Combustion mode switching judgment method, switching method and device of gas turbine

Abstract

The invention discloses a combustion mode switching judgment method, switching method and device for a gas turbine, and the method comprises the steps: collecting an operation data set and a hydrogen doping proportion set of the gas turbine, and constructing a mixed model according to the operation data set and the hydrogen doping proportion; inputting the real-time operation data of the gas turbine into the hybrid model, and judging whether to switch a combustion mode or not according to the load adjustment direction of the gas turbine to obtain a judgment result; according to the judgment result, the combustion modes of the gas turbine are switched, the combustion mode switching model is calculated in real time, the combustion mode switching is accurately controlled, the problem that a fixed limiting value cannot reflect the optimal combustion mode switching opportunity in the actual state is solved, and the combustion mode switching efficiency is improved. The combustion mode switching can adapt to the gas turbine under various operation conditions, and the advancement and the response speed of the control system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrogen - mixed gas turbines, and particularly to a method for judging combustion mode switching, a switching method and a device for a gas turbine. Background Art

[0002] In order to reduce NOX emissions and meet national environmental protection requirements, gas turbines usually adopt dry low - NOX combustion technology, and reduce the generation of NOX through premixed combustion. However, the premixed combustion flame is relatively unstable. If the combustion mode is not switched at the appropriate switching point, flameout or high combustion pulsation may easily occur, damaging the burner, and causing load fluctuations of the unit, affecting the stability of the power grid and endangering the safe operation of the unit.

[0003] Based on using natural gas as fuel, a hydrogen - mixed gas turbine can operate with a certain proportion of hydrogen doping or pure hydrogen fuel, and its combustion characteristics have also changed, making it more prone to thermo - acoustic oscillations, and the timing of combustion mode switching is more important.

[0004] Since the combustion temperature cannot be directly measured for a long time, currently, the combustion mode switching of gas turbines usually takes the turbine exhaust temperature or the gas turbine load reaching a fixed limit as the switching condition. However, influencing factors such as ambient temperature changes, unit operating state changes, and fuel component changes are not considered, and the switching condition of the fixed limit cannot accurately represent the optimal switching timing under the actual combustion situation.

[0005] The prior art has the following problems: (1) When the ambient temperature or the unit operating state changes, when the gas turbine reaches the same exhaust temperature or load, the combustion state is not the same. Taking a fixed limit as the combustion mode switching condition may lead to unstable combustion and even affect the safe operation of the unit;

[0006] (2) The introduction of hydrogen makes the flame propagation speed faster and the ignition delay shorter, making it more prone to thermo - acoustic oscillations. Switching combustion with a fixed limit cannot match the influence of the change in hydrogen doping ratio on the combustion state.

[0007] In the patent document CN118934271A, a control method and device for switching the combustion mode of a gas turbine are disclosed. By online analysis to determine the combustion chamber stability in the current stage of the gas turbine, the steady-state control parameters under different combustion modes and the transient control parameters during the combustion mode switching process are determined, including collecting data corresponding to each combustion mode during the combustion adjustment process and the operation process of the combustion chamber, and storing the data in a time series format; selecting the pressure time series that meet the requirements under each combustion mode of the combustion chamber, calculating the sample entropy of the data corresponding to the pressure time series, and constructing a steady-state combustion stability characterization parameter based on the sample entropy under each combustion model; selecting the pressure time series that meet the requirements under each combustion mode of the combustion chamber, calculating the damping ratio of the data corresponding to the pressure time series; selecting the minimum damping ratio that meets the switching stability requirements during the switching process of each combustion mode, and determining the transient control parameters for combustion mode switching based on the minimum damping ratio; calculating the real-time sample entropy and real-time damping ratio corresponding to the pressure data under the current combustion mode of the combustion chamber in real-time online; comparing the real-time sample entropy with the stability characterization parameter to determine the fuel ratio of the current valve; comparing the real-time damping ratio with the switching transient control parameters to determine whether to switch the combustion mode, without solving the problem of comprehensively making decisions on the switching of the combustion mode in combination with the operating state of the gas turbine and the hydrogen addition condition.

