A method and apparatus for predicting a dynamic tip clearance of a gas turbine

Abstract

The present application is suitable for the field of gas turbine technology, and provides a gas turbine blade tip dynamic gap prediction method and device. In the embodiment, first, a temperature field calculation method of the blade, turbine disc and turbine holding ring, and an initial blade tip gap calculation method are established, then the actual overall thermal parameters, metal wall surface temperature, rotating speed, rotor axis position and blade tip gap value of the gas turbine are obtained through sensors to correct the above calculation methods, and the dynamic change of the blade tip gap is determined through the corrected model and transient parameters under high rotating speed and high load. This solves the problem that the existing blade tip gap control technology based on the control strategy of static blade tip gap margin is prone to accidents under high rotating speed and high load transient conditions. And the gap value under the transient condition is predicted, which can effectively reduce the risk of blade tip rubbing, reduce the blade structure damage caused by rubbing, improve the operation safety of the gas turbine, and avoid the loss caused by unplanned shutdown.

Description

Technical Field

[0001] This application relates to the field of gas turbine technology, and in particular to a method and apparatus for predicting dynamic clearance at the tip of a gas turbine blade. Background Technology

[0002] Tip clearance, as a core structural parameter between rotating blades and stator turbine rings in turbine machinery, directly affects the aerodynamic efficiency, operational safety, and lifespan of critical components of the unit. With the trend towards higher turbine inlet temperatures and higher efficiency in modern gas turbines, precise control of tip clearance has become a key technological bottleneck restricting the improvement of unit performance.

[0003] The core challenge facing current tip clearance control technology lies in the difficulty of accurately predicting the clearance evolution under dynamic operating conditions during the design phase. Traditional methods mainly rely on thermodynamic models and empirical correction coefficients for clearance design, failing to fully consider the following key factors: First, during transient processes such as unit start-up and shutdown, and changes in operating conditions, the significant differences in thermal inertia exhibited by the rotor system and stator structure due to material differences and different heat transfer paths lead to a time phase difference in their thermal deformation response. Second, the three-dimensional temperature gradient field formed by large-mass components such as turbine disks and turbine bearings during heat transfer can induce local deformations that are difficult to quantify. The interaction of these factors makes existing control strategies based on static tip clearance margins highly susceptible to dynamic-static rubbing accidents under high-speed, high-load transient conditions, causing problems such as blade coating peeling and damage to the gas seal structure, and in severe cases, even leading to unplanned unit shutdowns. Summary of the Invention

[0004] In view of this, this application provides a method and apparatus for predicting dynamic tip clearance of gas turbine blades, so as to accurately quantify the change in tip clearance under transient conditions, thereby solving the problem that the control strategy of static tip clearance margin is prone to causing accidents under transient conditions of high speed and high load.

[0005] The first aspect of this application provides a method for predicting dynamic tip clearance of a gas turbine blade, the method comprising: Acquire first monitoring data for model correction under actual operating conditions of the gas turbine, and second monitoring data for comparison with the results calculated by the model; The turbine inlet flow rate is determined using the first monitoring data. The heat transfer coefficient of the target component is then determined using the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data. The heat transfer coefficient and the temperature data in the first monitoring data are input into the temperature field calculation model to obtain the temperature field information of the target component. The second monitoring data is then compared with the temperature field information, and the temperature field calculation model is corrected based on the comparison results to obtain the target temperature field calculation model. The target component includes blades, turbine disks, and turbine retaining rings. The estimated clearance value of the gas turbine is determined by using the temperature field information, the first monitoring data, and the initial tip clearance multi-field coupling calculation model based on the elongation calculation model of the target component. The estimated clearance value is compared with the actual measured clearance value, and the initial tip clearance multi-field coupling calculation model is corrected based on the comparison results to obtain the target tip clearance multi-field coupling calculation model. The dynamic change of the gas turbine tip clearance is determined by using a preset transient overall thermodynamic parameter time series, the target temperature field calculation model, and the target tip clearance multi-field coupling calculation model.

[0006] Optionally, the first monitoring data includes compressor inlet flow, turbine disk cooling gas flow, gas flow, and turbine ring cooling gas flow. Determining the turbine inlet flow rate using the first monitoring data includes: A flow pipe is installed in the compressor inlet system of the gas turbine to obtain the compressor inlet flow rate. ; An extraction system is installed at the compressor outlet of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine disk cooling gas flow rate. ; A flow meter is installed in the fuel gas system of the gas turbine to obtain the gas flow rate. ; An extraction system is installed between the compressor stages of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine ring cooling gas flow rate. ; Determine turbine inlet flow rate .

[0007] Optionally, the step of comparing the second monitoring data with the temperature field information and correcting the temperature field calculation model based on the comparison result includes: The metal wall temperature of the target component in the second monitoring data is compared with the temperature field information. When the comparison result is greater than the preset temperature difference threshold, the parameters of the temperature field calculation model are corrected using a long short-term memory neural network (LSTM). The temperature field information determined by the corrected temperature field calculation model is then compared with the metal wall temperature of the target component until the comparison result is less than the temperature difference threshold.

