Quantitative detection method and system for blockage severity of exhaust system of heavy-duty gas turbine

Abstract

The invention relates to a quantitative detection method and system for the blockage severity of an exhaust system of a heavy duty gas turbine. The method comprises the following steps: acquiring real-time environmental condition parameters and control condition parameters of the exhaust system of the heavy duty gas turbine to be detected; inputting the environmental condition parameters and the control condition parameters into a pre-trained exhaust system pressure difference prediction model, and outputting a theoretical healthy pressure difference of the to-be-detected heavy duty gas turbine exhaust system; and based on the theoretical health pressure difference, designing health characteristic parameters of the exhaust system of the to-be-detected heavy duty gas turbine, and using the health characteristic parameters to represent quantitative indexes of the pipeline blockage condition to complete the quantitative detection process of the blockage severity. Compared with the prior art, the method has the advantages of quantifying the blockage severity of the exhaust system of the heavy-duty gas turbine under complex working conditions and the like.

Description

Technical Field

[0001] The present invention relates to the field of energy and power engineering technology, and in particular to a method and system for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine. Background Art

[0002] Heavy-duty gas turbines have been widely used in the power generation field due to their flexible start and shutdown, high thermal efficiency and low environmental impact. With the further tightening of energy conservation, emission reduction and environmental protection requirements, the proportion of clean fuels such as natural gas in primary energy has gradually increased, and the proportion of heavy-duty gas turbines in my country's power plants has become increasingly larger.

[0003] The exhaust system of a heavy-duty gas turbine is one of the important flow components in a steam-gas combined cycle power plant. When the flue gas in the exhaust system flows through the superheater, reheater, evaporator, economizer, etc. of the waste heat boiler, it will generate resistance loss along the way and local resistance loss. When the flue gas damper of the waste heat boiler is not fully opened or the exhaust section is blocked internally, the pressure loss of the exhaust system will increase, and the exhaust pressure of the heavy-duty gas turbine will increase, which will lead to a decrease in the thermal efficiency of the unit, increase the stress on the last stage blades of the turbine, affect the safe operation of the gas turbine, and even cause exhaust leakage. At present, the health monitoring of the exhaust system in a steam-gas combined cycle power plant only judges the blockage of the exhaust system pipeline through which the flue gas working medium flows based on the exhaust pressure signal of the heavy-duty gas turbine. For example, when the exhaust pressure of the gas turbine is greater than the limit value, it is judged that the flue gas damper of the waste heat boiler is not fully opened or the exhaust section is blocked internally.

[0004] However, from a physical point of view, the pressure loss (i.e., pressure difference) of the exhaust system is not only related to the blockage of the exhaust system pipeline, but also to the mass flow rate, specific volume, etc. of the flue gas medium. When the mass flow rate, specific volume, and other parameters of the flue gas medium flowing through the exhaust system pipeline change due to changes in the unit's operating conditions, the pressure loss of the exhaust system will also change accordingly, resulting in changes in the corresponding exhaust pressure. When the exhaust system pipeline is blocked to varying degrees, the exhaust pressure signal of the heavy-duty gas turbine fluctuates and overlaps greatly at different operating points under different conditions. Therefore, the existing technology is too one-sided to judge the blockage of the exhaust system pipeline only from the exhaust pressure signal, which can easily lead to misleading diagnostic results. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for quantitatively detecting the severity of blockage in the exhaust system of a heavy-duty gas turbine, which can quantify the pipeline blockage under complex and variable operating conditions.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A method for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine comprises the following steps:

[0008] Acquire the real-time operation data of the exhaust system of the heavy-duty gas turbine to be tested, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0009] Based on the real-time operation data, a pre-trained exhaust system pressure difference prediction model is used for processing to output a theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0010] Based on the theoretical healthy pressure difference and the measured pressure difference, the healthy characteristic parameters of the exhaust system of the heavy-duty gas turbine to be tested are designed, and the healthy characteristic parameters are used as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

[0011] Furthermore, the real-time operating data includes environmental condition parameters, control condition parameters and other measurable gas circuit parameters, the environmental condition parameters include ambient pressure and ambient temperature, the control condition parameters include compressor inlet guide vane position feedback IGV, gas turbine speed, and gas turbine power, and the measurable gas circuit parameters include gas turbine inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

[0012] Furthermore, the input of the exhaust system pressure difference prediction model is the environmental condition parameters and control condition parameters in the real-time operation data.