[0008] In the patent document CN109654534B, a combustion adjustment method for a DLN1.0 combustion system is disclosed. The fuel ratio of the primary nozzle and the secondary nozzle of a GE9E gas turbine using a DLN1.0 burner is adjusted in the premixed stable stage to make it adapt to the influence brought by the change of ambient temperature and the Wobbe index of natural gas, optimize emissions, and improve combustion stability; after each parameter of the gas turbine changes and stabilizes, record the key data of the gas turbine, including time, gas turbine load, combustion reference temperature, instantaneous value of fuel ratio, ambient temperature, compressor pressure ratio, compressor exhaust temperature, turbine exhaust temperature, exhaust dispersion, natural gas inlet temperature, IGV opening, IBH opening, NOx emissions, CO emissions, and O2 content parameters. Draw the change relationship curves of the fuel ratio at the operating point with NOx and CO, and conduct data analysis and summary to determine the value of the FXKSPM array; if there are non-smooth points in the curve, repeat the measurement to ensure data accuracy; according to the currently adjusted combustion reference temperature, use the linear interpolation method to obtain the relationship between FXKTPM and TTRF1, without solving the problem of comprehensively making decisions on the switching of the combustion mode in combination with the operating state of the gas turbine and the hydrogen addition condition.

[0009] In summary, neither of the above two existing patents solves the problem of comprehensively making decisions on the switching of the combustion mode in combination with the operating state of the gas turbine and the hydrogen addition condition. Summary of the Invention

[0010] Based on the above technical problems, the present invention proposes a combustion mode switching judgment method, a switching method and a device for a gas turbine, to solve the problem of comprehensively making decisions on the switching of the combustion mode in combination with the operating state of the gas turbine and the hydrogen addition conditions.

[0011] To achieve the above object, the present invention proposes a combustion mode switching judgment method for a gas turbine.

[0012] A combustion mode switching judgment method for a gas turbine includes:

[0013] Collect the operating data set and the hydrogen addition ratio set of the gas turbine, and construct a hybrid model according to the operating data set and the hydrogen addition ratio;

[0014] Input the real-time operating data of the gas turbine into the hybrid model, and respectively judge whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine to obtain a judgment result.

[0015] Further, establishing the hybrid model according to the operating data set includes:

[0016] According to the operating data set, train the combustion mode switching point model, and then construct the hybrid model through linear regression according to the combustion mode switching point model and the hydrogen addition ratio set.

[0017] Further, the operating data set includes:

[0018] Compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel pressure, gas turbine speed and / or gas turbine load.

[0019] Further, the training to obtain the combustion mode switching point model according to the operating data set includes:

[0020] Randomly select 70% of the data in the operating data set as the training set and 30% of the data as the verification set to obtain the combustion mode switching point model.

[0021] Further, the constructing the hybrid model through linear regression according to the combustion mode switching point model and the hydrogen addition ratio includes:

[0022] Perform feature engineering on the hydrogen addition ratio set to obtain the processed hydrogen addition ratio set;

[0023] Perform linear regression analysis on the output of the processed hydrogen addition ratio set and the combustion mode switching point model to construct the hybrid model.

[0024] Further, the performing feature engineering on the hydrogen addition ratio set includes:

[0025] Perform maximum-minimum scaling or standardization on the set of hydrogen addition ratios.

[0026] Furthermore, according to the direction of adjustment of the load of the gas turbine, respectively determine whether to switch the combustion mode to obtain a judgment result, including:

[0027] When the gas turbine is in the process of increasing load, input the operation data of the gas turbine into the hybrid model to obtain a predicted load. If the unit load is greater than the predicted load plus the switching dead zone, the judgment result is to switch the combustion mode;

[0028] When the gas turbine is in the process of decreasing load, input the operation data of the gas turbine into the hybrid model to obtain a predicted load. If the unit load is less than the predicted load minus the switching dead zone, the judgment result is to switch the combustion mode.

[0029] Furthermore, the switching dead zone includes:

[0030] The switching dead zone includes a first switching dead zone and a second switching dead zone;

[0031] When the gas turbine is in the process of increasing load, the first switching dead zone is used to judge the switching of the combustion mode. When the gas turbine is in the process of decreasing load, the second switching dead zone is used to judge the switching of the combustion mode.

[0032] The present invention also proposes a combustion mode switching method based on the above-mentioned combustion mode switching judgment method for a gas turbine.

[0033] A combustion mode switching method for a gas turbine includes:

[0034] According to the judgment result, switch the combustion mode of the gas turbine.

[0035] Furthermore, the switching of the combustion mode of the gas turbine further includes:

[0036] First complete the pre-action, and then switch the combustion mode of the gas turbine.