[0008] Optionally, determining the estimated clearance value of the gas turbine using a multi-field coupled calculation model of the initial tip clearance determined by temperature field information, first monitoring data, and the elongation calculation model of the target component includes: The blade elongation L1(t) is determined using a blade elongation calculation model. The elongation L2(t) of the turbine disk is determined by the calculation model of the turbine disk elongation. The rotor's shaft offset L3(t) is determined by the shaft position; The elongation L4(t) of the turbine retainer was determined using a calculation model for the elongation of the turbine retainer. The estimated gap value is determined by the formula L(t)=L0- L1(t)- L2(t)- L3(t)+ L4(t), where L0 is the installation gap.

[0009] Optionally, comparing the estimated clearance value with the actual measured clearance value and correcting the initial tip clearance multi-field coupling calculation model based on the comparison results includes: When the comparison result is greater than the preset gap difference, the RBF-BPNN neural network is used to construct the inverse response surface model, and the gap value determined by the actual measurement is substituted into the inverse response surface model to correct the initial blade tip gap multi-field coupling calculation model. When the comparison result is less than the preset gap difference, the initial blade tip gap multi-field coupling calculation model is determined as the target blade tip gap multi-field coupling calculation model.

[0010] A second aspect of this application provides a device for predicting dynamic tip clearance of a gas turbine blade, the device comprising: The data acquisition unit is used to acquire first monitoring data for model correction and second monitoring data for comparison with the model calculation results under the actual operating conditions of the gas turbine. The temperature field model correction unit is used to determine the turbine inlet flow rate through the first monitoring data, then determine the heat transfer coefficient of the target component through the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data, input the heat transfer coefficient and the temperature data in the first monitoring data into the temperature field calculation model to obtain the temperature field information of the target component, then compare the second monitoring data with the temperature field information, and correct the temperature field calculation model based on the comparison result to obtain the target temperature field calculation model, wherein the target component includes blades, turbine disk, and turbine retainer ring; The tip clearance model correction unit is used to determine the estimated clearance value of the gas turbine through the temperature field information, the first monitoring data and the initial tip clearance multi-field coupling calculation model determined based on the elongation calculation model of the target component, compare the estimated clearance value with the actual measured clearance value, and correct the initial tip clearance multi-field coupling calculation model based on the comparison result to obtain the target tip clearance multi-field coupling calculation model. The blade tip clearance result determination unit is used to determine the dynamic change result of the gas turbine blade tip clearance by means of a preset transient overall thermodynamic parameter time series, the target temperature field calculation model, and the target blade tip clearance multi-field coupling calculation model.

[0011] Optionally, the data acquisition unit may be equipped with a flow pipe within the compressor intake system of the gas turbine to obtain the compressor inlet flow rate. ; An extraction system is installed at the compressor outlet of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine disk cooling gas flow rate. ; A flow meter is installed in the fuel gas system of the gas turbine to obtain the gas flow rate. ; An extraction system is installed between the compressor stages of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine ring cooling gas flow rate. ; Determine turbine inlet flow rate .

[0012] Optionally, the temperature field correction unit's comparison of the second monitoring data with the temperature field information, and the correction of the temperature field calculation model based on the comparison result, includes: The metal wall temperature of the target component in the second monitoring data is compared with the temperature field information. When the comparison result is greater than the preset temperature difference threshold, the parameters of the temperature field calculation model are corrected using a long short-term memory neural network (LSTM). The temperature field information determined by the corrected temperature field calculation model is then compared with the metal wall temperature of the target component until the comparison result is less than the temperature difference threshold.

[0013] Optionally, the tip clearance model correction unit determines the estimated clearance value of the gas turbine using an initial tip clearance multi-field coupled calculation model determined by temperature field information, first monitoring data, and an elongation calculation model based on the target component. The blade elongation L1(t) is determined using a blade elongation calculation model. The elongation L2(t) of the turbine disk is determined by the calculation model of the turbine disk elongation. The rotor's shaft offset L3(t) is determined by the shaft position; The elongation L4(t) of the turbine retainer was determined using a calculation model for the elongation of the turbine retainer. The estimated gap value is determined by the formula L(t)=L0- L1(t)- L2(t)- L3(t)+ L4(t), where L0 is the installation gap.

[0014] Optionally, the comparison of the estimated clearance value with the actual measured clearance value in the blade tip clearance model correction unit, and the correction of the initial blade tip clearance multi-field coupling calculation model based on the comparison results, includes: When the comparison result is greater than the preset gap difference, the RBF-BPNN neural network is used to construct the inverse response surface model, and the gap value determined by the actual measurement is substituted into the inverse response surface model to correct the initial blade tip gap multi-field coupling calculation model. When the comparison result is less than the preset gap difference, the initial blade tip gap multi-field coupling calculation model is determined as the target blade tip gap multi-field coupling calculation model.

