Heavy-duty gas turbine blade structure safety online monitoring method, device, equipment and storage medium

Abstract

The present application relates to the technical field of gas turbine health monitoring, in particular to a heavy-duty gas turbine blade structure safety online monitoring method and device. The present application acquires local dynamic strain signals and blade tip vibration displacement signals by setting strain gauge modules and blade tip timing sensor modules respectively, extracts single evidence representing blade state by evidence extraction module, and finally adopts D-S evidence theory for fusion decision by fusion decision module; firstly, the high-precision local measurement capability of strain gauge monitoring and the whole-blade non-contact measurement advantage of blade tip timing monitoring are fused, overcoming the inherent defects of single sensor in monitoring range and quantitative accuracy; secondly, through multi-evidence fusion decision, the subjectivity of manual weight setting is avoided, and the credibility and reliability of blade safety state discrimination are improved; thirdly, the online monitoring of the whole blade structure safety state is realized, providing more accurate technical support for early fault warning and health management of gas turbines.

Description

Technical Field

[0001] This invention relates to the field of gas turbine health monitoring technology, and in particular to a method, device, equipment, and computer storage medium for online monitoring of the structural safety of heavy-duty gas turbine blades. Background Technology

[0002] Gas turbine blades operate in harsh environments with high temperature, high pressure, and high speed for extended periods, and are subjected to complex combined loads. Vibration fatigue fracture is a major cause of unplanned shutdowns of the unit.

[0003] Currently, online monitoring for blade structural safety mainly employs two techniques: strain gauge monitoring and blade tip timing (BTT) monitoring. However, both methods have inherent limitations: strain gauge monitoring is a contact measurement method, only acquiring strain information from a very small local area of ​​the blade, failing to reflect the stress distribution across the entire blade, and requiring slotted leads on the blade, which significantly reduces the blade's service life; while blade tip timing monitoring enables non-contact monitoring of the entire blade, its signals suffer from undersampling issues and insufficient spatial identification accuracy. It can only indirectly infer blade stress and damage status through tip displacement vibration, with limited reliability in monitoring critical areas such as the blade root and middle, and is insensitive to early, minor damage.

[0004] Therefore, existing technologies are insufficient to achieve comprehensive, high-precision, and highly reliable online monitoring of the safety status of the blade structure. Summary of the Invention

[0005] Therefore, the technical problem to be solved by the present invention is to overcome the shortcomings of the existing single sensor monitoring method, which cannot take into account both the monitoring range and quantitative accuracy, and to realize online monitoring of the safety status of the blade structure with full range, high precision and high reliability.

[0006] To address the aforementioned technical problems, this invention provides a method for online monitoring of the structural safety of heavy-duty gas turbine blades, comprising:

[0007] Strain gauges are arranged on at least one blade to obtain local dynamic strain signals at critical parts of the blade;

[0008] A blade tip timing sensor is arranged circumferentially in the stage where the blade is located to obtain the blade tip timing pulse signal when all blades rotate past, and the blade tip vibration displacement signal is obtained by inversion based on the blade tip timing pulse signal.

[0009] Based on the local dynamic strain signal and the blade tip vibration displacement signal, multiple single pieces of evidence characterizing the blade structural state are extracted respectively. The single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal.

[0010] The DS evidence theory is used to fuse the multiple individual pieces of evidence for decision-making, and the safety status result of the blade structure is output.

[0011] Preferably, arranging strain gauges on at least one blade includes:

[0012] Select a portion of the blades along the circumferential direction and attach strain gauges to at least two key locations in the root maximum stress zone, the middle section of the inlet edge, and the middle of the blade body for each selected blade.

[0013] Preferably, the step of inverting the blade tip vibration displacement signal based on the blade tip timing pulse signal includes:

[0014] The ideal arrival pulse time for each blade under vibration-free conditions is calculated based on the rotor speed and the number of blades.

[0015] The time difference is obtained by comparing the actual measured pulse arrival time with the ideal pulse arrival time.

[0016] By combining the blade's rotation radius with the current rotation speed information, the time difference is converted into a blade tip vibration displacement signal.

[0017] Preferably, the first type of evidence includes a dynamic stress ratio and a localization index. The dynamic stress ratio is determined based on the ratio of the amplitude of the local dynamic strain signal at a preset modal frequency to the allowable dynamic stress at the corresponding location. The localization index is determined based on the ratio of the maximum value to the average value of the strain amplitude at multiple measuring points in the local dynamic strain signal.

[0018] Preferably, the second type of evidence includes at least two of the following: tip amplitude anomaly coefficient, natural frequency drift rate, damping anomaly coefficient, and multi-blade consistency coefficient.