[0037] Furthermore, after first completing the pre-action and then switching the combustion mode of the gas turbine, it further includes:

[0038] The pre-action includes air purging, nitrogen purging, and fuel pre-filling.

[0039] To achieve the above object, the present invention also proposes a combustion mode switching judgment device for a gas turbine.

[0040] A combustion mode switching judgment device for a gas turbine includes:

[0041] An acquisition and construction module, configured to acquire the operation data set and the hydrogen blending ratio set of the gas turbine, and construct a hybrid model according to the operation data set and the hydrogen blending ratio;

[0042] A comprehensive judgment module, configured to input the real-time operation data of the gas turbine into the hybrid model, and respectively judge whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine, so as to obtain a judgment result.

[0043] Further, in the acquisition and construction module:

[0044] Train the combustion mode switching point model according to the operation data set, and then construct the hybrid model through linear regression according to the combustion mode switching point model and the hydrogen blending ratio set.

[0045] Further, in the acquisition and construction module:

[0046] Compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel pressure, gas turbine speed and / or gas turbine load.

[0047] Further, in the acquisition and construction module:

[0048] Randomly select 70% of the data in the operation data set as the training set and 30% of the data as the validation set to obtain the combustion mode switching point model.

[0049] Further, in the acquisition and construction module:

[0050] Perform feature engineering on the hydrogen blending ratio set to obtain the processed hydrogen blending ratio set;

[0051] Perform linear regression analysis on the output of the processed hydrogen blending ratio set and the combustion mode switching point model to construct the hybrid model.

[0052] Further, in the acquisition and construction module:

[0053] Perform maximum-minimum scaling or standardization on the hydrogen blending ratio set.

[0054] Further, in the comprehensive judgment module:

[0055] When the gas turbine is in the process of increasing load, input the operation data of the gas turbine into the hybrid model to obtain the predicted load. If the unit load is greater than the predicted load plus the switching dead zone, the judgment result is to switch the combustion mode;

[0056] When the gas turbine is in the process of load reduction, the operating data of the gas turbine is input into the hybrid model to obtain the predicted load. If the unit load is less than the predicted load minus the switching dead zone, the judgment result is to switch the combustion mode.

[0057] Furthermore, in the comprehensive judgment module:

[0058] The switching dead zone includes a first switching dead zone and a second switching dead zone;

[0059] When the gas turbine is in the process of increasing load, it is determined that the switching combustion mode uses the first switching dead zone, and when the gas turbine is in the process of decreasing load, it is determined that the switching combustion mode uses the second switching dead zone.

[0060] Based on the above technical solution, the present invention has at least the following beneficial effects:

[0061] 1. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By calculating the combustion mode switching model in real time, the combustion mode switching is accurately controlled, avoiding the problem that the fixed limit value cannot reflect the optimal timing of the combustion mode switching under the actual state. At the same time, by establishing a mixed model under different hydrogen blending ratios, the adaptability of the system is significantly improved, so that the combustion mode switching can adapt to gas turbines under various operating conditions, the advancement and response speed of the control system are improved, and the flexibility and responsiveness under different hydrogen blending ratios are ensured. In addition, different adjustment strategies are set by different load adjustment directions, which further improves the accuracy of the combustion mode switching.

[0062] 2. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By calculating the combustion mode switching point in real time, the combustion fluctuation problem of the hydrogen-mixed gas turbine during the combustion mode switching is effectively suppressed, the main technical difficulties when burning hydrogen are solved, the risk of flameout or combustion pulsation is reduced, the success rate of combustion switching is improved, and the combustion efficiency is improved. At the same time, it helps to improve the stability of the power grid and ensure the reliability of power supply.

[0063] 3. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By optimizing the combustion mode switching, NOX emissions are effectively reduced, meeting national environmental protection requirements, and causing less pollution to the environment. It not only improves the overall performance of the gas turbine, but also optimizes combustion efficiency and environmental emissions, achieving the dual goals of improving efficiency and environmental performance while ensuring safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The accompanying drawings forming a part of this invention are used to provide a further understanding of the invention. The schematic embodiments and descriptions thereof of the invention are used to explain the invention and do not unduly limit the invention. In the drawings:

[0065] Figure 1 shows a flowchart of a method for judging the combustion mode switching of a gas turbine in an embodiment;

[0066] Figure 2 shows a flowchart of a method for switching the combustion mode of a gas turbine in an embodiment;

[0067] Figure 3 shows a schematic diagram of a device for judging the combustion mode switching of a gas turbine in an embodiment. Detailed implementation manners

[0068] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0069] The following further describes the present invention in detail with reference to specific embodiments, and these embodiments should not be construed as limiting the scope claimed by the present invention.