[0013] Furthermore, the calculation expression of the health characteristic parameter is:

[0014]

[0015] In the formula, HP EX is the health characteristic parameter, ΔP EX,act is the measured pressure difference, ΔP EX,pre is the theoretical healthy pressure difference.

[0016] Furthermore, the training step of the exhaust system pressure difference prediction model includes:

[0017] Obtain historical gas path measurable parameters when the exhaust system of the heavy-duty gas turbine is operating normally and healthily, wherein the historical gas path measurable parameters include historical environmental condition parameters, historical control condition parameters and historical remaining gas path measurable parameters, wherein the historical environmental condition parameters include ambient pressure, ambient temperature and relative humidity, the historical control condition parameters include speed, gas turbine power, fuel temperature, compressor inlet guide vane position feedback IGV, and the historical remaining gas path measurable parameters include compressor inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate;

[0018] Constructing a deep neural network, wherein the deep neural network includes one of a fully connected neural network and a convolutional neural network;

[0019] The historical gas path measurable parameters are input into the deep neural network for training until the training is completed, and an exhaust system pressure difference prediction model is obtained, wherein the historical operating state of the heavy-duty gas turbine exhaust system is determined by historical environmental condition parameters and historical control condition parameters. During the training process, the historical actual pressure difference calculated according to the historical gas path measurable parameters is used as a prediction target, so that the exhaust system pressure difference prediction model learns the mapping relationship between the historical environmental condition parameters, the historical control condition parameters and the prediction target.

[0020] The present invention also provides a heavy-duty gas turbine exhaust system blockage severity quantification detection system, comprising:

[0021] Measured pressure difference calculation module: used to obtain the real-time operation data of the exhaust system of the heavy-duty gas turbine to be tested, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0022] Theoretical healthy pressure difference prediction module: used for processing the real-time operation data using a pre-trained exhaust system pressure difference prediction model, and outputting the theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0023] Quantitative detection module: used to design the health characteristic parameters of the exhaust system of the heavy-duty gas turbine to be detected based on the theoretical healthy pressure difference and the measured pressure difference, and use the health characteristic parameters as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

[0024] Furthermore, the real-time operating data includes environmental condition parameters, control condition parameters and other measurable gas circuit parameters, the environmental condition parameters include ambient pressure and ambient temperature, the control condition parameters include compressor inlet guide vane position feedback IGV, gas turbine speed, and gas turbine power, and the measurable gas circuit parameters include gas turbine inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

[0025] Furthermore, the input of the exhaust system pressure difference prediction model is the environmental condition parameters and control condition parameters in the real-time operation data.

[0026] Furthermore, the calculation expression of the health characteristic parameter is:

[0027]

[0028] In the formula, HP EX is the health characteristic parameter, ΔP EX,actis the measured pressure difference, ΔP EX,pre is the theoretical healthy pressure difference.

[0029] Furthermore, the training step of the exhaust system pressure difference prediction model includes:

[0030] Obtain historical gas path measurable parameters when the exhaust system of the heavy-duty gas turbine is operating normally and healthily, wherein the historical gas path measurable parameters include historical environmental condition parameters, historical control condition parameters and historical remaining gas path measurable parameters, wherein the historical environmental condition parameters include ambient pressure, ambient temperature and relative humidity, the historical control condition parameters include speed, gas turbine power, fuel temperature, compressor inlet guide vane position feedback IGV, and the historical remaining gas path measurable parameters include compressor inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate;

[0031] Constructing a deep neural network, wherein the deep neural network includes one of a fully connected neural network and a convolutional neural network;

[0032] The historical gas path measurable parameters are input into the deep neural network for training until the training is completed, and an exhaust system pressure difference prediction model is obtained, wherein the historical operating state of the heavy-duty gas turbine exhaust system is determined by historical environmental condition parameters and historical control condition parameters. During the training process, the historical actual pressure difference calculated according to the historical gas path measurable parameters is used as a prediction target, so that the exhaust system pressure difference prediction model learns the mapping relationship between the historical environmental condition parameters, the historical control condition parameters and the prediction target.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] (1) For an actual healthy heavy-duty gas turbine, the exhaust system of the gas turbine can be determined solely by environmental conditions and control conditions. Under complex and variable operating conditions, the present invention only needs to collect environmental condition parameters and control condition parameters. The exhaust system pressure difference prediction model predicts the theoretical healthy pressure difference, and then by designing healthy characteristic parameters that characterize the pipeline blockage condition, the pipeline blockage condition under complex and variable operating conditions can be quantified.