[0015] In the embodiments provided in this application, a temperature field calculation model for the blades, turbine disk, and turbine retaining ring is first established, along with an initial tip clearance calculation method. Then, the actual overall thermodynamic parameters of the gas turbine, metal wall temperature, rotational speed, rotor shaft position, and tip clearance value are acquired through sensors to correct the aforementioned calculation model and method. Finally, the dynamic changes in tip clearance are determined using the corrected model and transient parameters under high speed and high load. This solves the problem that existing tip clearance control technologies based on static tip clearance margins are prone to causing accidents under transient conditions of high speed and high load. Furthermore, predicting the tip clearance value under transient conditions can effectively reduce the risk of tip rubbing, minimize blade structural damage caused by rubbing, improve gas turbine operational safety, and avoid losses from unplanned shutdowns. Attached Figure Description

[0016] Figure 1 A flowchart illustrating the method provided in this application embodiment; Figure 2 This is a structural diagram of the device provided in the embodiments of this application; Figure 3 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0017] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0018] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0019] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0020] This application provides a method and apparatus for predicting dynamic clearance at the tip of a gas turbine blade, in order to reduce the probability of turbine accidents.

[0021] The technical solutions of this application will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0022] like Figure 1 The diagram shown is a flowchart of a method for predicting dynamic tip clearance of a gas turbine blade provided in this application. The process may include the following steps: Step S101: Obtain first monitoring data for model correction and second monitoring data for comparison with the model calculation results under the actual operating conditions of the gas turbine.

[0023] In this embodiment, the first monitoring data includes compressor inlet flow rate, turbine disk cooling gas flow rate and temperature, gas flow rate, turbine shroud cooling gas flow rate and temperature, turbine inlet gas temperature, rotor speed, and shaft center position. The second monitoring data includes the gas turbine clearance value, and the metal wall temperatures of the turbine disk, turbine shroud, and blades. The acquisition process for each data point is as follows: 1. Compressor inlet flow rate. A flow pipe can be installed in the compressor inlet system of the gas turbine to obtain the compressor inlet flow rate.

[0024] 2. Turbine inlet gas temperature. This data is obtained by measuring the total temperature using a probe installed along the blade height at the leading edge of the turbine stator blade.

[0025] 3. Cooling gas flow rate of turbine disk and turbine retainer ring. The cooling gas flow rate of turbine disk is the compressor extraction cooling flow rate, and this flow rate and the cooling gas flow rate of turbine retainer ring can be directly measured by a flow meter.

[0026] 4. Cooling gas temperature of turbine disk and turbine retainer ring. Temperature probes can be installed at the blade cooling gas inlet channel to obtain the cooling gas temperature of the blade at different times; temperature probes can be installed in the turbine disk cavity to obtain the cooling gas temperature of the turbine disk at different times; and temperature probes can be installed in the turbine retainer ring to obtain the cooling gas temperature of the turbine retainer ring at different times.

[0027] 5. Rotor speed and shaft center position. Rotor speed, as one of the most fundamental and critical parameters in gas turbine monitoring and control, can be directly obtained in real time through sensors. Shaft center position can be obtained at different time intervals by installing eddy current sensors on the rotor journals.

[0028] 6. Temperature of the metal wall surfaces of the turbine disk, turbine retaining ring, and blades.

[0029] Tiny thermocouples are placed along the circumference of the blade wall to obtain the temperature of the metal wall of the blade. Armored thermocouples were installed on the turbine disk wall to obtain the turbine disk metal wall temperature for different durations. Armored thermocouples were installed on the turbine retainer wall to obtain the temperature of the turbine retainer metal wall at different durations.

[0030] 7. Clearance value. This data is the measured data of the blade tip clearance during gas turbine operation. It is determined by measuring the blade tip clearance using a rotor blade tip clearance probe installed on the turbine bearing ring.

[0031] Using the above method, we can obtain the first and second monitoring data for different durations, which can then be used to correct the model.

[0032] Step S102: Determine the turbine inlet flow rate using the first monitoring data, then determine the heat transfer coefficient of the target component using the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data, input the heat transfer coefficient and the temperature data in the first monitoring data into the temperature field calculation model to obtain the temperature field information of the target component, then compare the second monitoring data with the temperature field information, and correct the temperature field calculation model based on the comparison result to obtain the target temperature field calculation model.

[0033] In this embodiment, the target components include gas turbine blades, turbine shrouds, and turbine disks. Three temperature field models need to be trained separately: a blade temperature field calculation model, a turbine disk temperature field calculation model, and a turbine shroud temperature field calculation model. These models are trained using historical data consisting of the historical heat transfer coefficients, coolant temperatures, turbine inlet gas temperatures, and corresponding temperature fields of the blades, turbine shrouds, and turbine disks. This allows for the determination of the temperature field information for the blades, turbine shrouds, and turbine disks under different heat transfer coefficients, coolant temperatures, and turbine inlet gas temperatures.