[0019] The blade tip amplitude anomaly coefficient is determined based on the ratio of the measured blade tip vibration amplitude to the reference amplitude of a healthy blade; the natural frequency drift rate is determined based on the ratio of the difference between the measured blade frequency and the reference frequency of a healthy blade to the reference frequency; the damping anomaly coefficient is determined based on the ratio of the difference between the measured damping ratio and the reference damping ratio of a healthy blade to the reference damping ratio; and the multi-blade consistency coefficient is determined based on the ratio of the vibration amplitude of a single blade to the average vibration amplitude of all blades in the same class.

[0020] Preferably, the step of using DS evidence theory to fuse the multiple individual pieces of evidence for decision-making and outputting the safety status result of the blade structure includes:

[0021] Each piece of evidence is pre-assigned a confidence level corresponding to a different risk level;

[0022] The confidence levels of all individual pieces of evidence are merged to obtain the overall confidence level of the blade belonging to each risk level;

[0023] The result of determining the safety status of the blade structure is the risk level corresponding to the highest comprehensive confidence level.

[0024] Preferably, the risk level includes at least three levels: safety, early warning, and alarm. The overall confidence level includes the confidence level of the blade status for each risk level and a confidence level representing an uncertain state.

[0025] Preferably, when using the DS evidence theory to make a fusion decision on the multiple single pieces of evidence, a pre-constructed correction coefficient is used to correct the fusion process:

[0026] Multiple types of first-type evidence extracted from local dynamic strain signals are fused at the first level to obtain the first fused evidence;

[0027] Multiple pieces of second-type evidence extracted from the blade tip vibration displacement signal are fused at the first level to obtain the second fused evidence.

[0028] Based on the pre-constructed correction coefficients, the first fused evidence and the second fused evidence are fused at a second level to obtain the final fusion decision result. The correction coefficients are used as weighting factors to adjust the conflict coefficients between different evidence sources in the DS evidence theory fusion formula.

[0029] Preferably, the process of constructing the correction coefficient includes:

[0030] Based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key parts, correction coefficients are constructed to correct the fusion process.

[0031] The transfer ratio parameter is determined based on the ratio of the blade tip vibration displacement signal amplitude to the dynamic strain signal amplitude of the key part, and the dynamic strain signal amplitude of the key part is obtained through finite element analysis or experimental calibration.

[0032] Preferably, the step of constructing correction coefficients for correcting the fusion process based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key part includes:

[0033] Based on the transfer ratio parameter and strain threshold, the probability of blade failure is calculated, and a correction coefficient is obtained.

[0034] This invention also provides an online monitoring device for the structural safety of heavy-duty gas turbine blades, comprising:

[0035] A strain gauge module is installed on at least one blade to acquire local dynamic strain signals at key parts of the blade.

[0036] The blade tip timing sensor module is located in the circumferential direction of the blade level and is used to acquire the blade tip timing pulse signal when all blades rotate past, and to invert the blade tip vibration displacement signal based on the blade tip timing pulse signal.

[0037] An evidence extraction module, connected to the strain gauge module and the blade tip timing sensing module, is used to extract multiple single pieces of evidence characterizing the blade structure state based on the local dynamic strain signal and the blade tip vibration displacement signal, wherein the single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal.

[0038] The fusion decision module, connected to the evidence extraction module, is used to perform fusion decision-making on the multiple single pieces of evidence using DS evidence theory and output the safety status result of the blade structure.

[0039] This invention also provides an online monitoring device for the structural safety of heavy-duty gas turbine blades, comprising:

[0040] Memory, used to store computer programs;

[0041] A processor is used to execute the computer program to implement the steps of the above-described method for online safety monitoring of heavy-duty gas turbine blade structures.

[0042] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for online safety monitoring of heavy-duty gas turbine blade structures.

[0043] The technical solution of the present invention has the following advantages compared with the prior art:

[0044] The online monitoring method for the structural safety of heavy-duty gas turbine blades described in this invention acquires local dynamic strain signals and blade tip vibration displacement signals by setting up a strain gauge module and a blade tip timing sensing module, respectively. An evidence extraction module extracts multiple types of single evidence characterizing the blade's state, and finally, a fusion decision module uses DS evidence theory for fusion decision-making. This achieves the following beneficial effects: First, it integrates the high-precision local measurement capability of strain gauge monitoring with the non-contact measurement advantage of blade tip timing monitoring, overcoming the inherent defects of single sensors in monitoring range and quantitative accuracy. Second, through multi-evidence fusion decision-making, it avoids the subjectivity of manually setting weights, improving the credibility and reliability of blade safety status judgment. Third, it realizes online monitoring of the structural safety status of the entire blade, providing more accurate technical support for early fault warning and health management of gas turbines. Attached Figure Description

[0045] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein:

[0046] Figure 1 This is a flowchart illustrating the implementation of an online safety monitoring method for heavy-duty gas turbine blade structures provided by the present invention.

