Gas turbine natural gas hydrogen-doped dynamic combustion control optimization method and system

Abstract

The invention discloses a natural gas hydrogen-doped dynamic combustion control optimization method and system for a gas turbine, relates to the technical field of combustion control of gas turbines, and aims to solve the problems that the stability is poor, NOx emission exceeds the standard and the working condition adaptability is insufficient during hydrogen-doped combustion of the gas turbine. The method comprises the steps that 1, multi-dimensional operation parameters such as combustion chamber temperature distribution, pressure pulsation and tail gas components are obtained, and a combustion state matrix is constructed; 2, based on the matrix, temperature field space features and pressure and tail gas time sequence features are extracted and fused, and a combustion stability index and a NOx emission predicted value are obtained; 3, according to the indexes and the predicted value, the hydrogen doping proportion is dynamically adjusted, fuel distribution is optimized through a multi-stage cyclone ejector, and a local high-temperature area is restrained; and 4, determining an optimal combustion control strategy of premixing-dilution, graded diffusion or self-adaptive mixing in combination with real-time combustion state characteristics and load requirements. The system correspondingly comprises a combustion state module, a prediction module, a tail gas optimization module and a strategy control module. Dual optimization of combustion efficiency and pollutant emission is achieved, different load working conditions are adapted, and hydrogen-doped combustion stability and economical efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of gas turbine combustion control technology, specifically to a method and system for dynamic combustion control optimization of natural gas blending with hydrogen in gas turbines. Background Technology

[0002] Hydrogen-blended combustion technology in gas turbines is an important approach to achieving low-carbon power generation. However, the high diffusivity and rapid combustion characteristics of hydrogen lead to decreased combustion stability when the hydrogen blending ratio changes. In existing technologies, control strategies with a fixed hydrogen blending ratio are difficult to adapt to load fluctuations and changes in fuel characteristics, easily causing problems such as flame flashing, the formation of localized high-temperature zones, and excessive NOx emissions. Furthermore, traditional methods lack real-time monitoring and feedback control of the temperature field, pressure field, and flame dynamics within the combustion chamber, resulting in a trade-off between combustion efficiency and emission reduction targets.

[0003] In recent years, although dynamic hydrogen doping control technology has been proposed, the low degree of sensor data fusion during dynamic hydrogen doping makes it impossible to accurately reflect the combustion state. Secondly, the control strategy is singular and does not combine multi-modal combustion modes to adapt to different operating conditions, making it difficult to achieve long-term stable operation. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a method and system for dynamic combustion control optimization of natural gas blending with hydrogen in gas turbines. By fusing data from multiple sources of sensors, dynamically adjusting the hydrogen blending ratio, and coordinating the switching of multiple combustion modes, the invention achieves dual optimization of combustion efficiency and pollutant emissions.

[0005] This invention is achieved through the following technical solution: In a first aspect, this application provides a method for optimizing dynamic combustion control of hydrogen-blended natural gas in a gas turbine, comprising the following steps: Step 1: Obtain multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on the multi-dimensional operating parameters; Step 2: Based on the combustion state feature matrix, obtain the spatial feature vector of the temperature field in the combustion chamber, as well as the temporal feature vector of the pulsating pressure and exhaust gas composition. Then, fuse the spatial feature vector and the temporal feature vector to obtain the combustion stability index and the predicted NOx emission value. Step 3: Based on the current combustion stability index and NOx emission value, dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer, and optimize the fuel spatial distribution by using the swirl intensity of the multi-stage swirl fuel injector to suppress local high-temperature zones; Step 4: Determine the combustion state characteristics based on the real-time hydrogen doping ratio, swirl intensity, and fuel distribution. Based on the combustion state characteristics and load requirements, determine the optimal combustion control strategy.

[0006] Preferably, the step of constructing the combustion state matrix based on multidimensional operating parameters includes: The multidimensional operating parameters include temperature distribution, pressure pulsation, and exhaust gas composition; Infrared thermal imagers are used to collect the location and gradient of the high-temperature zone in the combustion chamber, and the temperature distribution is determined based on the location and gradient of the high-temperature zone. A high-frequency pressure sensor is used to acquire the combustion chamber pressure signal, and the pressure signal is subjected to fast Fourier transform to obtain the pressure pulsation. The exhaust gas components include the concentrations of NOx, CO, and O2; A combustion state matrix is ​​constructed based on temperature distribution, pressure pulsation, and exhaust gas composition.

