Training method of emission prediction model and emission prediction method
By constructing a numerical simulation model of biofuels, the initial emission concentration data is generated, which solves the problem of insufficient sample data in the training of biofuel emission prediction model, and achieves efficient and accurate emission prediction.
Patent Information
- Application Number
- CN202411795553.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-09
- Publication Date
- 2025-03-18
AI Technical Summary
The existing biofuel emission prediction methods are difficult to efficiently train the prediction model due to insufficient sample data, resulting in inaccuracy and inefficiency of prediction.
By obtaining the target alternative components of biofuel and the factors affecting initial combustion, combustion models, turbulence models and geometric models are constructed, and numerical simulations are performed to generate initial emission concentration data. The combined data set is used to train machine learning models.
It realizes efficient acquisition of large and accurate sample data, improves the training efficiency and prediction accuracy of emission prediction models, and solves the problem of insufficient sample data.
Smart Images
Figure CN119646517A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of aircraft emission prediction, and in particular to a training method for an emission prediction model and an emission prediction method. Background Art
[0002] With the increasing severity of environmental problems and the depletion of fossil energy, biomass energy, as a renewable energy source, has received extensive attention and research. Aircraft biofuel is made from biomass energy. The emissions generated by the combustion process of biofuel include carbon dioxide, carbon monoxide, nitrogen oxides, etc., and the emissions corresponding to biofuels of different components are also different. In order to optimize the combustion process and reduce harmful emissions, and at the same time achieve the optimization of biofuel components, it is necessary to establish an accurate biofuel emission prediction method. However, the existing biofuel emission prediction method usually takes the actual combustion results of biofuel as sample data, and uses the sample data to train a prediction model for emission prediction. This requires a lot of historical time to accumulate sample data, and the sample data of aircraft biofuel is relatively small, so it is difficult to collect enough sample data for training the prediction model. Summary of the invention
[0003] The present application provides an emission prediction model training method and an emission prediction method to solve the technical problem of not being able to efficiently obtain a large amount of sample data.
[0004] In a first aspect, the present application provides a method for training an emission prediction model, the method comprising:
[0005] Obtain target replacement components of biofuels for aircraft, multiple groups of initial combustion influencing factors, and initial machine learning models;
[0006] Construct the combustion model, turbulence model, and geometric model of the aircraft;
[0007] According to the target alternative component and the multiple groups of initial combustion influencing factors, numerical simulation is performed within the geometric model using the combustion model and the turbulence model to obtain multiple groups of initial emission concentrations;
[0008] combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set;
[0009] The sample data set is used to train and verify the initial machine learning model to obtain an emissions prediction model.
[0010] In some embodiments, obtaining a target replacement component for a biofuel for an aircraft includes:
[0011] Procurement of biofuels for aircraft; and identification of initial replacement components for biofuels;
[0012] Determine the target physical and chemical indicators of biofuels;
[0013] By using an optimization algorithm, the initial replacement component is adjusted so that the physical and chemical indexes of the initial replacement component are close to the target physical and chemical indexes, thereby obtaining the target replacement component.
[0014] In some embodiments, the target physical and chemical indicators include: density, viscosity, surface tension, net calorific value, distillation point, saturated vapor pressure, carbon content, hydrogen content, and cetane number.
[0015] In some embodiments, the optimization algorithm includes one of the following: a genetic algorithm, a particle swarm optimization algorithm.
[0016] In some embodiments, according to the target replacement component and the multiple groups of initial combustion influencing factors, numerical simulation is performed using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations, including:
[0017] Determine the chemical reaction kinetic mechanism file corresponding to the target alternative component;
[0018] The chemical reaction kinetic mechanism file is input into the combustion model, the multiple groups of initial combustion influencing factors are input into the combustion model and the turbulence model, and numerical simulation is performed using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations.
[0019] In some embodiments, the initial machine learning model is trained and verified using the sample data set to obtain an emission prediction model, including:
[0020] Randomly divide the sample data set into a training set and a test set;
[0021] Using the training set to train the initial machine learning model to obtain an intermediate model;
[0022] Using the test set to verify the accuracy of the intermediate model to obtain a verification result;
[0023] Whether the test verification result is qualified;
[0024] When the verification result is determined to be qualified, the intermediate model is used as the emission prediction model.