[0047] Figure 2 This is a structural block diagram of an online safety monitoring device for heavy-duty gas turbine blades provided in an embodiment of the present invention. Detailed Implementation

[0048] The core of this invention is to provide a method, device, equipment, and computer storage medium for online monitoring of the structural safety of heavy-duty gas turbine blades, which effectively solves the technical problem that existing single monitoring technologies cannot simultaneously address both monitoring range and quantitative accuracy.

[0049] To enable those skilled in the art to better understand the present invention, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Please refer to Figure 1. Figure 1 The flowchart illustrates the implementation of an online safety monitoring method for heavy-duty gas turbine blade structures provided by this invention; the specific operation steps are as follows:

[0051] S101: Strain gauges are arranged on at least one blade to obtain local dynamic strain signals at critical parts of the blade;

[0052] S102: A blade tip timing sensor is arranged circumferentially in the stage where the blade is located to obtain the blade tip timing pulse signal when all blades rotate past, and the blade tip vibration displacement signal is obtained by inverting the blade tip timing pulse signal.

[0053] S103: Based on the local dynamic strain signal and the blade tip vibration displacement signal, extract multiple single pieces of evidence characterizing the blade structural state, wherein the single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal.

[0054] S104: The DS evidence theory is used to fuse the multiple individual pieces of evidence for decision-making, and the safety status result of the blade structure is output.

[0055] Based on the above embodiments, this embodiment will provide a detailed description of step S101:

[0056] In some embodiments, arranging strain gauges on at least one blade includes:

[0057] Select a portion of the blades along the circumferential direction and attach strain gauges to at least two key locations in the root maximum stress zone, the middle section of the inlet edge, and the middle of the blade body for each selected blade.

[0058] Furthermore, the strain gauge bonding location is ground and surface-treated, and high-temperature adhesive is used for bonding. High-temperature shielded wires are used to lead the signal from the strain gauge and transmit it to the acquisition device through slip rings or telemetry to obtain the strain signal of that area.

[0059] In other embodiments, the method further includes: selecting multiple blades in the entire blade ring to arrange strain gauges, with three strain gauges attached to each blade, located at a. the maximum stress zone at the blade root, b. the middle section of the inlet edge, and c. the middle of the blade body, respectively, to obtain the state of the high stress zone and sensitive zone during blade operation.

[0060] In one specific embodiment: strain gauges are respectively attached to the root, the middle section of the air inlet edge, and the middle part of the blade of the i-th blade, and the three strain signals are recorded as the first strain signal. x i,a (t), second strain signal x i,b (t), third strain signal x i,c (t).

[0061] Specifically: by grinding and surface treatment of the strain gauge bonding location, applying high-temperature adhesive for bonding, and using high-temperature shielded wire to lead out the signal, the stable transmission of the strain signal is ensured under high temperature and high speed environment.

[0062] It should be noted that this step, by selecting a portion of the blades to place strain gauges in the circumferential direction, not only achieves high-precision stress measurement of the critical areas of the blades, but also avoids the high cost and blade life loss problems caused by placing strain gauges on all blades, thus providing high-precision local strain data for subsequent multi-evidence fusion.

[0063] Based on the above embodiments, this embodiment will provide a detailed description of step S102:

[0064] In some embodiments, obtaining the blade tip vibration displacement signal from the blade tip timing pulse signal includes:

[0065] The ideal arrival pulse time for each blade under vibration-free conditions is calculated based on the rotor speed and the number of blades.

[0066] The time difference is obtained by comparing the actual measured pulse arrival time with the ideal pulse arrival time.

[0067] By combining the blade's rotation radius with the current rotation speed information, the time difference is converted into a blade tip vibration displacement signal.

[0068] In other embodiments, the method further includes: arranging BTT sensor units circumferentially on each stage of the blade, adjusting the sensor axis with the rotor axis as the center using a positioning reference block to ensure uniform installation, and obtaining the pulse signal of the blade tip sweeping across the sensor.

[0069] In one specific embodiment: assuming the blades are in an ideal, vibration-free state, theoretically, the arrival time of each blade at the sensor is uniform and fixed. The ideal arrival pulse time for each blade can be calculated by combining the current rotor speed and the number of blades. However, actual blade rotation is often accompanied by vibration, causing the blades to oscillate, twist, or undergo a combination of oscillation and torsional deformation. This results in varying degrees of lead or lag when the blade tip arrives at the sensor. The time difference is calculated based on the ideal and actual arrival pulse times, and then the blade tip vibration displacement signal of the i-th blade is obtained. y i (t).