[0034] After training, the temperature field information of the gas turbine is determined by using the real-time heat transfer coefficients of the blades, turbine shrouds, and turbine disks, as well as the cold air temperature and turbine inlet gas temperature, based on the initial monitoring data. The process for determining the heat transfer coefficients is as follows: 1. The heat transfer coefficient of the blade. This heat transfer coefficient is expressed by the formula... Confirmed. In this formula, Nu b e represents the Nusselt number of the blade. b Re represents a constant. b and Cw b h represents the Reynolds number and flow coefficient of the blade, respectively. b The internal / external heat transfer coefficient of the blade, L. b Typically, the leaf spread or height is taken, λ b The thermal conductivity of the blade is represented by m. b Indicates turbine inlet flow rate, μ b The blade dynamic viscosity is represented by n1 and m1, which are exponential constants. Among the parameters above, the turbine inlet flow rate is expressed by the formula... Confirmed. In this formula, This is the compressor inlet flow rate. This refers to the turbine disk cooling airflow. For gas flow rate, The turbine ring cooling gas flow rate is determined in step S101 for all the above parameters. d The rotational Reynolds number is obtained through the formula... Confirmed, V b The characteristic velocity, in this case, is the relative velocity of the incoming flow at the leading edge of the blade, which can be calculated from the rotor speed and structural parameters. The gas density at the blade surface was obtained by referring to the "Table of Thermophysical Properties of Liquids and Gases". Cw b Through formula Determined. Constant e b n1 and m1 and μ b , λ b This was obtained by consulting "Air Systems and Heat Transfer Analysis, Volume 16 of the Aircraft Engine Design Manual".

[0035] 2. The heat transfer coefficient of the turbine disk. This coefficient is obtained through the formula... Confirmed. In this formula, Nu d e represents the Nusselt number of the turbine disk. d Re represents a constant. d and Cw d h represents the Reynolds number and flow coefficient of the turbine disk, respectively. d L is the convective heat transfer coefficient of the turbine disk. d Indicates the characteristic length, turbine disk L d Typically, the outer diameter r and λ are taken. d The value m represents the thermal conductivity of the turbine disk. d Indicates the turbine disk cooling air flow rate, μ d Represents the dynamic viscosity of the turbine disk surface, where n2 and m2 are exponential constants. Among the parameters mentioned above, the cooling gas flow rate of the turbine disk has already been determined in step S101. d The rotational Reynolds number is obtained through the formula... Confirmed, r d Where is the radius of the turbine disk. For rotational speed, r is the gas density at the turbine disk surface. d and It can be measured directly. Obtained by consulting the "Table of Thermophysical Properties of Liquids and Gases". Cw d Through formula Determined. Constant e d n2 and m2 and μ d , λ d This was obtained by consulting "Air Systems and Heat Transfer Analysis, Volume 16 of the Aircraft Engine Design Manual".

[0036] 3. The heat transfer coefficient of the turbine retaining ring. This coefficient is obtained through the formula... Confirmed. In this formula, Nu s Indicates the turbine ring Nusselt number, e s Re represents a constant. s and Pr s h represents the Reynolds number of the turbine bearing and the Prandtl number of the gas, respectively. s L is the convective heat transfer coefficient of the turbine ring. s The characteristic length, λ, can be the axial chord length of the turbine retainer ring. s The value m represents the thermal conductivity of the turbine ring. s Indicates the turbine ring cooling gas flow rate, μ s Re represents the dynamic viscosity of the turbine bearing surface, where n3 and m3 are exponential constants. Among the parameters mentioned above, the cooling gas flow rate of the turbine bearing has already been determined in step S101. s The rotational Reynolds number is obtained through the formula... Confirmed, Vs The characteristic velocity is the average axial velocity on the annular cross section. The gas density at the outer edge of the boundary layer on the turbine ring surface can be obtained through CFD simulation or empirical formulas. The constant e s n3 and m3 and μ s , λ s Pr s This was obtained by consulting "Air Systems and Heat Transfer Analysis, Volume 16 of the Aircraft Engine Design Manual".

[0037] After determining the heat transfer coefficients of the blades, turbine bearings, and turbine disks, the corresponding temperature field is then determined.

[0038] When determining the temperature field of the blade, the obtained blade heat transfer coefficient and turbine inlet gas temperature are set as the boundary conditions of the blade temperature field calculation model, and the temperature field of the blade at different durations can be calculated.

[0039] When determining the temperature field of the turbine disk, the obtained turbine disk heat transfer coefficient and turbine disk cold air temperature are set as the boundary conditions of the turbine disk temperature field calculation model, and the temperature field of the turbine disk at different durations is calculated.

[0040] When determining the temperature field of the turbine ring, the obtained turbine ring heat transfer coefficient and turbine ring coolant temperature are set as the boundary conditions of the turbine ring temperature field calculation model, and the temperature field of the turbine disk at different durations is calculated.

[0041] The temperature of the metal wall surface of the blades, turbine disk, and turbine ring is compared with the temperature field information. When the comparison result is greater than the preset temperature difference threshold, the temperature field calculation model is corrected, and the temperature field information determined by the corrected temperature field calculation model is compared with the temperature of the metal wall surface of the blades, turbine disk, and turbine ring until the comparison result is less than the temperature difference threshold.

