage blades of the low-pressure cylinder of a steam turbine based on the tip timing method described above.

[0061] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements online monitoring of the last stage blades of the low-pressure cylinder of a steam turbine based on the tip timing method described above.

[0062] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements online monitoring and fault diagnosis analysis of the last stage blades of the low-pressure cylinder of a steam turbine based on the blade tip timing method described above.

Claims

1. An online monitoring and diagnostic device for the low-pressure last-stage blades of a nuclear power turbine based on tip timing, characterized in that: A speed synchronization sensor is installed, with its monitoring probe facing the rotating blade disk. A groove is made on the end face of the rotating blade disk at a position corresponding to the monitoring probe of the speed synchronization sensor. At the same time, several blade tip timing sensors are arranged on the cylinder guide ring at circumferential positions corresponding to the radial direction of the rotating blade disk.

2. A method for online monitoring and diagnosis of low-pressure last-stage blades in nuclear power steam turbines based on tip timing, characterized in that, Includes the following steps: Step 1: Measure the arrival time t of two adjacent key phase pulses using a speed synchronization sensor. n and t n+1 Determine the time T for the rotating blade disk to complete one revolution. n =t n+1 -t n And the rotational speed of the nuclear power turbine impeller ω=60 / T n ; Step 2: Determine the order in which each blade passes the tip timing sensor based on the relative position of each blade on the rotating blade disk and the rotational speed synchronization sensor, and determine the number of each blade. ， Assumption t n At what time, the blade numbered by the tip timing sensor is b The blade numbers subsequently detected by the tip timing sensor are as follows: b、b+ 1. b +2..., the number of the blades that have already been passed is... b- 1 、b- 2. b -3...; Step 3: Based on the rotational speed of the nuclear power turbine impeller and the determined blade numbering... b Using the key phase pulse of the speed synchronization sensor as a reference, the theoretical arrival time of each blade when it is not vibrating and passes through the blade tip timing sensor is determined. And the actual arrival time of each blade through the blade tip timing sensor during vibration. The theoretical arrival time of the blade tip timing sensor when the blade is not vibrating. Calculations were performed, assuming the bond phase pulse time was... t n If the impeller has N blades per revolution, then the time interval between adjacent blades passing the blade tip timing sensor when the blades are not vibrating is (t... n+1 -t n ) / N, b The theoretical arrival time of blade number 1 after passing the tip timing sensor is: t nb , b The theoretical arrival time of blade +1 passing the tip timing sensor is: t nb +(t n+1 -t n ) / N, the actual arrival time of the blade tip timing sensor during blade vibration. This is obtained through data collection; Step 4: Based on the theoretical arrival time The actual arrival time The time required for the rotating blade disk to rotate one revolution The vibration displacement of each blade is determined by combining the distance from the blade tip to the center of the rotating blade disk.

3. The method for online monitoring and diagnosis of low-pressure last-stage blades of nuclear power steam turbines based on tip timing as described in claim 2, characterized in that: In step 4, based on the theoretical arrival time... The actual arrival time The time required for the rotating blade disk to rotate one revolution Based on the distance from the blade tip to the center of the rotating blade disk, the vibration displacement of each blade is calculated as follows: ， In the formula, r This refers to the distance from the tip of the blade to the center of the rotating blade disk. This refers to the actual arrival time of the blade; This refers to the theoretical arrival time of the blade. This refers to the time it takes for the impeller to rotate once. This represents the blade vibration displacement; subscript n Indicates the current lap number; subscript b Number the blades; subscript s This is the serial number of the leaf tip timing sensor.