[0070] It should be noted that this step utilizes BTT technology to achieve non-contact online vibration monitoring of the entire blade, overcoming the limitation of strain gauges that can only monitor local areas, and providing vibration information across the entire domain for subsequent multi-evidence fusion.

[0071] Based on the above embodiments, this embodiment will provide a detailed description of step S103:

[0072] In some embodiments, the first type of evidence includes a dynamic stress ratio and a localization index. The dynamic stress ratio is determined based on the ratio of the amplitude of the local dynamic strain signal at a preset modal frequency to the allowable dynamic stress at the corresponding location. The localization index is determined based on the ratio of the maximum value to the average value of the strain amplitude at multiple measuring points in the local dynamic strain signal.

[0073] Specifically, for x i,a (t), x i,b (t), x i,c The (t) signal undergoes a Fourier transform to obtain the frequency domain characteristics of the dynamic strain signal: amplitude and frequency. Let f be the modal frequency of the strain gauges in the three regions. n The amplitude below A i,a 、A i,b 、A i,c The allowable dynamic stress at the corresponding location is A i,al 、A i,bl、A i,cl Then the dynamic stress ratio η is defined as: ;

[0074] Define the localization exponent λ as: ;

[0075] It should be noted that the vibration of a healthy blade is a global mode. When a crack, damage, or loosening suddenly occurs at a certain point, the strain in that area will suddenly increase. To describe this amplitude change, the localization index λ is used. Finite element analysis is used to calculate the ratio of the maximum strain to the average strain. The localization index λ of a healthy blade is approximately 1.5.

[0076] In some embodiments, the second type of evidence includes at least two of the following: tip amplitude anomaly coefficient, natural frequency drift rate, damping anomaly coefficient, and multi-blade consistency coefficient.

[0077] The blade tip amplitude anomaly coefficient is determined based on the ratio of the measured blade tip vibration amplitude to the reference amplitude of a healthy blade; the natural frequency drift rate is determined based on the ratio of the difference between the measured blade frequency and the reference frequency of a healthy blade to the reference frequency; the damping anomaly coefficient is determined based on the ratio of the difference between the measured damping ratio and the reference damping ratio of a healthy blade to the reference damping ratio; and the multi-blade consistency coefficient is determined based on the ratio of the vibration amplitude of a single blade to the average vibration amplitude of all blades in the same class.

[0078] Specifically, for y i Perform a Fourier transform on (t) to obtain the tip vibration amplitude of the i-th blade. A i The frequency is f i Define the tip amplitude anomaly coefficient μ as: , The reference amplitude for a healthy blade;

[0079] It should be noted that when μ≈1, the amplitude is normal. A significant increase in μ may indicate intensified vibration. If it increases suddenly, it may indicate intensified high-cycle fatigue.

[0080] Define the natural frequency drift rate F as: ;

[0081] It should be noted that when cracks appear in the blade, the stiffness decreases, leading to a reduction in the blade's natural frequency. This phenomenon is characterized by the natural frequency drift rate F, where... This indicates the measured blade frequency. This represents the reference frequency for blade health; when F≈0, it indicates that the stiffness remains unchanged; the larger F is, the more dangerous the blade is.

[0082] Define the damping anomaly coefficient ;

[0083] It should be noted that when blades develop cracks or loosening, or in the early stages of blade flutter or instability, the blade damping ratio will become abnormal, which is characterized by the damping anomaly coefficient D; among which, D represents the baseline damping ratio of a healthy blade, and ζ represents the measured damping ratio. The damping ratio of a healthy blade is relatively constant. When a crack appears in the blade, the opening and closing friction of the crack surface causes the damping ratio to change. In the early stages of blade flutter and instability, the blade damping ratio drops sharply. Therefore, the larger D is, the greater the deviation of the damping from the healthy state, and the higher the risk.

[0084] Define the multi-blade consistency coefficient C as: Where N is the total number of blades.

[0085] It should be noted that the multi-blade consistency index C is used to measure whether the vibration of each blade on the same level of the disk is uniform. When the blades at each level are in very good condition, the vibration of each blade is very similar. When a blade has a crack, reduced stiffness, or loosened root, only the amplitude of that blade increases significantly, while the vibration of other blades remains basically unchanged. Therefore, the larger the C is, the greater the difference between the blades and the higher the risk of damage to a single blade.