[0025] In some embodiments, the method further comprises:
[0026] Determine the delayed ignition time corresponding to the target replacement component through shock tube experiments;
[0027] The delayed ignition time is used to detect whether the accuracy of the target alternative component meets the requirements; when it is determined that it meets the requirements, numerical simulation is performed within the geometric model using the combustion model and the turbulence model based on the target alternative component and the multiple groups of initial combustion influencing factors to obtain multiple groups of initial emission concentrations.
[0028] In some embodiments, the initial machine learning model includes one of the following: a random forest model, a neural network model.
[0029] In a second aspect, the present application also proposes an emission prediction method, the method comprising:
[0030] Obtaining target combustion influencing factors and emission prediction models of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio;
[0031] The target combustion influencing factors are input into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by any of the above-mentioned emission prediction model training methods.
[0032] In a third aspect, the present application proposes a training device for an emission prediction model, comprising:
[0033] A first acquisition module is used to obtain a target replacement component of the biofuel of the aircraft, multiple groups of initial combustion influencing factors, and an initial machine learning model;
[0034] Construction module, used to build the combustion model, turbulence model, and geometric model of the aircraft;
[0035] A calculation module, configured to perform numerical simulation using the combustion model and the turbulence model within the geometric model according to the target alternative component and the multiple groups of initial combustion influencing factors, to obtain multiple groups of initial emission concentrations;
[0036] A combination module, used for combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set;
[0037] A training module is used to train and verify the initial machine learning model using the sample data set to obtain an emission prediction model.
[0038] In a fourth aspect, the present application proposes an emission prediction device, comprising:
[0039] The second acquisition module is used to obtain the target combustion influencing factors and emission prediction model of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio;
[0040] A prediction module is used to input the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by any of the above-mentioned emission prediction model training methods.
[0041] In a fifth aspect, the present application proposes an electronic device, comprising a processor and a memory for storing instructions executable by the processor, wherein when the processor executes the instructions, the processor implements the steps of any one of the above-mentioned emission prediction model training methods, or the steps of any one of the above-mentioned emission prediction methods.
[0042] In a sixth aspect, the present application further proposes a storage medium having computer instructions stored thereon, which, when executed by a processor, implement the steps of any of the above-mentioned emission prediction model training methods, or the steps of any of the above-mentioned emission prediction methods.
[0043] The present application can achieve the following beneficial effects: the present application can utilize the target alternative component to accurately characterize the chemical properties of biofuels, and simulate the initial emission concentration of the target alternative component during and after combustion through numerical simulation, thereby efficiently obtaining a large amount of accurate sample data, improving the efficiency of acquiring sample data, and providing a data basis for the subsequent construction of an emission prediction model. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.
[0045] Figure 1 A schematic diagram of a flow chart of a method for training an emission prediction model provided in an embodiment of the present application;
[0046] Figure 2 A schematic diagram of a flow chart of an emission prediction method provided in an embodiment of the present application;
[0047] Figure 3 A schematic diagram of the structure of a training device for an emission prediction model provided in an embodiment of the present application;
[0048] Figure 4 A schematic diagram of the structure of an emission prediction device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0049] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.
[0050] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally indicates that the associated objects before and after are in an "or" relationship.
[0051] 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 existence of other identical elements in the process, method, article or device including the elements.
[0052] With the increasing severity of environmental problems and the depletion of fossil energy, biomass energy, as a renewable energy source, has received extensive attention and research. Aircraft biofuel is made from biomass energy as raw material. The emissions generated by the combustion process of biofuel include carbon dioxide, carbon monoxide, nitrogen oxides, etc., and the emissions corresponding to biofuels of different components are also different. In order to optimize the combustion process and reduce harmful emissions, and at the same time achieve the optimization of biofuel components, it is necessary to establish an accurate biofuel emission prediction method. However, the existing biofuel emission prediction method usually takes the actual combustion results of biofuel as sample data, and uses the sample data to train a prediction model to predict emissions. This requires a lot of historical time to accumulate sample data, and the sample data of biofuel for aircraft is relatively small, so it is difficult to collect enough sample data for training the prediction model. Therefore, in the prior art, there is a certain difficulty in training the prediction model.
[0053] The present application proposes a training method for an emission prediction model and an emission prediction method, which can use simulated data to train a neural network model to obtain an emission prediction model.