[0035] (2) The present invention designs an exhaust system health characteristic parameter as a quantitative indicator to characterize the exhaust system pipe blockage condition. The dimensionless health characteristic parameter reflects the physical essential changes in the exhaust system pipe blockage condition by dividing the measured pressure difference by the theoretical pressure difference and adding a bias value of 1. It can maintain good robustness and accuracy at different operating points under different variable operating conditions.

[0036] (3) The present invention can accurately quantify the pipe blockage condition of the exhaust system of a heavy-duty gas turbine under complex and variable operating conditions, providing a scientific basis and technical support for formulating appropriate and reasonable maintenance strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0038] Figure 2 A flow chart of establishing an exhaust system pressure difference prediction model according to the present invention;

[0039] Figure 3 It is a schematic diagram of the overall implementation process of the present invention;

[0040] Figure 4 The atmospheric temperature change of the embodiment of the present invention;

[0041] Figure 5 The IGV opening degree change of the embodiment of the present invention;

[0042] Figure 6 The fuel mass flow rate change of the embodiment of the present invention;

[0043] Figure 7 The diagnosis result obtained by the traditional method of the embodiment of the present invention;

[0044] Figure 8 The health characteristic parameters obtained by diagnosis when the exhaust system pipeline is blocked to different degrees in the embodiment of the present invention;

[0045] Fig. 9 is the relative percentage change of the measurable parameters of the air path when the intake system fails in the embodiment of the present invention;

[0046] Fig.10 is the relative percentage change of the measurable parameters of the gas circuit when the compressor fails in the embodiment of the present invention;

[0047] Fig.11 is the relative percentage change of the measurable parameters of the gas path when the turbine fails in the embodiment of the present invention;

[0048] Fig.12 It is the relative percentage change of the measurable parameters of the gas path when the combustion chamber fails in the embodiment of the present invention. DETAILED DESCRIPTION

[0049] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0050] Example 1

[0051] This embodiment provides a method for quantitatively detecting the severity of blockage in the exhaust system of a heavy-duty gas turbine. Figure 1 As shown, the method comprises the following steps:

[0052] S1. Collect real-time operating data of the exhaust system of the gas turbine to be diagnosed, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be detected.

[0053] The real-time operation data includes environmental condition parameters, control condition parameters and other measurable gas circuit parameters. The environmental condition parameters include ambient pressure and ambient temperature. The control condition parameters include compressor inlet guide vane position feedback IGV, gas turbine speed, and gas turbine power. The measurable gas circuit parameters include gas turbine inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

[0054] Based on the operating data, the measured pressure difference is calculated according to the existing calculation method.

[0055] S2. Based on the real-time operation data, a pre-trained exhaust system pressure difference prediction model is used for processing to output a theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested.

[0056] First, it is necessary to collect the historical gas path measurable parameters when the exhaust system of the gas turbine to be diagnosed is running normally and healthily, and establish the exhaust system pressure difference prediction model through deep neural network. The historical gas path measurable parameters include environmental condition parameters, control condition parameters and other gas path measurable parameters, among which the other gas path measurable parameters include compressor inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

[0057] For an actual healthy heavy-duty gas turbine, its operating performance status (including the exhaust system pressure difference) can be determined solely by the ambient conditions (including the ambient pressure P 0 , ambient temperature t 0 , relative humidity RH) and control conditions (including compressor inlet guide vane position feedback IGV, fuel temperature t f , gas turbine speed n, gas turbine power N e ) determines that the historical actual pressure difference calculated according to the historical gas path measurable parameters is used as the prediction target, so that the exhaust system pressure difference prediction model learns the mapping relationship between the historical environmental condition parameters, the historical control condition parameters and the prediction target, and obtains a trained exhaust system pressure difference prediction model.