[0042] In this embodiment, after determining the wall temperature of each target component, it is compared with the temperature field information under actual operating conditions. When the comparison result, that is, the difference between the two, exceeds a preset value, such as 10°C, a Long Short-Term Memory (LSTM) neural network is used to correct the temperature field calculation model parameters of the blades, turbine bearing rings, and turbine disk. The specific method is as follows: The convective heat transfer coefficient is corrected by using a long short-term memory neural network (LSTM) combined with experimental data on the metal wall temperature, the thermal boundary conditions of the structure are adjusted, and the temperature field calculation model is then corrected.

[0043] The specific calculation formula for short-term memory networks is as follows:

[0044]

[0045]

[0046]

[0047] In the above formula, x t The input to the Long Short-Term Memory (LSTM) network is specifically measured data of the metal wall temperature, h. t C represents the hidden state at the current moment. t For the memory of the current moment, i t f t o t These are the t-th input gate, forget gate, and output gate, respectively, where the output is the convective heat transfer coefficient to be corrected. The corrected convective heat transfer coefficient is used as the boundary condition for thermal analysis to obtain the predicted temperature values ​​at each temperature measurement point. The measured and predicted values ​​at each temperature measurement point are fused with the hidden state ht output by the LSTM network to construct the error loss function, as follows: In this formula, N represents the number of temperature measurement points, and α is the fusion coefficient, ranging from 0.1 to 0.3. The gradient descent algorithm is used to minimize the error loss function, adjusting the convective heat transfer coefficient in the temperature field calculation model until the root mean square error between the predicted and measured temperatures is less than 3%.

[0048] After correction using the aforementioned neural network, the temperature field information of the blades, turbine bearings, and turbine disk under actual operating conditions is determined using the corrected temperature field calculation model. This model is then compared with the measured metal wall temperature. If the difference is still greater than 10℃, the temperature field calculation model is corrected again until the difference is less than 10℃. This step uses LSTM to correct the temperature field model, thereby improving its accuracy.

[0049] Step S103: Using the temperature field information, the first monitoring data, and the initial tip clearance multi-field coupling calculation model determined based on the elongation calculation model of the target component, the estimated clearance value of the gas turbine is determined. The estimated clearance value is compared with the clearance value determined by actual measurement, and the initial tip clearance multi-field coupling calculation model is corrected based on the comparison result to obtain the target tip clearance multi-field coupling calculation model.

[0050] In this embodiment, the initial tip clearance multi-field coupled calculation model consists of three models: a blade elongation calculation model, a turbine disk elongation calculation model, and a turbine retainer elongation calculation model. These elongation calculation models are trained using the historical temperature fields, historical rotational speeds, and corresponding historical elongations of the blade, turbine disk, and turbine retainer, respectively, to determine the elongation of each model under different temperature fields and rotational speeds.

[0051] After training, the blade elongation calculation model determines the blade elongation L1(t) at different durations based on the blade temperature field and rotor speed under the actual operating conditions. The turbine disk elongation calculation model determines the turbine disk elongation L2(t) at different durations based on the turbine disk temperature field and rotor speed. The turbine retainer elongation calculation model determines the turbine retainer elongation L4(t) at different durations based on the turbine retainer temperature field and rotor speed. Combined with the shaft center position determined in step S101, i.e., the shaft center offset L3(t), the above initial tip clearance multi-field coupled calculation model determines the estimated tip clearance value of the gas turbine using the formula L(t) = L0 - L1(t) - L2(t) - L3(t) + L4(t). In the above formula, L0 is the installation clearance, which is measured during the assembly of the gas turbine.

[0052] After determining the estimated tip clearance value, it is compared with the measured tip clearance value. If the deviation is greater than the preset value, such as 0.02 mm, it indicates that the deviation is too large, and the initial tip clearance multi-field coupling calculation model needs to be corrected. The specific process is as follows: An RBF-BPNN neural network was used as the fitting function for the interval inverse response model. A sample database was established using finite element simulation data under different operating conditions to construct an inverse response surface surrogate model. Based on the trained RBF-BPNN, measured blade tip clearance data under different operating conditions were used as input, and the output was the parameters to be corrected. By combining the inverse response surface surrogate model with measured blade tip clearance data under different operating conditions, the physical model for calculating the clearance involving multiple fields was corrected.

[0053] The RBF-BPNN topology consists of three layers: an input layer, a hidden layer, and an output layer. The input layer inputs the blade tip clearance data samples X=(x1, x2, ..., xn) obtained from the initial finite element analysis into the network.

[0054] The hidden layer uses Gaussian radial basis functions and a Sigmoid function f(x) to establish the activation function, where the Gaussian radial basis functions and the hidden layer activation function are expressed as follows: , In this formula, , , represent the center and width of the radial basis function, i.e., the mean and variance of the input sample data; hj is the activation function of the j-th unit in the hidden layer; wij is the connection weight from the i-th neuron in the input layer to the j-th neuron in the hidden layer; is the threshold of the j-th neuron in the hidden layer.

[0055] The output layer outputs the parameters to be corrected, and their expression is: In the formula, yk' represents the output of the k-th unit of the output layer, i.e., the predicted value of the k-th parameter to be corrected. The thresholds of each layer and the connection weights between layers in the network are adjusted using the error function P through the backpropagation algorithm. P represents the error between the initial value and the predicted value of the parameter to be corrected. The error function P of the output result in RBF-BPNN is... In the formula, y k Let k be the sample to be corrected.