[0007] Preferably, the step of obtaining the spatial feature vector of the combustion chamber temperature field and the temporal feature vectors of pulsating pressure and exhaust gas composition based on the combustion state feature matrix includes: The combustion state feature matrix is ​​input into the CNN-LSTM fusion model. The CNN-LSTM fusion model extracts the spatial features of the temperature field through a convolutional neural network. The CNN-LSTM fusion model analyzes the temporal dependence of pressure pulsation and exhaust gas components through a long short-term memory network to obtain the time feature vector.

[0008] Preferably, the fusion of spatial and temporal feature vectors to obtain the combustion stability index and predicted NOx emissions includes: Based on the influence weights of spatial and temporal feature vectors on combustion stability assessment and NOx emissions, the spatial and temporal feature vectors are fused, and then the current combustion stability index and NOx emission value are output through the output layer of the CNN-LSTM fusion model.

[0009] Preferably, the method for dynamically adjusting the hydrogen blending ratio in the natural gas-hydrogen mixer is as follows: The current combustion stability index is compared with the preset value, and the hydrogen blending ratio is adjusted according to the comparison results. If the combustion stability index is below the threshold, it indicates that combustion is unstable, and the hydrogen doping ratio in the mixer should be reduced. If the combustion stability index is higher than the threshold and the NOx emission value exceeds the standard, the hydrogen blending ratio should be increased.

[0010] Preferably, the optimization of fuel spatial distribution and suppression of local high-temperature zones through the swirl intensity of the multi-stage swirl fuel injector includes: The blade angle of the multi-stage swirl fuel injector is controlled according to the combustion state to optimize the uniformity of fuel-air mixing, thereby suppressing the formation of local high-temperature zones and reducing NOx generation. The adjustment process involves adjusting parameters based on changes in the combustion stability index and NOx emission levels to ensure that the combustion state remains within the optimal range.

[0011] Preferably, the combustion state characteristics are determined based on the real-time hydrogen blending ratio, swirl intensity, and fuel distribution state. Based on these combustion state characteristics and load requirements, the optimal combustion control strategy is determined, including: The combustion control strategy includes a premix-dilution mode, a staged diffusion mode, and an adaptive mixing mode; At low loads, a premixed-dilution mode is used, where premixed combustion is combined with exhaust gas recirculation to reduce flame temperature and NOx formation. At high loads, a staged diffusion mode is adopted. The first stage of injection reduces the hydrogen ratio in the mixed fuel to form a stable ignition source; the second stage increases the hydrogen ratio in the mixed fuel to enhance combustion intensity. When the load changes, an adaptive mixing mode is adopted to dynamically adjust the ratio of premixed and diffusion combustion to balance efficiency and emissions.

[0012] Secondly, this application provides a dynamic combustion control optimization system for hydrogen blending in natural gas in a gas turbine, comprising: The combustion state module is used to acquire multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on these parameters. The prediction module is used to obtain the spatial feature vector of the temperature field of the combustion chamber and the temporal feature vector of the pulsating pressure and exhaust gas composition based on the combustion state feature matrix. The spatial feature vector and the temporal feature vector are fused to obtain the combustion stability index and NOx emission value. The exhaust gas optimization module is used to dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer according to the current combustion stability index and NOx emission value, and to optimize the fuel spatial distribution through the swirl intensity of the multi-stage swirl fuel injector to suppress local high temperature zones. The strategy control module is used to determine the combustion state characteristics based on the real-time hydrogen blending ratio, swirl intensity and fuel distribution, and to determine the optimal combustion control strategy based on the combustion state characteristics and load requirements.

[0013] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to implement the steps of the dynamic combustion control optimization method for natural gas blending in gas turbines when executing the computer program.

[0014] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the gas turbine natural gas hydrogen blending dynamic combustion control optimization method.

[0015] Compared with the prior art, the present invention has the following beneficial technical effects: This application provides a dynamic combustion control optimization method for hydrogen-blended natural gas in gas turbines. First, it acquires multi-dimensional operating parameters of the combustion chamber to construct a combustion state matrix, breaking the limitations of traditional single-parameter monitoring. This matrix comprehensively reflects the real-time combustion state from multiple dimensions, including temperature, pressure, and exhaust gas composition, laying a data foundation for subsequent accurate assessment. Next, by integrating the spatial characteristics of the temperature field with the temporal characteristics of pressure and exhaust gas, it obtains the combustion stability index and predicted NOx emissions, avoiding assessment biases caused by relying solely on single-dimensional spatial or temporal analysis, thus making combustion state judgment more accurate. Then, based on the assessment results, it dynamically adjusts the hydrogen blending ratio and optimizes the fuel spatial distribution, solving the problem that a fixed hydrogen blending ratio is difficult to adapt to changes in operating conditions, while simultaneously suppressing local high-temperature zones to control NOx emissions. Finally, it combines combustion state characteristics and load requirements to determine the optimal multi-modal combustion control strategy, achieving synergistic optimization of combustion efficiency and emissions under different load conditions. The overall technical solution forms a closed-loop control logic, effectively balancing combustion stability, low carbon emissions, and operating condition adaptability. It can effectively solve the industry pain point of balancing efficiency and emission reduction in hydrogen-blended gas turbine combustion, possessing strong practicality and innovation.