4. An online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing, characterized in that: It includes a data acquisition module, a data validity judgment module, a data cleaning module, a rotational speed calculation module, a vibration calculation module, a vibration analysis module, a condition assessment module, and a condition early warning module; The data acquisition module is used to acquire the arrival time of the impeller blades and the arrival time of the key phase; The data validity assessment module determines the validity of the blade arrival time data. The data cleaning module cleans the arrival time data of the blades that were mistakenly triggered. The rotational speed calculation module solves the rotational speed measurement matrix constructed by the arrival time of the key phase, and then calculates the rotational speed of the rotor blades; The vibration calculation module calculates the blade vibration displacement based on the blade arrival time, the key phase arrival time, the blade rotation speed, the blade tip rotation radius, and the angle between the blade and the key phase. The vibration analysis module analyzes and processes the blade vibration displacement, extracting the steady-state and transient displacements of the blade. The state assessment module reconstructs the transient and steady-state displacements of the blade into the blade stress, and performs blade state assessment by comparing the blade stress with stress thresholds for different states. The status warning module is used to send feedback information based on the blade status assessment results. When the blade is in a healthy state, the status warning module does not send warning information. When the blade is in a sub-healthy state, the status warning module issues a warning. When the blade is in a faulty state, the status warning module issues an alarm.

5. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The data acquisition module installs five blade vibration sensors in the diaphragm above the last-stage blades of the turbine, simultaneously collecting vibration data from the five sensors. A blade tip timing sensor is arranged on the exhaust guide ring of the last-stage turbine to measure the time t when the blade reaches the blade tip timing sensor. b,n A key phase is installed on the turbine rotor, and a key phase timing sensor is placed on top of the key phase to measure the time t when the key phase arrives at the key phase timing sensor. o,n The subscript b represents the blade number, the subscript o represents the bond phase, and the subscript n represents the number of revolutions.

6. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The data validity assessment module's assessment process includes: The period of one rotation of the rotor is: T=t o,n+1 -t o,n (1), The arrival time window width for each blade during this cycle is: t w =T / n b (2), n b The number of blades; The range of blade arrival times is: t o,n +(b-1)t w ≤t b,n ≤t o,n +bt w (3), In the formula, t o,n Let t be the arrival time of the bond phase in the nth rotational cycle. o,n+1 Let n be the arrival time of the (n+1)th rotating periodic bond phase. b b is the number of blades on the tested blade disk and b is the blade number; the validity of the blade arrival time data is determined by equation (3).

7. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 6, characterized in that, The data cleaning module cleans the arrival time data of the blades that were mistakenly triggered, including: If there are multiple arrival time signals or no blade arrival time signal within the range of values ​​of equation (3), this is considered a false triggering of the blade tip timing signal. The method for cleaning false triggering data is as follows: The frequency conversion of the bond phase is: f o =1 / (t o,n+1 -t o,n ) (4), The actual passing frequency between the arrival time signals of two adjacent blades is: f o act =1 / (t b+1,n -t b,n ) (5), If f o act >n b f b If there are redundant false triggering signals in the blade arrival time signal, the cleaning method in this case is to directly remove the redundant arrival time signal corresponding to this passing frequency. If f o act <n b f b If the blade arrival time signal contains a missing false trigger signal, the cleaning method in this case is to supplement the missing arrival time signal according to the key phase frequency. The supplementation method is as follows: t b+1,n =t b,n +1 / (n b f o ), t b,n =t b+1,n -1 / (n b f o ), In the formula, t b,n Let t be the arrival time of blade b in the nth rotation cycle. b+1,n This represents the arrival time of blade b+1 in the (n+1)th rotation cycle.

8. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The speed calculation module solves the speed measurement matrix constructed based on the key phase arrival time, and then calculates the speed of the rotor blades. The method is as follows: f=1 / (t o,n+1 -t o,n ) (7), The rotational speed calculated by equation (7) is the average speed of one revolution of the blade. The calculation method for the rotational speed under all operating conditions is as follows: The rotational speed of the turbine rotor blades is expressed as: f(t)=f0+Σ1 c a c t c (8), In the formula, f0 is the initial rotational speed of the blade when the key phase reaches the key phase timing sensor, Σ1 c a c t c Σ1 represents the change in rotational speed over time. c a c t c =0 indicates constant speed operation, c=1 indicates linear speed change operation, and c>1 indicates nonlinear speed change operation. The rotational speed measurement matrix is ​​constructed based on the arrival time of the key phase as follows: （9）， Equation (9) can be written in matrix form as follows: C=MF (10) In the formula, C is a matrix related to the number of blade rotations, and the value of c is related to the rate of change of rotational speed. In this invention, it is taken as 4. M is a matrix related to the arrival time of the bond phase, and F is the rotational speed matrix, which is obtained by the least squares method. F=(M T M) -1 M T C (11), After obtaining the rotational speed matrix, the blade rotational speed is calculated according to equation (8).

9. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The vibration calculation module calculates the blade vibration displacement based on the blade arrival time, key phase arrival time, blade rotation speed, blade tip rotation radius, and the angle between the blade and the key phase, including: The method for calculating blade vibration displacement is as follows: （12）， （13）， In the formula, Let R be the angle between blade b and the bond phase, and R be the blade tip rotation radius. The blade vibration displacement calculation method under all operating conditions is as follows: （14）， The blade vibration displacement calculated by equation (14) includes the blade tip vibration displacement caused by the blade modal response, as well as the blade tip steady-state displacement caused by axial movement, steam pressure, rotor thermal expansion and corrosion deformation. By reading the blade vibration data, the vibration data of each blade from the first sensor to the fifth sensor are rearranged, and the blade vibration waveform of each blade from the first sensor to the fifth sensor is sorted out. The time-domain vibration waveform of each blade is calculated, and the vibration spectrum of each blade is calculated by Fourier transform. The health status of the last stage blade of the turbine is analyzed by analyzing the shape, amplitude and spectrum of the time-domain waveform of each blade.

10. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The vibration analysis module analyzes and processes the blade vibration displacement, extracting the steady-state and transient displacements of the blade, including: The blade vibration displacement is analyzed and processed to extract the steady-state and transient displacements of the blade. The blade vibration displacement is expressed as: x=x s +x t (15), Where, x s For steady-state displacement, the extraction method involves using an SG filter to filter the blade vibration displacement; the low-frequency displacement component obtained by the filter is the steady-state displacement of the blade. t The transient displacement is obtained by subtracting the steady-state displacement from the blade vibration displacement.

11. The online monitoring and diagnostic system for the low-pressure last-stage blades of a nuclear power turbine based on tip timing as described in claim 4, characterized in that, The state assessment module reconstructs the blade's transient and steady-state displacements into blade stress. By comparing the blade stress with stress thresholds for different states, it performs blade state assessment, including: By performing blade modal analysis, the displacement-stress transfer function of the vibration mode corresponding to the transient displacement and the displacement-stress reconstruction coefficient of the steady-state displacement are obtained. The transient and steady-state displacements obtained from the analysis are then reconstructed into dynamic and steady-state stresses, and the stress of the blade is calculated as follows: （16）， in, The displacement-stress reconstruction coefficient is the steady-state displacement. Let σ be the displacement-stress transfer function of the vibration mode, and σ be the stress value of the blade. The blade's condition is divided into healthy, sub-healthy, and faulty states. The stress thresholds for different blade states are determined by dynamic simulation calculations combined with calibration tests. The stress threshold for the healthy state is σ1, and the stress threshold for the sub-healthy state is σ2. By comparing the relationship between σ and σ1 and σ2, the blade's condition can be assessed. Specifically, when σ < σ1, the blade is in a healthy state; when σ1 < σ < σ2, the blade is in a sub-healthy state; and when σ > σ2, the blade is in a faulty state.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to perform online monitoring of the last stage blades of a steam turbine low-pressure cylinder based on any of the above-described tip timing methods.

13. A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements online monitoring of the last stage blades of a steam turbine low-pressure cylinder based on the tip timing method described above.

14. A computer program product, comprising a computer program that, when executed by a processor, implements online monitoring and fault diagnosis analysis of the last stage blade of a steam turbine low-pressure cylinder based on the blade tip timing method described above.

Images