[0056] After correction by the aforementioned neural network, the initial tip clearance multi-field coupling calculation model is used to predict the tip clearance value. The predicted tip clearance value is then compared with the measured tip clearance value. If the difference between the two is still greater than 0.02 mm, the initial tip clearance multi-field coupling calculation model is corrected again until the difference is less than 0.02 mm. Then, the initial tip clearance multi-field coupling calculation model at this time is determined as the model required by this application, namely the aforementioned target tip clearance multi-field coupling calculation model.

[0057] Step S104: Determine the dynamic change result of the gas turbine tip clearance by using the preset transient overall thermodynamic parameter time series, the target temperature field calculation model, and the target tip clearance multi-field coupling calculation model.

[0058] In this embodiment, the time series of transient overall thermodynamic performance parameters includes transient compressor inlet flow rate, transient turbine disk cooling gas flow rate and temperature, transient gas flow rate, transient turbine ring cooling gas flow rate and temperature, transient turbine inlet gas temperature, transient speed, and shaft center position. This data consists of various parameter curves of the transient overall thermodynamic performance that the user wants to study, such as transient turbine inlet flow rate, gas flow rate, compressor inlet flow rate, compressor extraction flow rate, turbine ring flow rate variation curve, transient turbine blade cooling gas temperature variation curve, transient turbine ring cooling gas temperature variation curve, and transient turbine disk cooling gas temperature variation curve. The horizontal axis of each curve represents time, and the vertical axis represents the corresponding numerical value. These curves are discretized into scattered points, and then substituted into the previous temperature field model and the target blade tip clearance multi-field coupling calculation model to calculate the blade tip clearance. This yields the blade tip clearance values ​​at various times. Connecting these blade tip clearance values ​​then yields the dynamic blade tip clearance variation curve. This curve can be used to predict clearance values ​​under transient operating conditions, thereby effectively reducing the risk of blade tip rubbing, minimizing damage to blades and related structures caused by rubbing, and improving the operational safety of gas turbines.

[0059] This concludes the process. Figure 1 The process is shown below.

[0060] In the embodiments provided in this application, a temperature field calculation model for the blades, turbine disk, and turbine retaining ring is first established, along with an initial tip clearance calculation method. Then, the actual overall thermodynamic parameters of the gas turbine, metal wall temperature, rotational speed, rotor shaft position, and tip clearance value are acquired through sensors to correct the aforementioned calculation model and method. Finally, the dynamic changes in tip clearance are determined using the corrected model and transient parameters under high speed and high load. This solves the problem that existing tip clearance control technologies based on static tip clearance margins are prone to causing accidents under transient conditions of high speed and high load. Furthermore, predicting the tip clearance value under transient conditions can effectively reduce the risk of tip rubbing, minimize blade structural damage caused by rubbing, improve gas turbine operational safety, and avoid losses from unplanned shutdowns.

[0061] This application also provides a device for predicting dynamic tip clearance of gas turbine blades, such as... Figure 2 As shown, the device includes: The data acquisition unit 201 is used to acquire first monitoring data for model correction and second monitoring data for comparison with the model calculation results under the actual operating conditions of the gas turbine. The temperature field model correction unit 202 is used to determine the turbine inlet flow rate through the first monitoring data, then determine the heat transfer coefficient of the target component through the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data, input the heat transfer coefficient and the temperature data in the first monitoring data into the temperature field calculation model to obtain the temperature field information of the target component, then compare the second monitoring data with the temperature field information, and correct the temperature field calculation model based on the comparison result to obtain the target temperature field calculation model, wherein the target component includes blades, turbine disk, and turbine retainer ring; The tip clearance model correction unit 203 is used to determine the estimated clearance value of the gas turbine through the temperature field information, the first monitoring data and the initial tip clearance multi-field coupling calculation model determined based on the elongation calculation model of the target component, compare the estimated clearance value with the actual measured clearance value, and correct the initial tip clearance multi-field coupling calculation model based on the comparison result to obtain the target tip clearance multi-field coupling calculation model. The blade tip clearance result determination unit 204 is used to determine the dynamic change result of the gas turbine blade tip clearance by means of a preset transient overall thermodynamic parameter time series, the target temperature field calculation model and the target blade tip clearance multi-field coupling calculation model.

[0062] In another embodiment, the data acquisition unit is equipped with a flow pipe in the compressor intake system of the gas turbine to obtain the compressor inlet air flow rate. ; An extraction system is installed at the compressor outlet of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine disk cooling gas flow rate. ; A flow meter is installed in the fuel gas system of the gas turbine to obtain the gas flow rate. ; An extraction system is installed between the compressor stages of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine ring cooling gas flow rate. ; Determine turbine inlet flow rate .