[0016] This application also proposes a dynamic combustion control optimization system for natural gas blending with hydrogen in a gas turbine, an electronic device, and a computer storage medium, which possess all the advantages of the aforementioned dynamic combustion control optimization method for natural gas blending with hydrogen in a gas turbine. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the optimization of dynamic combustion control for hydrogen blending in natural gas in a gas turbine, as described in this invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0021] A method for optimizing dynamic combustion control of hydrogen-blended natural gas in a gas turbine includes the following steps: Step 1: Obtain multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on the multi-dimensional operating parameters.

[0022] Multidimensional operating parameters include temperature distribution, flame oscillation frequency, pressure pulsation, and exhaust gas composition, including NOx, CO, and O2 concentrations.

[0023] The temperature distribution is monitored in real time by an infrared thermal imager to determine the location and gradient of the high-temperature zone within the combustion chamber. The flame oscillation frequency is captured using a high-speed camera combined with image processing technology to capture the dynamic characteristics of the flame. The pressure pulsation is achieved by using a high-frequency pressure sensor (sampling rate ≥10kHz) to acquire combustion chamber pressure fluctuation data, and by using FFT to obtain the pressure pulsation frequency and amplitude. The exhaust gas components, including NOx, CO, and O2 concentrations, are detected in real time using a spectrometer.

[0024] Temperature distribution directly reflects the heat concentration in the combustion zone. Localized high-temperature zones not only affect combustion efficiency but may also increase pollutant generation. Flame oscillation frequency relates to flame stability; abnormal oscillations can lead to combustion instability and even affect equipment safety. Pressure pulsation reflects dynamic pressure changes during combustion; abnormal pressure fluctuations can impact the combustion chamber structure. The concentrations of nitrogen oxides (NOx), carbon monoxide (CO), and oxygen in the exhaust gas are crucial indicators of complete combustion and compliance with emission standards. Therefore, by collecting these multi-dimensional operating parameters and integrating them to construct a combustion state matrix, the real-time status of the combustion system can be presented comprehensively and accurately from multiple dimensions, laying the foundation for subsequent combustion stability assessment and control strategy adjustments.

[0025] Multi-dimensional operating parameter acquisition: For temperature distribution, infrared thermal imaging is used for real-time monitoring. It can clearly capture the temperature differences in various areas of the combustion chamber, accurately determine the location of the high-temperature zone and the temperature gradient, and intuitively understand the distribution of heat. To obtain the flame oscillation frequency, a high-speed camera is used to capture the dynamic process of the flame. Then, image processing technology is used to analyze the captured image sequence to extract the frequency characteristics of the flame oscillation and understand the stable state of the flame. Pressure pulsations are collected using a high-frequency pressure sensor with a sampling rate of no less than 10 kHz to ensure that rapidly changing pressure signals can be captured. The collected pressure fluctuation data is then processed using a fast Fourier transform (FFT) to obtain the frequency and amplitude of the pressure pulsations and analyze the dynamic change patterns of the pressure. The exhaust gas composition is detected in real time using a spectrometer, which can accurately detect the concentrations of NOx, CO, and O2 in the exhaust gas, providing data support for judging combustion efficiency and pollutant emissions.

[0026] Finally, a combustion state matrix is ​​constructed based on the collected multidimensional operating parameters.

[0027] By collecting multi-dimensional operating parameters and constructing a combustion state matrix covering key dimensions such as temperature, flame, pressure, and exhaust gas composition, the constructed combustion state matrix can comprehensively reflect the operating status of the combustion system, avoiding the problem of one-sided judgment of combustion status due to a single parameter. From a real-time perspective, all equipment is monitored and data is collected in real time, which can promptly capture changes in combustion status. This allows personnel or subsequent control systems to grasp the dynamics of the combustion system immediately, creating conditions for rapid adjustment of control strategies and effectively preventing the escalation of combustion problems caused by data lag. This provides a strong guarantee for the stable and efficient operation of gas turbine natural gas blended with hydrogen combustion.

[0028] Step 2: Construct a combustion state feature matrix based on the data from Step 1, input it into the combustion stability assessment model, and obtain the current combustion stability index and NOx emission prediction value; The combustion stability assessment model is a CNN-LSTM fusion model. The CNN module processes the time-series temperature field image to extract the spatial feature vector of the temperature field (such as the shape and gradient of the high-temperature zone); the LSTM module analyzes the pressure pulsation spectrum and the time sequence of exhaust gas components to capture the time dependence and obtain the time feature vector. Then, the combustion stability index and NOx emission prediction value are fused and output.