[0054] The following introduces an emission prediction model training method and an emission prediction method provided by the present application in conjunction with the accompanying drawings.
[0055] In some embodiments, please refer to Figure 1 , the present application proposes a method for training an emission prediction model. In specific implementation, the method may include the following contents.
[0056] S101: Obtain target replacement components of biofuel for aircraft, multiple groups of initial combustion influencing factors, and an initial machine learning model.
[0057] In some embodiments, the real components of biofuels are very complex, including hundreds of different compounds, and their molecular formulas and chemical reaction processes are difficult to describe in detail, which is not conducive to subsequent numerical simulations. Therefore, the present application uses target surrogate components to represent the real components. The molecular formulas and chemical reactions of the target surrogate components are relatively simple, which helps to simplify the calculation process of subsequent numerical simulations. The physical and chemical indicators of the target surrogate components are very close to the physical and chemical indicators of the real components. Therefore, in the subsequent process of solving the combustion model based on the numerical simulation method, the emission concentration of the target surrogate component can be considered to be equal to the emission concentration of the real component.
[0058] In some embodiments, obtaining a target replacement component of a biofuel for an aircraft comprises:
[0059] S1: Obtaining biofuel for the aircraft; and determining the initial replacement components of the biofuel;
[0060] S2: Determine the target physical and chemical indicators of biofuel;
[0061] S3: By using an optimization algorithm, the initial replacement component is adjusted so that the physical and chemical indicators of the initial replacement component are close to the target physical and chemical indicators, thereby obtaining a target replacement component.
[0062] In some embodiments, the initial replacement components of the biofuel can be set based on experience or component testing experiments, and commonly used replacement components of biofuel can be used as the initial replacement components.
[0063] In some embodiments, after obtaining the biofuel of the aircraft, the target physical and chemical indicators of the biofuel are obtained through experimental measurement, and the target physical and chemical indicators include: density, viscosity, surface tension, net calorific value, distillation point, saturated vapor pressure, carbon content, hydrogen content, and cetane number. The target physical and chemical indicators can reflect the chemical properties of the biofuel during the combustion process.
[0064] In some embodiments, a prediction model for the physicochemical properties of aviation biofuel is established in advance through a neural network and a display function. The prediction model for the physicochemical properties of aviation biofuel can predict the physicochemical indicators (also called physicochemical properties) under different component mixing ratios. Then, the physicochemical indicators of the initial replacement components (the physicochemical indicators of the initial replacement components also include density, viscosity, surface tension, net calorific value, distillation point, saturated vapor pressure, carbon content, hydrogen content, and cetane number) are taken as the object to be optimized, and the molar fraction of the initial replacement components is continuously adjusted using an optimization algorithm (for example, a genetic algorithm or a particle swarm optimization algorithm) so that the physicochemical indicators of the initial replacement components gradually approach the target physicochemical indicators. The physicochemical indicators of the initial replacement components are the output results of the prediction model for the physicochemical properties of aviation biofuel; after multiple adjustments and iterations, a target replacement component whose physicochemical indicators are very close to the real components of the biofuel can be obtained. Among them, the density, viscosity, and distillation point errors of the target replacement components are all controlled to be no more than 10%.
[0065] In some embodiments, the combustion influencing factors are factors that affect the aircraft fuel combustion process and the process of generating emissions. The initial combustion influencing factors include: initial combustion chamber pressure, initial combustion chamber air inlet temperature, and initial fuel blending ratio.
[0066] In some embodiments, the aircraft fuel is burned in the combustion chamber. The aircraft fuel is not pure biofuel, but a mixture of traditional petroleum-based fuel (aviation fuel) and biofuel, and the mixture ratio of the two is the fuel blending ratio.
[0067] In some embodiments, the combustion chamber pressure used in historical operating conditions can be used as the initial combustion chamber pressure, the combustion chamber air inlet temperature used in historical operating conditions can be used as the initial combustion chamber air inlet temperature, and the fuel blending ratio used in historical operating conditions can be used as the initial fuel blending ratio to obtain multiple groups of different initial combustion influencing factors.
[0068] In some embodiments, the initial machine learning model can be a random forest model, a neural network model, etc.
[0069] S102: Construct the combustion model, turbulence model, and geometric model of the aircraft.