[0063] In another embodiment, the temperature field correction unit's comparison of the second monitoring data with the temperature field information, and the correction of the temperature field calculation model based on the comparison result, includes: The metal wall temperature of the target component in the second monitoring data is compared with the temperature field information. When the comparison result is greater than a preset temperature difference threshold, a Long Short-Term Memory (LSTM) neural network is used to correct the parameters of the temperature field calculation model. The temperature field information determined by the corrected temperature field calculation model is then compared with the metal wall temperature of the target component until the comparison result is less than the temperature difference threshold.

[0064] In another embodiment, the tip clearance model correction unit determines the estimated tip clearance value of the gas turbine using an initial tip clearance multi-field coupled calculation model determined by temperature field information, first monitoring data, and an elongation calculation model based on the target component. The blade elongation L1(t) is determined using a blade elongation calculation model. The elongation L2(t) of the turbine disk is determined by the calculation model of the turbine disk elongation. The rotor's shaft offset L3(t) is determined by the shaft position; The elongation L4(t) of the turbine retainer was determined using a calculation model for the elongation of the turbine retainer. The estimated gap value is determined by the formula L(t)=L0- L1(t)- L2(t)- L3(t)+ L4(t), where L0 is the installation gap.

[0065] In another embodiment, the comparison of the estimated clearance value with the actual measured clearance value in the tip clearance model correction unit, and the correction of the initial tip clearance multi-field coupling calculation model based on the comparison result, includes: When the comparison result is greater than the preset gap difference, the RBF-BPNN neural network is used to construct the inverse response surface model, and the gap value determined by the actual measurement is substituted into the inverse response surface model to correct the initial blade tip gap multi-field coupling calculation model. When the comparison result is less than the preset gap difference, the initial blade tip gap multi-field coupling calculation model is determined as the target blade tip gap multi-field coupling calculation model.

[0066] The above embodiments of the present invention provide a method for predicting dynamic clearance at the tip of a gas turbine blade, and based on the method, provide a device for predicting dynamic clearance at the tip of a gas turbine blade. The above method and device can reduce the probability of gas turbine accidents.

[0067] This embodiment also discloses a computer device, such as... Figure 3 As shown, the computer device includes a processor and a memory, the memory storing at least one instruction, which is loaded and executed by the processor to implement any of the above-described methods for predicting dynamic clearance at the tip of a gas turbine blade.

[0068] Furthermore, in the above-described embodiment of the gas turbine blade tip dynamic clearance prediction device, the logical division of each program module is merely illustrative. In practical applications, the above functions can be assigned to different program modules as needed, for example, for the sake of corresponding hardware configuration requirements or the convenience of software implementation. That is, the internal structure of the gas turbine blade tip dynamic clearance prediction device can be divided into different program modules to complete all or part of the functions described above.

[0069] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for predicting dynamic tip clearance of a gas turbine blade, characterized in that, The method includes: Acquire first monitoring data for model correction under actual operating conditions of the gas turbine, and second monitoring data for comparison with the results calculated by the model; The turbine inlet flow rate is determined using the first monitoring data. The heat transfer coefficient of the target component is then determined using the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data. The heat transfer coefficient and the temperature data in the first monitoring data are input into the temperature field calculation model to obtain the temperature field information of the target component. The second monitoring data is then compared with the temperature field information, and the temperature field calculation model is corrected based on the comparison results to obtain the target temperature field calculation model. The target component includes blades, turbine disks, and turbine retaining rings. The estimated clearance value of the gas turbine is determined by using the temperature field information, the first monitoring data, and the initial tip clearance multi-field coupling calculation model based on the elongation calculation model of the target component. The estimated clearance value is compared with the actual measured clearance value, and the initial tip clearance multi-field coupling calculation model is corrected based on the comparison results to obtain the target tip clearance multi-field coupling calculation model. The dynamic change of the gas turbine tip clearance is determined by using a preset transient overall thermodynamic parameter time series, the target temperature field calculation model, and the target tip clearance multi-field coupling calculation model.

2. The method according to claim 1, characterized in that, The first monitoring data includes compressor inlet flow rate, turbine disk cooling gas flow rate, gas flow rate, and turbine retaining ring cooling gas flow rate; Determining the turbine inlet flow rate using the first monitoring data includes: A flow pipe is installed in the compressor inlet system of the gas turbine to obtain the compressor inlet flow rate. ; An extraction system is installed at the compressor outlet of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine disk cooling gas flow rate. ; A flow meter is installed in the fuel gas system of the gas turbine to obtain the gas flow rate. ; An extraction system is installed between the compressor stages of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine ring cooling gas flow rate. ; Determine turbine inlet flow rate .

3. The method according to claim 1, characterized in that, The step of comparing the second monitoring data with the temperature field information and correcting the temperature field calculation model based on the comparison results includes: The metal wall temperature of the target component in the second monitoring data is compared with the temperature field information. When the comparison result is greater than the preset temperature difference threshold, the parameters of the temperature field calculation model are corrected using a long short-term memory neural network (LSTM). The temperature field information determined by the corrected temperature field calculation model is then compared with the metal wall temperature of the target component until the comparison result is less than the temperature difference threshold.