[0086] It should be noted that this step extracted six pieces of evidence characterizing the blade state from two types of sensor signals, including dynamic stress ratio, localization index, tip amplitude anomaly coefficient, natural frequency drift rate, damping anomaly coefficient, and multi-blade consistency coefficient. These pieces of evidence reflect the health status of the blade from different dimensions and provide a rich information foundation for subsequent multi-evidence fusion.

[0087] Based on the above embodiments, this embodiment will provide a detailed description of step S104:

[0088] In some embodiments, the DS evidence theory is used to fuse the multiple individual pieces of evidence for decision-making, and the resulting safety status of the blade structure is output as follows:

[0089] Each piece of evidence is pre-assigned a confidence level corresponding to a different risk level;

[0090] The confidence levels of all individual pieces of evidence are merged to obtain the overall confidence level of the blade belonging to each risk level;

[0091] The result of determining the safety status of the blade structure is the risk level corresponding to the highest comprehensive confidence level.

[0092] Specifically, the risk level set is constructed as follows: Let R1 represent safety, R2 represent early warning, and R3 represent alarm. For any subsets R1, R2, and R3, basic probability values ​​m(R1), m(R2), and m(R3) represent the confidence levels of R1, R2, and R3. The uncertainty level is represented by m(σ). A trust function and a likelihood function are used to describe the confidence interval of a subset.

[0093] Furthermore, the risk level includes at least three levels: safety, early warning, and alarm. The comprehensive confidence level includes the confidence level of the blade state for each risk level and a confidence level representing an uncertain state.

[0094] The final blade state can be one of four types: safe, warning, alarm, or uncertain, with corresponding confidence levels of [m(R1), m(R2), m(R3), m(σ)]; the confidence function is: The risk level is: If the calculated blade state is R1, then its confidence interval is: .

[0095] Based on experience, examples of the confidence levels for blades to be in safe, early warning, and alarm states are shown in Table 1 below, using the dynamic stress ratio η as an example:

[0096] Table 1. Confidence level using dynamic stress ratio η as an example

[0097]

[0098] Based on experience, examples of the confidence levels for blades to be in safe, early warning, and alarm states are shown in Table 2 below, using the blade tip amplitude anomaly coefficient μ as an example:

[0099] Table 2 Confidence level using the tip amplitude anomaly coefficient μ as an example

[0100]

[0101] Based on experience, examples of the confidence levels for blades to be in safe, early warning, and alarm states are shown in Table 3 below, using the natural frequency drift rate F as an example:

[0102] Table 3 Confidence level using natural frequency drift rate F as an example

[0103]

[0104] Based on experience, examples of the confidence levels for blades to be in safe, early warning, and alarm states are shown in Table 4 below, using the damping anomaly coefficient D as an example:

[0105] Table 4 shows the confidence level using the damping anomaly coefficient D as an example.

[0106]

[0107] Based on experience, examples of the confidence levels for blades to be in safe, early warning, and alarm states are shown in Table 5 below, using the multi-blade consistency index C as an example:

[0108] Table 5. Confidence level of multi-leaf consistency index C as an example

[0109]

[0110] In other embodiments, when using the DS evidence theory to make a fusion decision on the multiple individual pieces of evidence, a pre-constructed correction coefficient is used to modify the fusion process:

[0111] Multiple types of first-type evidence extracted from local dynamic strain signals are fused at the first level to obtain the first fused evidence;

[0112] Multiple pieces of second-type evidence extracted from the blade tip vibration displacement signal are fused at the first level to obtain the second fused evidence.

[0113] Based on the pre-constructed correction coefficients, the first fused evidence and the second fused evidence are fused at a second level to obtain the final fusion decision result. The correction coefficients are used as weighting factors to adjust the conflict coefficients between different evidence sources in the DS evidence theory fusion formula.

[0114] Specifically, for the two types of evidence for judging the blade state obtained from dynamic strain acquisition—dynamic stress ratio and localization—the following formula is used for first-level fusion to obtain the first fused evidence. R i The confidence level of the event is:

[0115]

[0116]

[0117] in, The basic probability assignment represents the evidence of dynamic stress ratio, and the confidence level for judging the blade to be at each risk level based on the dynamic stress ratio η is represented. λ represents the basic probability assignment of the evidence of localization index, and λ represents the confidence level of judging the leaf to be at each risk level based on the localization index λ. This represents the basic probability assignment of the first fused evidence; This indicates the risk level, where i=1,2,3 correspond to safe, early warning, and alarm, respectively. This represents the focal element (i.e., the subset of risk levels supported by the evidence) corresponding to each piece of evidence. This represents the conflict coefficient.