[0070] Embodiment

[0071] To solve the problem of comprehensively deciding the switching of the combustion mode by combining the operating state of the gas turbine and the hydrogen addition condition, the present invention proposes a method, a switching method and a device for judging the combustion mode switching of a gas turbine.

[0072] To achieve the above object, the present invention also proposes a method for judging the combustion mode switching of a gas turbine.

[0073] As Figure 1 shows a flowchart of a method for judging the combustion mode switching of a gas turbine in an embodiment of the present invention. As Figure 1 described therein, the process mainly includes the following steps:

[0074] Collect the operating data set and the hydrogen addition ratio set of the gas turbine, and construct a hybrid model according to the operating data set and the hydrogen addition ratio;

[0075] Furthermore, during the operation of the gas turbine, collect the inlet temperature of the compressor, the inlet pressure of the compressor, the outlet temperature of the compressor, the outlet pressure of the compressor, the fuel temperature, the fuel quantity pressure and the gas turbine speed under different environmental parameters and different operating conditions at a fixed hydrogen addition ratio to construct an operating data set. At the same time, construct a corresponding hydrogen addition ratio set according to the operating data set.

[0076] Furthermore, a combustion mode switching point model is constructed, and the combustion mode switching point model can be expressed as:

[0077] W0 = f1(T1, P1, T2, P2, T f , P f , n),

[0078] where W0 is the optimal combustion mode switching load at the switching point, f1 is a neural network function, T1 is the compressor inlet temperature, P1 is the compressor inlet pressure, T2 is the compressor outlet temperature, P2 is the compressor outlet pressure, T f is the fuel temperature, P f is the fuel quantity pressure, and n is the gas turbine speed.

[0079] In this embodiment, the RBF neural network algorithm is used as an example to establish a calculation model. With the compressor inlet temperature T1, the compressor inlet pressure P1, the compressor outlet temperature T2, the compressor outlet pressure P2, the fuel temperature T f , the fuel quantity pressure P f , and the gas turbine speed n as the inputs of the neural network and the optimal combustion mode switching load W0 as the output, the neural network can be expressed as:

[0080] W0 = f(x),

[0081] where f(x) is the function of the neural network. This neural network adopts a three-layer structure. The number of neurons i in the input layer is 7, the number of neurons o in the output layer is 1, and the number of neurons p in the hidden layer can be calculated as:

[0082]

[0083] where z is a constant with a value range of [1, 10].

[0084] The Gaussian kernel function selected in this embodiment can be expressed as:

[0085]

[0086] And this is used as the radial basis function. Among them, k(‖x - x c ‖) is the output value of the Gaussian radial basis function, x is the input vector, x c is the center of the kernel function, and σ is the width of the radial basis function. Therefore, the output of the RBF network can be expressed as:

[0087]

[0088] where y j represents the jth output variable of the neural network, w ijrepresents the connection value between the i-th hidden node and the j-th output node, σ is the width of the radial basis function, and x p is the variable vector of the input layer of the neural network, and c i is the center vector of the i-th radial basis function.

[0089] In the constructed operating dataset, 70% of the data is randomly selected as the training set, and the remaining 30% is used as the validation set. The combustion mode switching point model is trained and obtained, and the accuracy of the combustion mode switching point model is verified. If the accuracy is insufficient, the hyperparameters are modified and the training steps are repeated until the model prediction accuracy reaches more than 95%.

[0090] In other embodiments, if the collected data of the operating dataset of the gas turbine is not limited to the compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel quantity pressure, and gas turbine speed under different environmental parameters and different operating conditions, etc., when a relatively large-scale dataset is obtained, the stochastic gradient optimization algorithm can be used to further optimize the model.