4. The method according to claim 1, characterized in that, The determination of the estimated clearance value of the gas turbine using the temperature field information, the first monitoring data, and the initial tip clearance multi-field coupled calculation model determined based on the elongation calculation model of the target component includes: The blade elongation L1(t) is determined using a blade elongation calculation model. The elongation L2(t) of the turbine disk is determined by the calculation model of the turbine disk elongation. The rotor's shaft offset L3(t) is determined by the shaft position; The elongation L4(t) of the turbine retainer was determined using a calculation model for the elongation of the turbine retainer. The estimated gap value is determined by the formula L(t)=L0- L1(t)- L2(t)- L3(t)+ L4(t), where L0 is the installation gap.

5. The method according to claim 1, characterized in that, The step of comparing the estimated clearance value with the actual measured clearance value and correcting the initial tip clearance multi-field coupling calculation model based on the comparison results includes: When the comparison result is greater than the preset gap difference, the RBF-BPNN neural network is used to construct the inverse response surface model, and the gap value determined by the actual measurement is substituted into the inverse response surface model to correct the initial blade tip gap multi-field coupling calculation model. When the comparison result is less than the preset gap difference, the initial blade tip gap multi-field coupling calculation model is determined as the target blade tip gap multi-field coupling calculation model.

6. A device for predicting dynamic tip clearance of a gas turbine blade, characterized in that, The device includes: The data acquisition unit is used to acquire first monitoring data for model correction and second monitoring data for comparison with the model calculation results under the actual operating conditions of the gas turbine. The temperature field model correction unit is used to determine the turbine inlet flow rate through the first monitoring data, then determine the heat transfer coefficient of the target component through the turbine inlet flow rate and the cooling gas flow rate in the first monitoring data, input the heat transfer coefficient and the temperature data in the first monitoring data into the temperature field calculation model to obtain the temperature field information of the target component, then compare the second monitoring data with the temperature field information, and correct the temperature field calculation model based on the comparison result to obtain the target temperature field calculation model, wherein the target component includes blades, turbine disk, and turbine retainer ring; The tip clearance model correction unit is used to determine the estimated clearance value of the gas turbine through the temperature field information, the first monitoring data and the initial tip clearance multi-field coupling calculation model determined based on the elongation calculation model of the target component, compare the estimated clearance value with the actual measured clearance value, and correct the initial tip clearance multi-field coupling calculation model based on the comparison result to obtain the target tip clearance multi-field coupling calculation model. The blade tip clearance result determination unit is used to determine the dynamic change result of the gas turbine blade tip clearance by means of a preset transient overall thermodynamic parameter time series, the target temperature field calculation model, and the target blade tip clearance multi-field coupling calculation model.

7. The apparatus according to claim 6, characterized in that, The data acquisition unit is equipped with a flow pipe in the compressor intake system of the gas turbine to obtain the compressor inlet flow rate. ; An extraction system is installed at the compressor outlet of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine disk cooling gas flow rate. ; A flow meter is installed in the fuel gas system of the gas turbine to obtain the gas flow rate. ; An extraction system is installed between the compressor stages of the gas turbine, and a flow pipe is installed within the extraction system to obtain the turbine ring cooling gas flow rate. ; Determine turbine inlet flow rate .

8. The apparatus according to claim 6, characterized in that, The temperature field model correction unit compares the second monitoring data with the temperature field information and corrects the temperature field calculation model based on the comparison result, including: The metal wall temperature of the target component in the second monitoring data is compared with the temperature field information. When the comparison result is greater than the preset temperature difference threshold, the parameters of the temperature field calculation model are corrected using a long short-term memory neural network (LSTM). The temperature field information determined by the corrected temperature field calculation model is then compared with the metal wall temperature of the target component until the comparison result is less than the temperature difference threshold.

9. The apparatus according to claim 6, characterized in that, The tip clearance model correction unit uses the initial tip clearance multi-field coupled calculation model determined by the temperature field information, the first monitoring data, and the elongation calculation model based on the target component to determine the estimated tip clearance value of the gas turbine, including: The blade elongation L1(t) is determined using a blade elongation calculation model. The elongation L2(t) of the turbine disk is determined by the calculation model of the turbine disk elongation. The rotor's shaft offset L3(t) is determined by the shaft position; The elongation L4(t) of the turbine retainer was determined using a calculation model for the elongation of the turbine retainer. The estimated gap value is determined by the formula L(t)=L0- L1(t)- L2(t)- L3(t)+ L4(t), where L0 is the installation gap.

10. The apparatus according to claim 6, characterized in that, The blade tip clearance model correction unit compares the estimated clearance value with the actual measured clearance value, and corrects the initial blade tip clearance multi-field coupling calculation model based on the comparison results, including: When the comparison result is greater than the preset gap difference, the RBF-BPNN neural network is used to construct the inverse response surface model, and the gap value determined by the actual measurement is substituted into the inverse response surface model to correct the initial blade tip gap multi-field coupling calculation model. When the comparison result is less than the preset gap difference, the initial blade tip gap multi-field coupling calculation model is determined as the target blade tip gap multi-field coupling calculation model.

Images