[0029] The combustion state feature matrix includes time-series temperature field image data, as well as pressure pulsation spectrum data and exhaust gas composition time-series data.

[0030] For the CNN module, the time-series temperature field image data in the constructed combustion state feature matrix is ​​input. Through network structures such as convolutional layers and pooling layers, the spatial features in the temperature field are gradually extracted, such as whether the specific shape of the high-temperature area is concentrated blocky or dispersed sheety, and the magnitude of the temperature gradient between different regions, to obtain the spatial feature vector of the temperature field.

[0031] The LSTM module receives the pressure pulsation spectrum data and exhaust gas composition time series data from the feature matrix as input. The LSTM module then uses a gating mechanism (input gate, forget gate, output gate) to progressively analyze this time series data, capturing the patterns of data change over time, such as the changing trends of pressure pulsation frequency and amplitude at different times, and the fluctuations in NOx, CO, and O2 concentrations. This allows it to extract the pressure pulsation spectrum and exhaust gas composition time series, resulting in a time feature vector.

[0032] Finally, the temperature field spatial feature vector output by the CNN module and the time feature vector output by the LSTM module are fused. During the fusion, the respective proportions are adjusted according to the influence weights of the two features on combustion stability assessment and NOx emission prediction. Then, the current combustion stability index and NOx emission prediction value are calculated and output through the output layer of the model.

[0033] A CNN-LSTM fusion model is employed to assess combustion stability and predict NOx emissions. This model considers both spatial and temporal characteristics, avoiding the biases inherent in assessments that rely on a single dimension. For example, previous methods focusing solely on the temperature field may overlook anomalous pressure fluctuations, while relying solely on time-series data may miss localized high-temperature zones. This new model, by fusing both types of features, provides a more comprehensive reflection of the combustion state, resulting in a more accurate assessment of the combustion stability index and NOx emission predictions that more closely approximate actual conditions.

[0034] In terms of response speed, the collaborative working mode of CNN and LSTM modules can quickly process the input feature matrix data. The CNN module efficiently extracts spatial features, while the LSTM module quickly captures temporal dependencies. In addition, the feature fusion process of the two modules is optimized, resulting in a fast overall data processing speed. It can output evaluation and prediction results in a short time, meeting the real-time requirements of gas turbine combustion control and avoiding the inability to adjust combustion problems in a timely manner due to data processing delays.

[0035] This model can accurately assess combustion stability and predict NOx emissions by extracting corresponding spatial and temporal characteristics under different operating conditions such as changes in the hydrogen blending ratio of the gas turbine and load fluctuations. It will not have significant errors due to changes in operating conditions, providing a stable and reliable analytical basis for adjusting combustion control strategies under different operating conditions, and ensuring that the gas turbine can operate stably and with low carbon emissions in various operating scenarios.

[0036] Step 3: Based on the combustion stability index and NOx emission value output in Step 2, dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer, and optimize the fuel spatial distribution through a multi-stage swirl fuel injector to suppress local high-temperature zones; The dynamic mixer supports continuous adjustment of the natural gas-hydrogen blending ratio from 0% to 50%. The blending ratio is adjusted in real time based on a comparison of the combustion stability index with a preset threshold. The multi-stage swirl fuel injector uses adjustable guide vanes. By adjusting the vane angle and swirl intensity, it optimizes the uniformity of fuel-air mixing, suppresses local high-temperature zones, and reduces the formation of thermal NOx.

[0037] Based on the combustion stability index and NOx emission prediction values ​​obtained in step 2, an adjustment strategy is formulated. First, examine the combustion stability index: compare it with a preset threshold. If the index is below the threshold, it indicates unstable combustion, possibly due to excessively high hydrogen blending ratio leading to overly vigorous combustion. In this case, reduce the hydrogen blending ratio of the dynamic mixer. If the index is above the threshold and the NOx emission prediction value exceeds the limit, it may indicate excessively concentrated local combustion. The hydrogen blending ratio can be adjusted appropriately, while simultaneously optimizing fuel distribution. It is important to note that the dynamic mixer supports continuous adjustment of the hydrogen blending ratio from 0% to 50%. During the adjustment process, the stability index changes are monitored in real time to ensure accurate ratio adjustment.