[0118] For the four types of second-order evidence for judging blade status obtained from BTT data collection—tip amplitude anomaly coefficient, frequency drift, damping anomaly coefficient, and multi-blade consistency—the following formula is used for first-level fusion to obtain second-order fused evidence. The R-value of the second-order fused evidence is... i The confidence level of the event is:

[0119]

[0120]

[0121] in, The basic probability assignment of the leaf tip amplitude anomaly coefficient evidence represents the confidence level of judging the leaf blade's risk level based on the leaf tip amplitude anomaly coefficient μ. The basic probability assignment represents the evidence of frequency drift rate, and the confidence level of judging the blade's risk level based on the inherent frequency drift rate F is represented. The basic probability assignment represents the evidence of the damping anomaly coefficient, and the confidence level of judging the blade's risk level based on the damping anomaly coefficient D is: The basic probability assignment represents the evidence of the multi-leaf consistency coefficient, and the confidence level of judging the leaf's risk level based on the multi-leaf consistency coefficient C is; This represents the basic probability assignment of the second fused evidence. This represents the focal element (i.e., the subset of risk levels supported by the evidence) corresponding to each piece of evidence. Indicates the conflict coefficient; The empty set represents a risk level where the evidence is completely conflicting and has no common support.

[0122] Furthermore, the process of constructing the correction coefficients includes:

[0123] Based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key parts, correction coefficients are constructed to correct the fusion process.

[0124] The transfer ratio parameter is determined based on the ratio of the blade tip vibration displacement signal amplitude to the dynamic strain signal amplitude of the key part, and the dynamic strain signal amplitude of the key part is obtained through finite element analysis or experimental calibration.

[0125] Specifically, tip timing (BTT) and strain transfer ratio (DST) are the core parameters for non-contact dynamic strain monitoring of blades. Their calculation essentially establishes a quantitative mapping relationship between tip vibration displacement and dynamic strain at key locations (such as the blade root). The value of h represents the relationship between the tip vibration amplitude (BTT) and the maximum local strain.

[0126]

[0127] In the formula, A i Ac represents the amplitude of the blade tip vibration displacement signal, which is obtained from finite element analysis or experimental calibration.

[0128] Furthermore, based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key parts, correction coefficients for correcting the fusion process are constructed, including:

[0129] Based on the transfer ratio parameter and strain threshold, the probability of blade failure is calculated, and a correction coefficient is obtained.

[0130] Specifically, based on the transfer ratio parameter of BTT and strain, the difference between the maximum strain and the strain threshold of the blade is obtained. This probability is described exponentially, and the probability is:

[0131]

[0132] Thus, we obtain evidence from the strain-displacement transfer ratio parameter of the blade, indicating that the probability of blade failure is P and the probability of no failure is 1-P. The adjustment coefficient (adjustable parameter) is used to control the steepness of the probability curve, that is, to control the sensitivity of the failure probability to strain exceeding the limit. This is the strain threshold.

[0133] It should be noted that this formula maps the degree of strain exceeding the limit to the failure probability P through an exponential function. much smaller When P approaches 0; when Approaching or exceeding At this point, P rapidly increases and approaches 1, enabling continuous quantitative assessment of blade failure risk.

[0134] In one specific embodiment, the DS formula is combined with the blade failure probability obtained considering the strain transfer ratio parameter h for a second-level fusion to obtain the final fusion decision result. The confidence fusion calculation formula for the BTT measurement results and dynamic strain measurement results is as follows:

[0135]

[0136]

[0137]

[0138]

[0139]

[0140]

[0141] It should be noted that, Assign a base probability value to the final fusion decision result.

[0142] It should be noted that the final determination of the blade condition and its confidence level are obtained by sequentially fusing these six sets of evidence. Compared with the conventional weighted average, this method does not require setting weights. It intelligently calculates the conflict coefficient based on the evidence conflicts of multiple data, avoiding the subjectivity of manually setting weights and improving the credibility of blade safety monitoring.