[0091] The hydrogen addition ratio set obtained by collection is [10, 15, 20, 30, 50, 55]. The hydrogen addition ratio set is subjected to maximum-minimum scaling and can be expressed as:

[0092]

[0093] where x' is the data after maximum-minimum scaling, taking the first three digits of the calculation result, x is the original data in the hydrogen addition ratio set, min(x) is the minimum value in the hydrogen addition ratio set, and max(x) is the maximum value in the hydrogen addition ratio set. The calculated hydrogen addition ratio set is [0, 0.11, 0.22, 0.44, 0.88, 1].

[0094] A hybrid function is constructed according to the hydrogen addition ratio set and the combustion point switching model, and the expression is as follows:

[0095] W i = f2(W0, C H ),

[0096] where W i is the predicted load at the switching point, f2 is the function of the hybrid model, W0 is the optimal combustion mode switching load at the switching point, and C H is the hydrogen addition ratio.

[0097] In specific implementation, linear regression analysis is used to fit the hybrid model. The hybrid model f2 can be specifically expressed as:

[0098] f2(W0, C H ) = a·W0 + b·C H + c,

[0099] Where W0 is the optimal combustion mode switching load at the switching point, C H is the hydrogen blending ratio, and a, b, and c are the parameters of the linear regression model respectively. 80% of the hydrogen blending ratio set and the corresponding optimal combustion mode switching load are randomly used as the training set of the model, and the remaining 20% of the data is used as the validation set. Through linear regression analysis, the estimated values of a, b, and c are obtained, and finally the hybrid model is obtained.

[0100] Input the real-time operation data of the gas turbine into the hybrid model, and respectively judge whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine to obtain a judgment result.

[0101] In the switching dead zone in the present invention, it means a set load change range or time period as a buffer zone, and the combustion mode is not switched within this range or time period. The specific load change range data is set as the first switching dead zone and the second switching dead zone according to the empirical values of load increase and load decrease.

[0102] Furthermore, collect the real-time operation data of the gas turbine. The collected data has the same data type as the data for constructing the model. In this embodiment, it includes the compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel quantity pressure, gas turbine speed, and hydrogen blending ratio. Input the collected real-time operation data into the hybrid model, and output a predicted load W i , if the gas turbine is in the process of increasing load, the relationship between the unit load, the predicted load, and the first switching dead zone satisfies:

[0103] W>W i +D u ,

[0104] Where W is the unit load, W i is the predicted load, D u is the first switching dead zone. When the unit load W of the gas turbine is greater than the predicted load W i and the first switching dead zone D u , then the judgment result is to switch the combustion mode, otherwise the judgment result is to continue running; if the gas turbine is in the process of reducing load, the relationship between the unit load, the predicted load, and the second switching dead zone satisfies:

[0105] W<W i -D d ,

[0106] Where W is the unit load, W i is the predicted load, D d is the second switching dead zone. When the unit load W of the gas turbine is less than the predicted load W i and the second switching dead zone Dd Then the judgment result is to switch the combustion mode; otherwise, the judgment result is to continue operation.

[0107] To achieve the above object, the present invention further provides a method for switching the combustion mode of a gas turbine.

[0108] As Figure 2 shown in Figure 2 a flowchart of a method for switching the combustion mode of a gas turbine according to an embodiment of the present invention, and as described in

[0109] Further, according to the obtained judgment result, the next decision is made for the gas turbine. If the judgment result is to switch the combustion mode, the gas turbine is first pre-acted. In this embodiment, the pre-action may be air purging, nitrogen purging, and fuel pre-filling in the fuel pipeline. After all the pre-actions are completed, the combustion mode of the gas turbine is switched.

[0110] To achieve the above object, the present invention further provides a device for judging the switching of the combustion mode of a gas turbine.

[0111] As Figure 3 shown in Figure 3 a schematic diagram of a device for judging the switching of the combustion mode of a gas turbine according to an embodiment of the present invention, and as described

[0112] The acquisition and construction module is used to acquire the operation data set and the hydrogen addition ratio set of the gas turbine, and construct a hybrid model according to the operation data set and the hydrogen addition ratio;

[0113] Further, in the acquisition and construction module:

[0114] According to the operation data set, the combustion mode switching point model is trained, and then the hybrid model is constructed by linear regression according to the combustion mode switching point model and the hydrogen addition ratio set.

[0115] Further, in the acquisition and construction module:

[0116] The compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel pressure, gas turbine speed, and / or gas turbine load.

[0117] Further, in the acquisition and construction module:

[0118] Randomly select 70% of the data in the operation data set as the training set and 30% of the data as the validation set to obtain the combustion mode switching point model.