[0038] Next, fuel spatial distribution optimization is addressed using a multi-stage swirling fuel injector. The core of this injector is its adjustable guide vanes. Operators or the control system adjust the vane angle based on combustion conditions (such as the detection of localized high-temperature zones and NOx emission prediction levels). A larger angle enhances swirling intensity, resulting in more thorough fuel-air mixing; a smaller angle weakens swirling intensity, preventing combustion instability caused by over-mixing. This adjustment optimizes fuel-air mixing uniformity, spatially suppressing the formation of localized high-temperature zones and reducing thermal NOx generation. The entire adjustment process is real-time and dynamic, continuously adjusting parameters based on changes in the combustion stability index and NOx emission prediction levels to ensure combustion conditions remain within the optimal range.

[0039] From the perspective of combustion stability, dynamically adjusting the hydrogen blending ratio ensures that the fuel combustion rate always adapts to the current operating conditions, avoiding combustion instability issues caused by a fixed hydrogen blending ratio. For example, under load fluctuations, a previously fixed hydrogen blending ratio might have resulted in flame flickering, while dynamic adjustment keeps the combustion stability index stable above the preset threshold, resulting in a more stable combustion state and reducing equipment losses caused by combustion instability.

[0040] In terms of pollutant control, the multi-stage swirl fuel injector optimizes fuel spatial distribution, resulting in more uniform fuel-air mixing, effective suppression of localized high-temperature zones, and a significant reduction in thermal NOx formation. Combined with dynamic hydrogen blending ratio adjustment, NOx emissions can be controlled at even lower levels, meeting low-carbon emission requirements, while also avoiding damage to the combustion chamber structure caused by localized high temperatures.

[0041] In addition, the dynamic mixer's continuous adjustment range of 0%-50% can adapt to different hydrogen source supply conditions and different load requirements. For example, when the hydrogen source is sufficient, the hydrogen blending ratio can be appropriately increased to further reduce carbon emissions; when the load is low, it can be adjusted to a hydrogen blending ratio suitable for low load conditions, balancing efficiency and stability.

[0042] Step 4: Determine the combustion state characteristics based on the real-time hydrogen doping ratio, swirl intensity, and fuel distribution. Based on the combustion state characteristics and load requirements, determine the multi-mode combustion control strategy, i.e., the optimal combustion control strategy.

[0043] The multimodal combustion control strategy, namely the premixed-dilution mode, uses premixed combustion combined with exhaust gas recirculation (EGR) at low loads to reduce flame temperature and NOx formation. The staged diffusion mode employs phased fuel injection at high loads. The first stage injects a low-hydrogen-ratio blend (e.g., 10% hydrogen) to form a stable ignition source. The second stage injects a high-hydrogen-ratio blend (e.g., 40% hydrogen) to enhance combustion intensity. The adaptive mixing mode dynamically adjusts the premixed and diffusion combustion ratios when the load changes rapidly, balancing efficiency and emissions.

[0044] Gas turbines have significantly different core combustion requirements under different load conditions. At low loads, priority should be given to ensuring stable combustion and controlling NOx emissions. At high loads, combustion intensity needs to be increased on the basis of stability to meet power demands. When the load changes rapidly, efficiency and emissions must be balanced to avoid combustion imbalance caused by sudden changes in operating conditions. Combustion characteristics (determined by real-time hydrogen blending ratio, swirl intensity, and fuel distribution state) directly reflect the basic combustion conditions. For example, the hydrogen blending ratio affects the combustion rate, and the swirl intensity and fuel distribution state determine the mixing effect of fuel and air. Whether these conditions are suitable for the current load requirements is the key to selecting a combustion strategy.

[0045] The reason why the premixed-dilution mode is suitable for low load is that premixed combustion allows fuel and air to be fully mixed in advance, reducing uneven local combustion. Combined with exhaust gas recirculation (EGR) to introduce low-temperature exhaust gas, it can directly reduce the flame temperature. The low-temperature environment can effectively suppress the formation of thermal NOx, which perfectly meets the requirements of "stable combustion and low emissions" under low load.

[0046] The staged diffusion mode is used at high loads because low-hydrogen-proportion fuels burn more stably. Injecting them first can establish a stable ignition source and avoid excessively intense combustion caused by direct injection of high-hydrogen fuels. Subsequent injection of high-hydrogen-proportion fuels can take advantage of the high combustion efficiency of hydrogen to improve combustion intensity and meet the power requirements of high loads.

[0047] The adaptive hybrid mode is designed because when the load changes rapidly, neither a single premixed nor a diffusion mode can adapt in time. The premixed mode has a slow response, and the diffusion mode is prone to exceeding emission standards. Dynamically adjusting the ratio of the two modes can flexibly cope with changes in operating conditions, balancing efficiency and emissions.