[0058] Through the sensitivity analysis of various environmental condition parameters and control condition parameters on the accuracy of exhaust system pressure difference prediction, the final gas path measurable parameters input by the deep neural network are selected including environmental conditions (including environmental pressure P 0 , ambient temperature t 0 ) and control conditions (including IGV position feedback IGV, engine speed n, engine power N e ), the measurable gas path parameter output by the deep neural network is the exhaust system pressure difference ΔP EX (i.e. turbine exhaust pressure and ambient pressure P 0 Since the input of the deep neural network model is only 5 parameters and the output is only 1 parameter, a fully connected neural network or a common convolutional neural network can achieve good training and prediction effects. The present invention adopts a one-dimensional convolutional neural network, and the structure of the deep convolutional neural network model is shown in Table 1 below.

[0059] The exhaust system pressure difference prediction model is used to calculate the theoretical pressure difference ΔP of the exhaust system of the actual gas turbine unit under complex variable operating conditions in real time. EX,pre ,like Figure 3 The exhaust system pressure difference prediction model is used to calculate the theoretical healthy pressure difference ΔP of the exhaust system of the actual gas turbine unit under complex variable operating conditions in real time. EX,pre .

[0060] Table 1 Deep convolutional neural network model structure

[0061]

[0062] S3. Based on the theoretical healthy pressure difference and the measured pressure difference, the healthy characteristic parameters of the exhaust system of the heavy-duty gas turbine to be tested are designed, and the healthy characteristic parameters are used as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

[0063] The health characteristic parameter HP EX Equal to the measured pressure difference ΔP of the exhaust system EX,act The difference between the theoretical healthy pressure and the EX,pre The difference is then divided by the theoretical healthy pressure difference ΔP EX,pre , plus a bias value of 1, that is:

[0064]

[0065] The dimensionless health characteristic parameters obtained by the diagnosis reflect the physical nature of the changes in the exhaust system pipe blockage condition, and can maintain good robustness and accuracy at different operating points. EXWhen the value is 1, it indicates that the exhaust system is in a baseline healthy state. When the value is greater than 1, it indicates that there is a blockage in the exhaust system pipes, and the larger the value, the more serious it is.

[0066] Table 2 Environmental and control parameter conditions and other measurable parameters of the gas path

[0067]

[0068] In order to test the effectiveness of the detection method proposed in the present invention, the AE94.3A F-class heavy-duty gas turbine produced by Ansaldo was selected as the diagnostic object. This AE94.3A heavy-duty gas turbine has been widely used in domestic power plants. Its compressor is a 15-stage large-flow, axial-flow compressor equipped with 2-stage inlet adjustable guide vanes. The combustion chamber is an annular combustion chamber equipped with 24 low-NOx burners. The detailed gas path measurable parameters of this heavy-duty gas turbine are shown in Table 2.

[0069] like Figure 2 As shown in the figure, firstly, the historical gas path measurable parameters of the gas turbine exhaust system under normal and healthy operation are collected to establish the exhaust system pressure difference prediction model through deep neural network. For an actual healthy heavy-duty gas turbine, its operating performance status (including exhaust system pressure difference) can be uniquely determined by the environmental conditions (including the ambient pressure P 0 , ambient temperature t 0 , relative humidity RH) and control conditions (including compressor inlet guide vane position feedback IGV, fuel temperature t f , gas turbine speed n, gas turbine power N e ) is determined by analyzing the sensitivity of various environmental condition parameters and control condition parameters to the prediction accuracy of the exhaust system pressure difference. Finally, the gas path measurable parameters input by the deep neural network are selected including environmental conditions (including environmental pressure P 0 , ambient temperature t 0 ) and control conditions (including IGV position feedback IGV, engine speed n, engine power N e ), the measurable gas path parameter output by the deep neural network is the exhaust system pressure difference ΔP EX (i.e. turbine exhaust pressure and ambient pressure P 0 The exhaust system pressure difference prediction model is used to calculate the theoretical pressure difference ΔP of the exhaust system of the actual gas turbine unit under complex variable operating conditions in real time. EX,pre The exhaust system pressure difference prediction model is used to calculate the theoretical healthy pressure difference ΔP of the exhaust system of the actual gas turbine unit under complex variable operating conditions in real time. EX,pre .