[0070] In some embodiments, the combustion model represents the chemical reaction between the target surrogate species and air in the combustion chamber. The turbulence model represents the complex flow process of air and the target surrogate species in the combustion chamber. The geometric model represents the geometric shape of the aircraft combustion chamber.
[0071] S103: According to the target replacement component and the multiple groups of initial combustion influencing factors, numerical simulation is performed within the geometric model using the combustion model and the turbulence model to obtain multiple groups of initial emission concentrations.
[0072] In some embodiments, according to the target replacement component and the multiple groups of initial combustion influencing factors, numerical simulation is performed using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations, specifically including:
[0073] S1: Determine the chemical reaction kinetic mechanism file corresponding to the target alternative component;
[0074] S2: Input the chemical reaction kinetic mechanism file into the combustion model, input the multiple groups of initial combustion influencing factors into the combustion model and the turbulence model, perform numerical simulation using the combustion model and the turbulence model within the geometric model, and obtain multiple groups of initial emission concentrations.
[0075] In some embodiments, the chemical reaction kinetic mechanism file is also called a chemical reaction dynamic mechanism file, which is obtained according to the chemical reaction kinetic theory. The chemical reaction kinetic mechanism file can be regarded as a set of equations describing the combustion process, which specifically describes the combustion reaction path, reaction rate, concentration changes of reactants and products of the target replacement component, and can characterize the combustion characteristics of the target replacement component, predict key behaviors such as energy release, intermediate product generation, temperature change, material conversion and combustion rate during its combustion process, and is regarded as a mathematical description of the combustion behavior.
[0076] In some embodiments, the chemical reaction kinetic mechanism file is input into the combustion model, the multiple groups of initial combustion influencing factors are input into the combustion model and the turbulence model, and the combustion model and the turbulence model are solved by numerical simulation inside the geometric model (i.e., under the condition that the geometric model is satisfied), so that the emission concentration (initial emission concentration) generated by the target alternative component during and after combustion under the action of different initial combustion influencing factors can be obtained. Each group of initial combustion influencing factors generates a corresponding group of initial emission concentrations.
[0077] Specifically, the initial emission concentration includes: the concentration of nitrogen oxides (NOx), the concentration of carbon monoxide, and the concentration of carbon dioxide at each time step.
[0078] S104: Combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set.
[0079] In some embodiments, each group of initial combustion influencing factors and their corresponding initial emission concentrations are combined in a one-to-one correspondence to obtain a sample data set. Each group of combustion influencing factors and their corresponding initial emission concentrations constitute a sample data set.
[0080] Based on the above embodiments, the target alternative component can be used to accurately characterize the chemical properties of the biofuel, and the initial emission concentration of the target alternative component during and after combustion can be simulated by numerical simulation. A large amount of accurate sample data can be obtained efficiently, thereby improving the efficiency of obtaining sample data and providing a data basis for the subsequent construction of an emission prediction model without having to rely on experimental measurement to obtain the emission concentration of the biofuel.
[0081] S105: Using the sample data set to train and verify the initial machine learning model to obtain an emission prediction model.
[0082] In some embodiments, the sample data set is used to train and verify the initial machine learning model to obtain an emission prediction model, specifically including:
[0083] S1: randomly divide the sample data set into a training set and a test set;
[0084] S2: Using the training set to train the initial machine learning model to obtain an intermediate model;
[0085] S3: Verify the accuracy of the intermediate model using the test set to obtain a verification result;
[0086] S4: Check whether the verification result is qualified;
[0087] S5: When the verification result is determined to be qualified, the intermediate model is used as an emission prediction model.
[0088] In some embodiments, initial combustion influencing factors are used as input data of an initial machine learning model, and initial emission concentrations are used as output data of the initial machine learning model for training.
[0089] In some embodiments, 70% of the sample data sets can be randomly selected as training sets, and the remaining 30% of the sample data sets can be used as test sets. The initial machine learning model is trained using the training set to obtain a trained intermediate model. The accuracy of the intermediate model is verified by cross-validation and other methods. When the error between the output data of the intermediate model and the initial emission concentration in the test set is less than a preset error threshold, the verification result is considered to be "qualified". When the verification result is "qualified", the intermediate model is used as an emission prediction model for emission prediction under actual working conditions. The accuracy and generalization ability of the emission prediction model are better.