[0048] First, the combustion state characteristics are determined by real-time data collection of hydrogen doping ratio (obtained from the control parameters of the dynamic mixer), swirl intensity (calculated from the blade angle of the multi-stage swirl injector), and fuel distribution state (combined with the temperature field distribution monitored by the infrared thermal imager and the flame morphology captured by the high-speed camera to determine the current basic combustion state).

[0049] Then, a strategy is selected based on load demand. First, the current load type is determined by the operating parameters of the gas turbine (such as output power and speed): if it is a low load (e.g., output power is less than 30% of rated power), the premix-dilution mode is activated—the fuel supply system is controlled to deliver natural gas-hydrogen mixed fuel in a premixed manner, while the EGR valve is opened to introduce 15%-25% of the exhaust gas into the combustion chamber. During the process, the flame temperature and NOx concentration are monitored in real time, and the EGR rate and hydrogen blending ratio are finely adjusted to ensure that the flame temperature is controlled below 1500℃ and the NOx concentration does not exceed the standard.

[0050] When under high load (e.g., output power exceeding 70% of rated power), the system switches to staged diffusion mode. In the first stage, the fuel injectors are controlled to inject a low-hydrogen-ratio blend (typically 10%-15% hydrogen) for 0.5-1 seconds. After flame monitoring confirms ignition source stability, the system enters the second stage, switching to injecting a high-hydrogen-ratio blend (35%-45% hydrogen). Simultaneously, the injection interval and fuel injection quantity are adjusted according to changes in combustion chamber pressure to avoid excessive pressure fluctuations.

[0051] If a rapid load change is detected (e.g., a power change rate exceeding 5% / second), the adaptive mixing mode is automatically triggered. The control system then calculates the load change trend in real time. For example, when the load rises rapidly, the diffusion combustion ratio is gradually increased (from the initial 30% to 60%) to quickly enhance combustion intensity; when the load drops rapidly, the premixed combustion ratio is increased (from 40% to 70%) to ensure stable combustion. Simultaneously, the hydrogen blending ratio is adjusted (appropriately increased when the load rises and appropriately decreased when it falls) and the swirl intensity is fine-tuned (by adjusting the blade angle based on the fuel distribution). The control parameters are updated every 0.1 seconds throughout the entire process to ensure that the combustion state always adapts to load changes.

[0052] The multimodal strategy can accurately respond to different load requirements. At low loads, the premixed-dilution mode ensures stable combustion and reduces NOx emissions by 30%-40% compared to the traditional stationary mode, avoiding the problems of flame flickering and excessive emissions at low loads. At high loads, the staged diffusion mode can ensure safety through a low-hydrogen ignition source and increase combustion intensity by more than 20% with high-hydrogen fuel, meeting the rated power output requirements. This avoids the contradiction of "unstable combustion at high loads and low efficiency at low loads" that occurs in the traditional single mode.

[0053] In scenarios with dynamic load changes, the advantages of the adaptive mixing mode are even more pronounced. Previously, fixed combustion modes were prone to combustion imbalances of 1-2 seconds when faced with sudden load changes (such as a sudden increase in NOx concentration or a sudden decrease in combustion efficiency). This new mode, however, can adjust the premixing and diffusion ratios within 0.3 seconds, keeping combustion efficiency fluctuations within ±0.5% and NOx concentration fluctuations within ±5 mg / m³, fully meeting the needs of high-frequency load changes such as grid peak shaving.

[0054] Each mode allows for precise matching of combustion status with load demand, avoiding problems such as "local overheating at high loads" and "fuel waste at low loads." This not only extends the service life of the combustion chamber (expected to be 3-5 years longer than the traditional mode) but also improves the overall operating efficiency of the gas turbine by 1.5%-2%, taking into account environmental protection, safety, and economy. It is particularly suitable for the current low-carbon operation requirements of hydrogen-blended gas turbines.

[0055] Step 5: Based on the operating parameters of the switched multimodal combustion control strategy, repeat steps 2-4 to achieve real-time optimization of the multimodal combustion control strategy.

[0056] This application's dynamic combustion control and multimodal optimization method for natural gas blending with hydrogen improves the accuracy of combustion state monitoring through dynamic evaluation using multi-source real-time sensor data and a CNN-LSTM fusion model, solving the problem of delayed response of traditional experimental methods to complex operating conditions. It employs dynamic hydrogen blending ratio control and synergistic optimization with multi-stage swirl injectors, using adjustable guide vanes to suppress local high-temperature zones in real time, reducing NOx emissions and concentration, improving fuel distribution uniformity, and overcoming the operating condition adaptability bottleneck of fixed hydrogen blending strategies. Based on a reinforcement learning closed-loop optimization mechanism, it iteratively updates control parameters using historical data, achieving adaptive adjustment of the hydrogen blending ratio, swirl intensity, and multimodal strategy, improving long-term operational stability, and overcoming the drift defect of open-loop control parameters. This invention has the advantages of real-time sensing (replacing repetitive experiments), intelligent models (reducing the complexity of numerical simulation), and closed-loop optimization (reducing the cost of manual parameter tuning), reducing overall development costs. Furthermore, through modular design, it can be extended to other fuel types (such as biomass gas-hydrogen blends), requiring only updates to numerical calculations and machine learning models for rapid adaptation to new combustion systems.