[0070] The exhaust system health characteristic parameter is then designed as a quantitative indicator to characterize the exhaust system pipeline blockage condition. EXEqual to the measured pressure difference ΔP of the exhaust system EX,act The difference between the theoretical healthy pressure and the EX,pre The difference is then divided by the theoretical healthy pressure difference ΔP EX,pre , plus a bias value of 1, that is The dimensionless health characteristic parameters obtained by the diagnosis reflect the physical nature of the changes in the exhaust system pipe blockage condition, and can maintain good robustness and accuracy at different operating points. EX When the value is 1, it indicates that the exhaust system is in a baseline healthy state. When the value is greater than 1, it indicates that there is a blockage in the exhaust system pipes, and the larger the value, the more serious it is.

[0071] In order to test the effectiveness of the proposed method in the present invention in accurately quantifying the blockage of the gas turbine exhaust system pipeline under complex variable operating conditions (from the minimum opening load of the IGV after grid connection to the basic load of the IGV fully opened), the heavy-duty gas turbine unit was selected under complex variable operating conditions (such as Figures 4 to 6 Diagnostic test is performed on samples of measurable gas path parameters when the exhaust system pipeline has blockage conditions of different severity (mild, medium, severe) as shown in Table 3. Blockage conditions of different severity of the exhaust system pipeline are shown in Table 3.

[0072] Table 3 Description of blockage conditions of different severity in exhaust system pipes

[0073]

[0074] Depend on Figure 7 As shown, since the pressure loss of the exhaust system is not only related to the blockage of the exhaust system pipeline, but also to the mass flow rate, specific volume and other variable operating conditions of the flue gas, when the mass flow rate, specific volume and other parameters of the flue gas flowing through the exhaust system pipeline change due to the change of the unit operating conditions, the pressure loss of the exhaust system will also change accordingly, resulting in a corresponding change in the exhaust pressure. When the exhaust system pipeline is blocked to different degrees of severity, the exhaust pressure signal of the heavy-duty gas turbine fluctuates and overlaps greatly at different variable operating points. Therefore, the existing technology is too one-sided to judge the blockage of the exhaust system pipeline through which the flue gas flows only from the exhaust pressure signal, which is easy to produce misleading diagnostic results.

[0075] The results of the diagnosis obtained by the method proposed by the present invention are as follows Figure 8 As shown. Figure 8It can be seen that the quantitative numerical results diagnosed by the method proposed in the present invention are basically consistent with the blockage conditions of different severity (mild, medium, severe) occurring in the exhaust system pipelines of the heavy-duty gas turbine unit under complex variable operating conditions, and the health characteristic parameters of the exhaust system diagnosed maintain good robustness and accuracy at different variable operating points, and are not affected by changes in the operating conditions of the unit, which illustrates the higher diagnostic accuracy and anti-interference ability of the method proposed in the present invention.

[0076] In addition, the anti-interference performance of the method proposed in the present invention on the influence of other flow-through components failure of heavy-duty gas turbine units is analyzed. When the intake system, compressor, combustion chamber, and turbine fail to meet the requirements of different severity, the relative percentage change of the measurable parameters of the gas path (compared with the measurable parameters of the gas path when the unit is healthy) can be obtained. Figures 9 to 12 shown.

[0077] Depend on Fig. 9 It can be seen that when the intake system has faults of different severity, the compressor inlet pressure loss increases, and the compressor inlet pressure P 1 The gas turbine output power N e , the turbine outlet flue gas mass flow rate and the gas turbine thermal efficiency are reduced, while the turbine outlet temperature t 5 However, the exhaust pressure signal P of the heavy-duty gas turbine 5 Almost unchanged, that is, the measured pressure difference ΔP of the exhaust system EX,act (Turbine exhaust pressure P 5 and ambient pressure P 0 The difference) is almost unchanged, so the dimensionless health characteristic parameter diagnosed by the present invention is There is no change either, which illustrates the anti-interference ability of the method proposed in the present invention on the influence of the failure of the air intake system of the heavy-duty gas turbine unit.