[0090] In some embodiments, if the error between the output data of the intermediate model and the initial emission concentration in the test set is greater than or equal to a preset error threshold, the verification result is considered to be "unqualified". At this time, more sample data can be added to form an expanded training set, and the expanded training set can be used to train the intermediate model again and verify the accuracy. Alternatively, the steps of inputting the chemical reaction kinetic mechanism file into the combustion model, inputting multiple groups of initial combustion influencing factors into the combustion model and the turbulence model, and performing numerical simulation using the combustion model and the turbulence model within the geometric model can be repeated to regenerate the sample data set and use the regenerated sample data set for training and accuracy verification until the verification result is "qualified".
[0091] In some embodiments, the method further comprises:
[0092] S1: Determine the delayed ignition time corresponding to the target replacement component through shock tube experiments;
[0093] S2: Using the delayed ignition time to detect whether the accuracy of the target alternative component meets the requirements; when it is determined that it meets the requirements, executing numerical simulation based on the target alternative component and the multiple groups of initial combustion influencing factors, using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations.
[0094] In some embodiments, the delayed ignition time is an important indicator for measuring the combustion characteristics of fuel. Through shock tube experiments, the delayed ignition time of the biofuel and the delayed ignition time of the target replacement component are respectively measured; the difference between the delayed ignition time of the biofuel and the delayed ignition time of the target replacement component is calculated, and it is detected whether the difference is less than a preset threshold; if it is determined that the difference is less than the preset threshold, it is considered that the accuracy of the target replacement component meets the requirements; if it is determined that the difference is greater than or equal to the preset threshold, it is considered that the accuracy of the target replacement component does not meet the requirements.
[0095] This application proposes a method for predicting emissions. Figure 2 As shown, the method includes the following contents:
[0096] S201: Obtain target combustion influencing factors and emission prediction model of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio;
[0097] S202: Inputting the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by any of the above-mentioned emission prediction model training methods.
[0098] In some embodiments, the target combustion influencing factors may include: a combustion chamber pressure in a current operating condition, a combustion chamber air inlet temperature in a current operating condition, and a fuel blending ratio in a current operating condition.
[0099] In some embodiments, the target combustion influencing factors are input into the emission prediction model, and the emission prediction model outputs a target emission concentration. The target emission concentration specifically includes: a predicted concentration value of nitrogen oxides (NOx), a predicted concentration value of carbon monoxide, and a predicted concentration value of carbon dioxide at each time step.
[0100] Based on the above embodiments, different target combustion influencing factors may be input for different operating conditions to accurately predict the emission concentrations in different operating conditions.
[0101] Based on the above-mentioned emission prediction model training method, one or more embodiments of the present application also provide an emission prediction model training device. Figure 3 It is a structural schematic diagram of an embodiment of a training device for an emission prediction model provided in the present application. The training device for an emission prediction model provided in the present application may include: a first acquisition module 301, a construction module 302, a calculation module 303, a combination module 304, and a training module 305.
[0102] The first acquisition module 301 is used to acquire a target replacement component of the biofuel of the aircraft, multiple groups of initial combustion influencing factors, and an initial machine learning model;
[0103] A construction module 302 is used to construct a combustion model, a turbulence model, and a geometric model of the aircraft;
[0104] A calculation module 303 is used to perform numerical simulation using the combustion model and the turbulence model within the geometric model according to the target alternative component and the multiple groups of initial combustion influencing factors to obtain multiple groups of initial emission concentrations;
[0105] A combining module 304 is used to combine the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set;
[0106] The training module 305 is used to train and verify the initial machine learning model using the sample data set to obtain an emission prediction model.
[0107] In some embodiments, the first acquisition module 301 is specifically used to obtain biofuel for an aircraft; determine the initial substitute components of the biofuel; determine the target physicochemical indicators of the biofuel; and adjust the initial substitute components through an optimization algorithm so that the physicochemical indicators of the initial substitute components are close to the target physicochemical indicators, thereby obtaining the target substitute components.
[0108] In some embodiments, the calculation module 303 is specifically used to determine the chemical reaction kinetic mechanism file corresponding to the target alternative component; input the chemical reaction kinetic mechanism file into the combustion model, input the multiple groups of initial combustion influencing factors into the combustion model and the turbulence model, and perform numerical simulation using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations.