[0143] This invention discloses a method, device, equipment, and computer-readable storage medium for online monitoring of the structural safety of heavy-duty gas turbine blades, belonging to the field of gas turbine health monitoring technology. Addressing the technical problem that existing single-sensor monitoring methods cannot simultaneously achieve both monitoring range and quantitative accuracy, this invention proposes a blade safety monitoring scheme based on BTT / local strain coordination. In the technical solution, strain gauges are first arranged on at least one blade to obtain local dynamic strain signals, and simultaneously, blade tip timing sensors are arranged circumferentially in the stage where the blade is located to obtain blade tip vibration displacement signals. Then, based on the dynamic strain signals, two types of dynamic strain evidence, dynamic stress ratio and localization index, are extracted, and based on the blade tip vibration displacement signals, four types of blade tip timing evidence, blade tip amplitude anomaly coefficient, natural frequency drift rate, damping anomaly coefficient, and multi-blade consistency coefficient, are extracted. Furthermore, the DS evidence theory is used to fuse the above six types of evidence for decision-making. Specifically, this includes pre-setting a confidence level assignment for each piece of evidence corresponding to three risk levels: safety, warning, and alarm. The confidence level assignments of all evidence are fused to obtain a comprehensive confidence level, and the risk level corresponding to the highest comprehensive confidence level is determined as the final safe state. As a preferred solution, a correction coefficient based on the transfer ratio parameter is also introduced to optimize the fusion process. This transfer ratio parameter is determined based on the ratio of the blade tip vibration displacement amplitude to the dynamic strain amplitude of the key parts. The probability of blade failure is calculated through an exponential function, and this probability is used as a weighting factor to adjust the conflict coefficient between different evidence sources in the DS evidence theory. This invention combines the high-precision local measurement capabilities of strain gauges with the non-contact measurement advantages of BTT across the entire blade, and employs DS evidence theory to intelligently fuse multi-source evidence. This effectively overcomes the inherent defects of single monitoring technologies in terms of monitoring range, quantitative accuracy, and sensitivity to early faults, avoids the subjectivity of manually setting weights, and achieves highly reliable online monitoring of the structural safety status of the entire blade.

[0144] Please refer to Figure 2 , Figure 2 A structural block diagram of a heavy-duty gas turbine blade structure safety online monitoring device provided in an embodiment of the present invention is shown. The specific device includes:

[0145] The strain gauge module 100 is disposed on at least one blade and is used to acquire local dynamic strain signals of key parts of the blade.

[0146] The blade tip timing sensor module 200 is located in the circumferential direction of the blade stage and is used to acquire the blade tip timing pulse signal when all blades rotate past, and to invert the blade tip vibration displacement signal based on the blade tip timing pulse signal.

[0147] The evidence extraction module 300 is connected to the strain gauge module and the blade tip timing sensor module, and is used to extract multiple single pieces of evidence characterizing the blade structure state based on the local dynamic strain signal and the blade tip vibration displacement signal, wherein the single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal.

[0148] The fusion decision module 400 is connected to the evidence extraction module and is used to perform fusion decision-making on the multiple single pieces of evidence using DS evidence theory, and output the safety status result of the blade structure.

[0149] The heavy-duty gas turbine blade structure safety online monitoring device of this embodiment is used to implement the aforementioned heavy-duty gas turbine blade structure safety online monitoring method. Therefore, the specific implementation of the heavy-duty gas turbine blade structure safety online monitoring device can be found in the embodiment section of the heavy-duty gas turbine blade structure safety online monitoring method above. For example, the strain gauge module 100, the blade tip timing sensor module 200, the evidence extraction module 300, and the fusion decision module 400 are respectively used to implement steps S101, S102, S103, and S104 in the above-mentioned heavy-duty gas turbine blade structure safety online monitoring method. Therefore, its specific implementation can be referred to the description of the corresponding embodiments, and will not be repeated here.

[0150] A specific embodiment of the present invention also provides an online monitoring device for the safety of heavy-duty gas turbine blade structures, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the above-described online monitoring method for the safety of heavy-duty gas turbine blade structures.

[0151] A specific embodiment of the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for online safety monitoring of heavy-duty gas turbine blade structures.

[0152] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0153] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0154] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0155] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0156] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for online monitoring of the structural safety of heavy-duty gas turbine blades, characterized in that, include: Strain gauges are placed on at least one blade to obtain local dynamic strain signals at critical parts of the blade; A blade tip timing sensor is arranged circumferentially in the stage where the blade is located to obtain the blade tip timing pulse signal when all blades rotate past, and the blade tip vibration displacement signal is obtained by inversion based on the blade tip timing pulse signal. Based on the local dynamic strain signal and the blade tip vibration displacement signal, multiple single pieces of evidence characterizing the blade structural state are extracted respectively. The single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal. The DS evidence theory is used to fuse multiple individual pieces of evidence for decision-making, and output the safety status result of the blade structure. This includes: pre-setting a confidence level assignment for each individual piece of evidence corresponding to different risk levels; fusing the confidence level assignments of all individual pieces of evidence to obtain the comprehensive confidence level of the blade belonging to each risk level; and determining the safety status result of the blade structure as the risk level corresponding to the highest comprehensive confidence level. In the process of using the DS evidence theory to fuse multiple single pieces of evidence for decision-making, a pre-constructed correction coefficient is used to correct the fusion process: multiple first-type pieces of evidence extracted from local dynamic strain signals are fused at the first level to obtain first fused evidence; multiple second-type pieces of evidence extracted from blade tip vibration displacement signals are fused at the first level to obtain second fused evidence; based on the pre-constructed correction coefficient, the first fused evidence and the second fused evidence are fused at the second level to obtain the final fused decision result. The correction coefficient is used in the DS evidence theory fusion formula as a weighting factor to adjust the conflict coefficient between different evidence sources. The construction process of the correction coefficient includes: constructing correction coefficients for correcting the fusion process based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key part, wherein the transfer ratio parameter is determined based on the ratio of the amplitude of the blade tip vibration displacement signal to the amplitude of the dynamic strain signal of the key part, and the amplitude of the dynamic strain signal of the key part is obtained through finite element analysis or experimental calibration.