[0078] Depend on Fig.10 It can be seen that when the compressor fails to a certain degree, the compressor flow capacity is reduced, the compressor pressure ratio is reduced, and the compressor outlet pressure P 2 As the compressor pressure ratio decreases, the turbine expansion ratio decreases, and the turbine outlet temperature t 5 However, the exhaust pressure signal P of the heavy-duty gas turbine 5 Almost unchanged, that is, the measured pressure difference ΔP of the exhaust system EX,act (Turbine exhaust pressure P 5 and ambient pressure P 0 The difference) is almost unchanged, so the dimensionless health characteristic parameter diagnosed by the present invention is There is no change either, which illustrates the anti-interference ability of the method proposed in the present invention on the influence of compressor failure of heavy-duty gas turbine units.

[0079] Depend on Fig.11 It can be seen that when the turbine fails to a certain degree, the flow capacity of the turbine increases and the expansion ratio of the turbine decreases, resulting in a decrease in the compressor outlet pressure P 2 As the expansion ratio of the turbine decreases, the compressor pressure ratio decreases, and the turbine outlet temperature t 5 However, the exhaust pressure signal P of the heavy-duty gas turbine 5 Almost unchanged, that is, the measured pressure difference ΔP of the exhaust system EX,act (Turbine exhaust pressure P 5 and ambient pressure P 0 The difference) is almost unchanged, so the dimensionless health characteristic parameter diagnosed by the present invention is There is no change either, which illustrates the anti-interference ability of the method proposed in the present invention on the influence of turbine failure of heavy-duty gas turbine units.

[0080] Depend on Fig.12 It can be seen that when the combustion chamber fails to a certain degree, the fuel mass flow rate remains unchanged, the combustion efficiency decreases, and the turbine outlet temperature t 5 Reduce, gas turbine output power N e However, the exhaust pressure signal P of the heavy-duty gas turbine 5 Almost unchanged, that is, the measured pressure difference ΔP of the exhaust system EX,act (Turbine exhaust pressure P 5 and ambient pressure P 0 The difference) is almost unchanged, so the dimensionless health characteristic parameter diagnosed by the present invention is There is no change either, which illustrates the anti-interference ability of the method proposed in the present invention to the influence of combustion chamber failure of heavy-duty gas turbine units.

[0081] In summary, when other flow components (intake system, compressor, combustion chamber, turbine) of the heavy-duty gas turbine unit fail to meet the requirements of different severity, the dimensionless health characteristic parameters diagnosed by the method proposed in the present invention are Almost no change occurs, with excellent anti-interference performance.

[0082] Finally, the sensitivity analysis of each gas path measurable parameter on the diagnostic accuracy is carried out. For an actual healthy heavy-duty gas turbine, its operating performance status (including the exhaust system pressure difference) can be uniquely determined by the environmental conditions (including the ambient pressure P 0 , ambient temperature t 0 , relative humidity RH) and control conditions (including compressor inlet guide vane position feedback IGV, fuel temperature t f , gas turbine speed n, gas turbine power N e) is determined, so theoretically, the selection of measurable gas path parameters input by the neural network of the present invention may include environmental conditions (including environmental pressure P 0 , ambient temperature t 0 , relative humidity RH) and control conditions (including IGV position feedback IGV, fuel temperature t f , gas turbine speed n, gas turbine power N e ), the measurable gas path parameter output by the neural network is the exhaust system pressure difference ΔP EX,pre (Turbine exhaust pressure P 5 and ambient pressure P 0 The exhaust system pressure difference prediction model is used to calculate the theoretical healthy pressure difference ΔP of the exhaust system of the actual gas turbine unit under complex variable operating conditions in real time. IN,pre Next, we further analyzed the sensitivity of each input parameter of the selected neural network to the diagnostic accuracy, as shown in Table 4.