[0119] Further, in the acquisition and construction module:

[0120] Perform feature engineering on the hydrogen doping ratio set to obtain the processed hydrogen doping ratio set;

[0121] Perform linear regression analysis on the processed hydrogen doping ratio set and the output of the combustion mode switching point model to construct the hybrid model.

[0122] Further, in the acquisition and construction module:

[0123] Perform min-max scaling or normalization on the hydrogen doping ratio set.

[0124] The comprehensive judgment module is used to input the real-time operation data of the gas turbine into the hybrid model, and respectively judge whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine to obtain a judgment result.

[0125] Further, in the comprehensive judgment module:

[0126] When the gas turbine is in the process of increasing load, input the operation data of the gas turbine into the hybrid model to obtain a predicted load. If the unit load is greater than the predicted load plus the switching dead zone, the judgment result is to switch the combustion mode;

[0127] When the gas turbine is in the process of decreasing load, input the operation data of the gas turbine into the hybrid model to obtain a predicted load. If the unit load is less than the predicted load minus the switching dead zone, the judgment result is to switch the combustion mode.

[0128] Further, in the comprehensive judgment module:

[0129] The switching dead zone includes a first switching dead zone and a second switching dead zone;

[0130] When the gas turbine is in the process of increasing load, the first switching dead zone is used to judge the switching of the combustion mode. When the gas turbine is in the process of decreasing load, the second switching dead zone is used to judge the switching of the combustion mode.

[0131] In summary, from the above description, it can be seen that the above embodiments of the present invention achieve the following technical effects:

[0132] 1. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By calculating the combustion mode switching model in real time, the combustion mode switching is accurately controlled, avoiding the problem that the fixed limit value cannot reflect the optimal timing of the combustion mode switching under the actual state. At the same time, by establishing a mixed model under different hydrogen blending ratios, the adaptability of the system is significantly improved, so that the combustion mode switching can adapt to gas turbines under various operating conditions, the advancement and response speed of the control system are improved, and the flexibility and responsiveness under different hydrogen blending ratios are ensured. In addition, different adjustment strategies are set by different load adjustment directions, which further improves the accuracy of the combustion mode switching.

[0133] 2. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By calculating the combustion mode switching point in real time, the combustion fluctuation problem of the hydrogen-mixed gas turbine during the combustion mode switching is effectively suppressed, the main technical difficulties when burning hydrogen are solved, the risk of flameout or combustion pulsation is reduced, the success rate of combustion switching is improved, and the combustion efficiency is improved. At the same time, it helps to improve the stability of the power grid and ensure the reliability of power supply.

[0134] 3. The present invention proposes a combustion mode switching judgment method, switching method and device for a gas turbine. By optimizing the combustion mode switching, NOX emissions are effectively reduced, meeting national environmental protection requirements, and causing less pollution to the environment. It not only improves the overall performance of the gas turbine, but also optimizes combustion efficiency and environmental emissions, achieving the dual goals of improving efficiency and environmental performance while ensuring safety.

[0135] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0136] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0137] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definable sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in conjunction with these instruction execution systems, apparatus, or devices.

[0138] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application specific integrated circuits with suitable combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0139] It should be noted that in the description of this specification, the description of reference terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

Claims

1. A method for determining combustion mode switching of a gas turbine, characterized in that: include: Collecting an operating data set and a hydrogen blending ratio set of the gas turbine, and constructing a hybrid model according to the operating data set and the hydrogen blending ratio; The real-time operation data of the gas turbine is input into the hybrid model, and whether to switch the combustion mode is determined according to the direction of the load adjustment of the gas turbine to obtain a determination result.

2. The method according to claim 1, characterized in that: Establishing a hybrid model according to the running data set includes: The combustion mode switching point model is trained based on the operating data set, and then the hybrid model is constructed through linear regression based on the combustion mode switching point model and the hydrogen blending ratio set.

3. The method according to claim 2, characterized in that: The operating data set includes: Compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel pressure, turbine speed and / or turbine load.

4. The method according to claim 2, characterized in that: The step of training the combustion mode switching point model according to the operating data set includes: 70% of the data in the operating data set are randomly selected as a training set, and 30% of the data are randomly selected as a verification set to obtain the combustion mode switching point model.