[0057] Correspondingly, this application also provides a dynamic combustion control optimization system for hydrogen blending of natural gas in a gas turbine, comprising: The combustion state module is used to acquire multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on these parameters. The prediction module is used to obtain the spatial feature vector of the temperature field of the combustion chamber and the temporal feature vector of the pulsating pressure and exhaust gas composition based on the combustion state feature matrix. The spatial feature vector and the temporal feature vector are fused to obtain the combustion stability index and NOx emission value. The exhaust gas optimization module is used to dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer according to the current combustion stability index and NOx emission value, and to optimize the fuel spatial distribution through the swirl intensity of the multi-stage swirl fuel injector to suppress local high temperature zones. The strategy control module is used to determine the combustion state characteristics based on the real-time hydrogen blending ratio, swirl intensity and fuel distribution, and to determine the optimal combustion control strategy based on the combustion state characteristics and load requirements.

[0058] It should be noted that, in the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another device, or some features may be ignored or not executed. The modules described as separate components may or may not be physically separated. The components shown as modules may be one or more physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the modules can be selected to achieve the purpose of the solution in this embodiment according to actual needs.

[0059] Furthermore, in the various embodiments of the present invention, the modules can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated unit described above can be implemented in hardware or as a software functional unit.

[0060] An electronic device provided in this application includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the dynamic combustion control optimization method for natural gas blending with hydrogen in a gas turbine as described in any of the above embodiments.

[0061] Another electronic device provided in this application embodiment may further include: an input port connected to a processor for transmitting multimodal data collected by an external acquisition device to the processor; a display unit connected to the processor for displaying the processor's processing results to the outside world; and a communication module connected to the processor for enabling communication between the electronic device and the outside world. The display unit may be a display panel, a laser scanning display, etc.; the communication method adopted by the communication module includes, but is not limited to, Mobile High Definition Link (HML), Universal Serial Bus (USB), High Definition Multimedia Interface (HDMI), and wireless connection (including Wi-Fi, Bluetooth, Bluetooth Low Energy, and IEEE 802.11s-based communication technology).

[0062] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the steps of the dynamic combustion control optimization method for natural gas blending in gas turbines as described in any of the above embodiments.

[0063] For descriptions of relevant parts of the gas turbine natural gas hydrogen blending dynamic combustion control optimization system, electronic equipment, and computer-readable storage medium provided in this application's embodiments, please refer to the detailed descriptions of the corresponding parts in the gas turbine natural gas hydrogen blending dynamic combustion control optimization method provided in this application's embodiments; they will not be repeated here. Furthermore, parts of the technical solutions provided in this application that are consistent with the implementation principles of corresponding technical solutions in the prior art have not been described in detail to avoid excessive elaboration.

[0064] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method for optimizing dynamic combustion control of hydrogen-blended natural gas in a gas turbine, characterized in that, Includes the following steps: Step 1: Obtain multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on the multi-dimensional operating parameters; Step 2: Based on the combustion state feature matrix, obtain the spatial feature vector of the temperature field in the combustion chamber, as well as the temporal feature vector of the pulsating pressure and exhaust gas composition. Then, fuse the spatial feature vector and the temporal feature vector to obtain the combustion stability index and the predicted NOx emission value. Step 3: Based on the current combustion stability index and NOx emission value, dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer, and optimize the fuel spatial distribution by using the swirl intensity of the multi-stage swirl fuel injector to suppress local high-temperature zones; Step 4: Determine the combustion state characteristics based on the real-time hydrogen doping ratio, swirl intensity, and fuel distribution. Based on the combustion state characteristics and load requirements, determine the optimal combustion control strategy.

2. The method for dynamic combustion control optimization of natural gas blending with hydrogen in a gas turbine according to claim 1, characterized in that, The construction of the combustion state matrix based on multidimensional operating parameters includes: The multidimensional operating parameters include temperature distribution, pressure pulsation, and exhaust gas composition; Infrared thermal imagers are used to collect the location and gradient of the high-temperature zone in the combustion chamber, and the temperature distribution is determined based on the location and gradient of the high-temperature zone. A high-frequency pressure sensor is used to acquire the combustion chamber pressure signal, and the pressure signal is subjected to fast Fourier transform to obtain the pressure pulsation. The exhaust gas components include the concentrations of NOx, CO, and O2; A combustion state matrix is ​​constructed based on temperature distribution, pressure pulsation, and exhaust gas composition.