[0083] It can be seen from Table 4 that when the heavy-duty gas turbine is running under variable conditions, among the various input parameters of the selected deep neural network, the compressor inlet guide vane position feedback IGV has the greatest sensitivity to the diagnostic accuracy, followed by the gas turbine power N e , followed by the ambient temperature t 0 , and again the ambient pressure P 0 and engine speed n, while relative humidity RH, and fuel temperature t f The sensitivity to diagnostic accuracy is relatively small (the maximum error percentage does not exceed 0.3%), so the neural network input parameters for modeling the exhaust system pressure difference prediction model are selected to include environmental conditions (including ambient pressure P 0 , ambient temperature t 0 ) and control conditions (including IGV position feedback IGV, engine speed n, engine power N e )These 5 parameters.

[0084] Table 4 Sensitivity analysis of the selected neural network input parameters on diagnostic accuracy

[0085]

[0086] In summary, in order to effectively extract the healthy characteristic parameters of the exhaust system that are not affected by changes in the operating conditions of the unit, the present invention first collects the historical gas path measurable parameters when the exhaust system of the gas turbine to be diagnosed is healthy, and establishes an exhaust system pressure difference prediction model through a neural network. Then, the exhaust system pressure difference prediction model is used to calculate in real time the theoretical pressure difference ΔP of the exhaust system of the actual gas turbine unit when it is operating under complex variable conditions. EX,pre Finally, the exhaust system health characteristic parameter is designed as a quantitative indicator to characterize the exhaust system pipeline blockage condition. The dimensionless health characteristic parameter HP EXEqual to the measured pressure difference ΔP of the exhaust system EX,act The theoretical pressure difference ΔP EX,pre The difference is divided by the theoretical pressure difference ΔP EX,pre , plus a bias value of 1, that is The dimensionless health characteristic parameters obtained by this diagnosis reflect the physical essential changes in the blockage condition of the exhaust system pipes, and can maintain good robustness and accuracy at different operating points.

[0087] Example 2

[0088] This embodiment provides a system for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine, comprising:

[0089] Measured pressure difference calculation module: used to obtain the real-time operation data of the exhaust system of the heavy-duty gas turbine to be tested, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0090] Theoretical healthy pressure difference prediction module: used for processing the real-time operation data using a pre-trained exhaust system pressure difference prediction model, and outputting the theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested;

[0091] Quantitative detection module: used to design the health characteristic parameters of the exhaust system of the heavy-duty gas turbine to be detected based on the theoretical healthy pressure difference and the measured pressure difference, and use the health characteristic parameters as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

[0092] The rest is the same as in Example 1.

[0093] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0094] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for quantitatively detecting the severity of blockage in a heavy-duty gas turbine exhaust system, characterized in that: The following steps are involved: Acquire the real-time operation data of the exhaust system of the heavy-duty gas turbine to be tested, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be tested; Based on the real-time operation data, a pre-trained exhaust system pressure difference prediction model is used for processing to output a theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested; Based on the theoretical healthy pressure difference and the measured pressure difference, the healthy characteristic parameters of the exhaust system of the heavy-duty gas turbine to be tested are designed, and the healthy characteristic parameters are used as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

2. A method for quantitatively detecting the severity of blockage in a heavy-duty gas turbine exhaust system according to claim 1, characterized in that: The real-time operating data includes environmental condition parameters, control condition parameters and other measurable gas circuit parameters. The environmental condition parameters include ambient pressure and ambient temperature. The control condition parameters include compressor inlet guide vane position feedback IGV, gas turbine speed, and gas turbine power. The measurable gas circuit parameters include gas turbine inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

3. A method for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine according to claim 2, characterized in that: The input of the exhaust system pressure difference prediction model is the environmental condition parameters and control condition parameters in the real-time operation data.

4. A method for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine according to claim 1, characterized in that: The calculation expression of the health characteristic parameter is: In the formula, HP EX is the health characteristic parameter, ΔP EX,act is the measured pressure difference, ΔP EX,pre is the theoretical healthy pressure difference.