5. The method according to claim 2, characterized in that: The hybrid model is constructed by linear regression according to the combustion mode switching point model and the hydrogen blending ratio, including: Performing feature engineering on the hydrogen doping ratio set to obtain the processed hydrogen doping ratio set; The processed hydrogen blending ratio set and the output of the combustion mode switching point model are subjected to linear regression analysis to construct the hybrid model.

6. The method according to claim 5, characterized in that: The step of subjecting the hydrogen doping ratio set to feature engineering comprises: The hydrogen blending ratio set is subjected to maximum and minimum scaling or normalization.

7. The method according to claim 1, characterized in that: The step of determining whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine and obtaining the determination result comprises: When the gas turbine is in the process of increasing load, the operation data of the gas turbine is input into the hybrid model to obtain the predicted load. If the unit load is greater than the predicted load plus the switching dead zone, the judgment result is to switch the combustion mode; When the gas turbine is in the process of load reduction, the operating data of the gas turbine is input into the hybrid model to obtain the predicted load. If the unit load is less than the predicted load minus the switching dead zone, the judgment result is to switch the combustion mode.

8. The method according to claim 7, characterized in that: The switching dead zone includes: The switching dead zone includes a first switching dead zone and a second switching dead zone; When the gas turbine is in the process of increasing load, it is determined that the switching combustion mode uses the first switching dead zone, and when the gas turbine is in the process of decreasing load, it is determined that the switching combustion mode uses the second switching dead zone.

9. A combustion mode switching method based on the combustion mode switching judgment method for a gas turbine according to any one of claims 1 to 8, characterized in that: include: The combustion mode of the gas turbine is switched according to the determination result.

10. The method according to claim 9, characterized in that: The switching of the combustion mode of the gas turbine further includes: The pre-action is completed first, and then the combustion mode of the gas turbine is switched.

11. The method according to claim 10, characterized in that: After the pre-action is completed, the combustion mode of the gas turbine is switched, and the method further includes: The pre-actions include air purging, nitrogen purging and fuel pre-filling.

12. A combustion mode switching judgment device for a gas turbine, characterized in that: include: A collection and construction module, used for collecting an operation data set and a hydrogen blending ratio set of the gas turbine, and constructing a hybrid model according to the operation data set and the hydrogen blending ratio; The comprehensive judgment module is used to input the real-time operation data of the gas turbine into the hybrid model, and judge whether to switch the combustion mode according to the direction of the load adjustment of the gas turbine to obtain the judgment result.

13. The device according to claim 12, characterized in that: In the acquisition building block: The combustion mode switching point model is trained based on the operating data set, and then the hybrid model is constructed through linear regression based on the combustion mode switching point model and the hydrogen blending ratio set.

14. The device according to claim 13, characterized in that: In the acquisition building block: Compressor inlet temperature, compressor inlet pressure, compressor outlet temperature, compressor outlet pressure, fuel temperature, fuel pressure, turbine speed and / or turbine load.

15. The device according to claim 13, characterized in that: In the acquisition building block: 70% of the data in the operating data set are randomly selected as a training set, and 30% of the data are randomly selected as a verification set to obtain the combustion mode switching point model.

16. The device according to claim 13, characterized in that: In the acquisition building block: Performing feature engineering on the hydrogen doping ratio set to obtain the processed hydrogen doping ratio set; The processed hydrogen blending ratio set and the output of the combustion mode switching point model are subjected to linear regression analysis to construct the hybrid model.

17. The device according to claim 16, characterized in that: In the acquisition building block: The hydrogen blending ratio set is subjected to maximum and minimum scaling or normalization.

18. The device according to claim 12, characterized in that: In the comprehensive judgment module: When the gas turbine is in the process of increasing load, the operation data of the gas turbine is input into the hybrid model to obtain the predicted load. If the unit load is greater than the predicted load plus the switching dead zone, the judgment result is to switch the combustion mode; When the gas turbine is in the process of load reduction, the operating data of the gas turbine is input into the hybrid model to obtain the predicted load. If the unit load is less than the predicted load minus the switching dead zone, the judgment result is to switch the combustion mode.

19. The device according to claim 18, characterized in that In the comprehensive judgment module: The switching dead zone includes a first switching dead zone and a second switching dead zone; When the gas turbine is in the process of increasing load, it is determined that the switching combustion mode uses the first switching dead zone, and when the gas turbine is in the process of decreasing load, it is determined that the switching combustion mode uses the second switching dead zone.

Images