3. The method for optimizing dynamic combustion control of natural gas with hydrogen in a gas turbine according to claim 1, characterized in that, The process of obtaining the spatial feature vector of the combustion chamber temperature field and the temporal feature vectors of pulsating pressure and exhaust gas composition based on the combustion state feature matrix includes: The combustion state feature matrix is ​​input into the CNN-LSTM fusion model. The CNN-LSTM fusion model extracts the spatial features of the temperature field through a convolutional neural network. The CNN-LSTM fusion model analyzes the temporal dependence of pressure pulsation and exhaust gas components through a long short-term memory network to obtain the time feature vector.

4. The method for dynamic combustion control optimization of natural gas blending with hydrogen in a gas turbine according to claim 3, characterized in that, The process of fusing spatial and temporal feature vectors to obtain the combustion stability index and predicted NOx emissions includes: Based on the influence weights of spatial and temporal feature vectors on combustion stability assessment and NOx emissions, the spatial and temporal feature vectors are fused, and then the current combustion stability index and NOx emission value are output through the output layer of the CNN-LSTM fusion model.

5. The method for optimizing dynamic combustion control of natural gas with hydrogen in a gas turbine according to claim 1, characterized in that, The method for dynamically adjusting the hydrogen blending ratio in the natural gas-hydrogen mixer is as follows: The current combustion stability index is compared with the preset value, and the hydrogen blending ratio is adjusted according to the comparison results. If the combustion stability index is below the threshold, it indicates that combustion is unstable, and the hydrogen doping ratio in the mixer should be reduced. If the combustion stability index is higher than the threshold and the NOx emission value exceeds the standard, the hydrogen blending ratio should be increased.

6. The method for dynamic combustion control optimization of natural gas blending with hydrogen in a gas turbine according to claim 5, characterized in that, The optimization of fuel spatial distribution and suppression of local high-temperature zones through the swirl intensity of the multi-stage swirl fuel injector includes: The blade angle of the multi-stage swirl fuel injector is controlled according to the combustion state to optimize the uniformity of fuel-air mixing, thereby suppressing the formation of local high-temperature zones and reducing NOx generation. The adjustment process involves adjusting parameters based on changes in the combustion stability index and NOx emission levels to ensure that the combustion state remains within the optimal range.

7. The method for dynamic combustion control optimization of natural gas blending with hydrogen in a gas turbine according to claim 1, characterized in that, Combustion state characteristics are determined based on real-time hydrogen blending ratio, swirl intensity, and fuel distribution. Based on these characteristics and load requirements, the optimal combustion control strategy is determined, including: The combustion control strategy includes a premix-dilution mode, a staged diffusion mode, and an adaptive mixing mode; At low loads, a premixed-dilution mode is used, where premixed combustion is combined with exhaust gas recirculation to reduce flame temperature and NOx formation. At high loads, a staged diffusion mode is adopted. The first stage of injection reduces the hydrogen ratio in the mixed fuel to form a stable ignition source; the second stage increases the hydrogen ratio in the mixed fuel to enhance combustion intensity. When the load changes, an adaptive mixing mode is adopted to dynamically adjust the ratio of premixed and diffusion combustion to balance efficiency and emissions.

8. A dynamic combustion control and optimization system for hydrogen blending in natural gas in a gas turbine, characterized in that, include: The combustion state module is used to acquire multi-dimensional operating parameters of the gas turbine combustor and construct a combustion state matrix based on these parameters. The prediction module is used to obtain the spatial feature vector of the temperature field of the combustion chamber and the temporal feature vector of the pulsating pressure and exhaust gas composition based on the combustion state feature matrix. The spatial feature vector and the temporal feature vector are fused to obtain the combustion stability index and NOx emission value. The exhaust gas optimization module is used to dynamically adjust the hydrogen blending ratio of the natural gas-hydrogen mixer according to the current combustion stability index and NOx emission value, and to optimize the fuel spatial distribution through the swirl intensity of the multi-stage swirl fuel injector to suppress local high temperature zones. The strategy control module is used to determine the combustion state characteristics based on the real-time hydrogen blending ratio, swirl intensity and fuel distribution, and to determine the optimal combustion control strategy based on the combustion state characteristics and load requirements.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the dynamic combustion control optimization method for natural gas blending in gas turbines as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the dynamic combustion control optimization method for natural gas blending in gas turbines as described in any one of claims 1-8.

Images