5. A method for quantitatively detecting the severity of blockage in an exhaust system of a heavy-duty gas turbine according to claim 1, characterized in that: The training steps of the exhaust system pressure difference prediction model include: Obtain historical gas path measurable parameters when the exhaust system of the heavy-duty gas turbine is operating normally and healthily, wherein the historical gas path measurable parameters include historical environmental condition parameters, historical control condition parameters and historical remaining gas path measurable parameters, wherein the historical environmental condition parameters include ambient pressure, ambient temperature and relative humidity, the historical control condition parameters include speed, gas turbine power, fuel temperature, compressor inlet guide vane position feedback IGV, and the historical remaining gas path measurable parameters include compressor inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate; Constructing a deep neural network, wherein the deep neural network includes one of a fully connected neural network and a convolutional neural network; The historical gas path measurable parameters are input into the deep neural network for training until the training is completed, and an exhaust system pressure difference prediction model is obtained, wherein the historical operating state of the heavy-duty gas turbine exhaust system is determined by historical environmental condition parameters and historical control condition parameters. During the training process, the historical actual pressure difference calculated according to the historical gas path measurable parameters is used as a prediction target, so that the exhaust system pressure difference prediction model learns the mapping relationship between the historical environmental condition parameters, the historical control condition parameters and the prediction target.

6. A heavy-duty gas turbine exhaust system blockage severity quantitative detection system, characterized in that: include: Measured pressure difference calculation module: used to obtain the real-time operation data of the exhaust system of the heavy-duty gas turbine to be tested, and calculate the measured pressure difference of the exhaust system of the heavy-duty gas turbine to be tested; Theoretical healthy pressure difference prediction module: used for processing the real-time operation data using a pre-trained exhaust system pressure difference prediction model, and outputting the theoretical healthy pressure difference of the exhaust system of the heavy-duty gas turbine to be tested; Quantitative detection module: used to design the health characteristic parameters of the exhaust system of the heavy-duty gas turbine to be detected based on the theoretical healthy pressure difference and the measured pressure difference, and use the health characteristic parameters as quantitative indicators to characterize the pipeline blockage status, thereby completing the blockage severity quantitative detection process.

7. A heavy-duty gas turbine exhaust system blockage severity quantification detection system according to claim 6, characterized in that: The real-time operating data includes environmental condition parameters, control condition parameters and other measurable gas circuit parameters. The environmental condition parameters include ambient pressure and ambient temperature. The control condition parameters include compressor inlet guide vane position feedback IGV, gas turbine speed, and gas turbine power. The measurable gas circuit parameters include gas turbine inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate.

8. A heavy-duty gas turbine exhaust system blockage severity quantification detection system according to claim 7, characterized in that: The input of the exhaust system pressure difference prediction model is the environmental condition parameters and control condition parameters in the real-time operation data.

9. A heavy-duty gas turbine exhaust system blockage severity quantification detection system according to claim 6, characterized in that: The calculation expression of the health characteristic parameter is: In the formula, HP EX is the health characteristic parameter, ΔP EX,act is the measured pressure difference, ΔP EX,pre is the theoretical healthy pressure difference.

10. A heavy-duty gas turbine exhaust system blockage severity quantification detection system according to claim 6, characterized in that: The training steps of the exhaust system pressure difference prediction model include: Obtain historical gas path measurable parameters when the exhaust system of the heavy-duty gas turbine is operating normally and healthily, wherein the historical gas path measurable parameters include historical environmental condition parameters, historical control condition parameters and historical remaining gas path measurable parameters, wherein the historical environmental condition parameters include ambient pressure, ambient temperature and relative humidity, the historical control condition parameters include speed, gas turbine power, fuel temperature, compressor inlet guide vane position feedback IGV, and the historical remaining gas path measurable parameters include compressor inlet pressure, compressor inlet temperature, compressor outlet pressure, compressor outlet temperature, turbine exhaust pressure, turbine outlet temperature and fuel mass flow rate; Constructing a deep neural network, wherein the deep neural network includes one of a fully connected neural network and a convolutional neural network; The historical gas path measurable parameters are input into the deep neural network for training until the training is completed, and an exhaust system pressure difference prediction model is obtained, wherein the historical operating state of the heavy-duty gas turbine exhaust system is determined by historical environmental condition parameters and historical control condition parameters. During the training process, the historical actual pressure difference calculated according to the historical gas path measurable parameters is used as a prediction target, so that the exhaust system pressure difference prediction model learns the mapping relationship between the historical environmental condition parameters, the historical control condition parameters and the prediction target.

Images