2. The method according to claim 1, characterized in that, The arrangement of strain gauges on at least one blade includes: Select a portion of the blades along the circumferential direction and attach strain gauges to at least two key locations in the root maximum stress zone, the middle section of the inlet edge, and the middle of the blade body for each selected blade.

3. The method according to claim 1, characterized in that, The step of obtaining the blade tip vibration displacement signal by inverting the blade tip timing pulse signal includes: The ideal arrival pulse time for each blade under vibration-free conditions is calculated based on the rotor speed and the number of blades. The time difference is obtained by comparing the actual measured pulse arrival time with the ideal pulse arrival time. By combining the blade's rotation radius with the current rotation speed information, the time difference is converted into a blade tip vibration displacement signal.

4. The method according to claim 1, characterized in that, The first type of evidence includes dynamic stress ratio and localization index. The dynamic stress ratio is determined based on the ratio of the amplitude of the local dynamic strain signal at a preset modal frequency to the allowable dynamic stress at the corresponding location. The localization index is determined based on the ratio of the maximum value to the average value of the strain amplitude at multiple measuring points in the local dynamic strain signal.

5. The method according to claim 1, characterized in that, The second type of evidence includes at least two of the following: tip amplitude anomaly coefficient, natural frequency drift rate, damping anomaly coefficient, and multi-blade consistency coefficient. The blade tip amplitude anomaly coefficient is determined based on the ratio of the measured blade tip vibration amplitude to the reference amplitude of a healthy blade; the natural frequency drift rate is determined based on the ratio of the difference between the measured blade frequency and the reference frequency of a healthy blade to the reference frequency; the damping anomaly coefficient is determined based on the ratio of the difference between the measured damping ratio and the reference damping ratio of a healthy blade to the reference damping ratio; and the multi-blade consistency coefficient is determined based on the ratio of the vibration amplitude of a single blade to the average vibration amplitude of all blades in the same class.

6. The method according to claim 1, characterized in that, The risk level includes at least three levels: safety, early warning, and alarm. The overall confidence level includes the confidence level of the blade status for each risk level and a confidence level representing an uncertain state.

7. The method according to claim 1, characterized in that, The step of constructing correction coefficients for the fusion process based on the transfer ratio parameter between the blade tip vibration displacement signal and the dynamic strain signal of the key parts includes: Based on the transfer ratio parameter and strain threshold, the probability of blade failure is calculated, and a correction coefficient is obtained.

8. A heavy-duty gas turbine blade structure safety online monitoring device, employing the heavy-duty gas turbine blade structure safety online monitoring method according to any one of claims 1-7, characterized in that, include: A strain gauge module is installed on at least one blade to acquire local dynamic strain signals at key parts of the blade. The blade tip timing sensor module is located in the circumferential direction of the blade level and is used to acquire the blade tip timing pulse signal when all blades rotate past, and to invert the blade tip vibration displacement signal based on the blade tip timing pulse signal. An evidence extraction module, connected to the strain gauge module and the blade tip timing sensing module, is used to extract multiple single pieces of evidence characterizing the blade structure state based on the local dynamic strain signal and the blade tip vibration displacement signal, wherein the single pieces of evidence include at least a first type of evidence extracted from the local dynamic strain signal and a second type of evidence extracted from the blade tip vibration displacement signal. The fusion decision module, connected to the evidence extraction module, is used to perform fusion decision-making on the multiple single pieces of evidence using DS evidence theory and output the safety status result of the blade structure.

9. A heavy-duty gas turbine blade structure safety online monitoring device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the online monitoring method for the safety of heavy-duty gas turbine blade structures as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the online monitoring method for the safety of heavy-duty gas turbine blade structures as described in any one of claims 1 to 7.

Images