[0109] In some embodiments, the training module 305 is specifically used to randomly divide the sample data set into a training set and a test set; use the training set to train the initial machine learning model to obtain an intermediate model; use the test set to verify the accuracy of the intermediate model to obtain a verification result; detect whether the verification result is qualified; and when it is determined that the verification result is qualified, use the intermediate model as an emission prediction model.
[0110] Based on the above-mentioned emission prediction method, one or more embodiments of the present application also provide an emission prediction device. Figure 4 It is a structural diagram of an embodiment of an emission prediction device provided by the present application. The emission prediction device provided by the present application may include: a second acquisition module 401 and a prediction module 402.
[0111] The second acquisition module 401 is used to acquire the target combustion influencing factors and emission prediction model of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio;
[0112] The prediction module 402 is used to input the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by any of the above-mentioned emission prediction model training methods.
[0113] It should be noted that the above-mentioned device may also include other implementation modes according to the description of the method embodiment. The specific implementation mode may refer to the description of the relevant method embodiment, which will not be described one by one here.
[0114] The present application also provides an electronic device, including at least one processor and a memory for storing processor executable instructions, and the processor can implement any of the above method embodiments when executing the instructions. For example, when the instructions are executed by the processor, the following steps are implemented: obtaining the target replacement components, multiple groups of initial combustion influencing factors, and initial machine learning models of the aircraft's biofuel; constructing the aircraft's combustion model, turbulence model, and geometric model; according to the target replacement components and the multiple groups of initial combustion influencing factors, using the combustion model and the turbulence model to perform numerical simulations within the geometric model to obtain multiple groups of initial emission concentrations; combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set; using the sample data set to train and verify the initial machine learning model to obtain an emission prediction model.
[0115] The present application also provides an electronic device, comprising at least one processor and a memory for storing processor executable instructions, wherein the processor can implement any of the above method embodiments when executing the instructions. For example, when the instructions are executed by the processor, the following steps are implemented: obtaining target combustion influencing factors and emission prediction models of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, fuel blending ratio; inputting the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by any of the above emission prediction model training methods.
[0116] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.
[0117] It should be noted that the electronic device described above may also include other implementations according to the description of the method or device embodiment. The specific implementation methods may refer to the description of the relevant method embodiment, which will not be described one by one here.
[0118] The present application also provides a storage medium, which is a computer-readable storage medium, and the storage medium stores computer program instructions, which, when executed, implement the following: obtaining a target alternative component of the aircraft's biofuel, multiple groups of initial combustion influencing factors, and an initial machine learning model; constructing a combustion model, a turbulence model, and a geometric model of the aircraft; performing numerical simulations using the combustion model and the turbulence model within the geometric model based on the target alternative component and the multiple groups of initial combustion influencing factors to obtain multiple groups of initial emission concentrations; combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set; and using the sample data set to train and verify the initial machine learning model to obtain an emission prediction model.
[0119] The present application also provides a storage medium, which is a computer-readable storage medium, and which stores computer program instructions. When the computer program instructions are executed, the following are achieved: obtaining target combustion influencing factors and an emission prediction model of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio; inputting the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; and the emission prediction model is trained by any of the above-mentioned emission prediction model training methods.
[0120] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0121] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0122] In summary, although the present application has been disclosed as above with preferred embodiments, the above preferred embodiments are not intended to limit the present application. Ordinary technicians in this field can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined in the claims.
Claims
1. A method for training an emission prediction model, characterized in that: include: Obtain target replacement components of biofuels for aircraft, multiple groups of initial combustion influencing factors, and initial machine learning models; Construct the combustion model, turbulence model, and geometric model of the aircraft; According to the target alternative component and the multiple groups of initial combustion influencing factors, numerical simulation is performed within the geometric model using the combustion model and the turbulence model to obtain multiple groups of initial emission concentrations; combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set; The sample data set is used to train and verify the initial machine learning model to obtain an emissions prediction model.
2. The method according to claim 1, characterized in that Obtaining targeted alternative components for aircraft biofuels, including: Procurement of biofuels for aircraft; and identification of initial replacement components for biofuels; Determine the target physical and chemical indicators of biofuels; By using an optimization algorithm, the initial replacement component is adjusted so that the physical and chemical indexes of the initial replacement component are close to the target physical and chemical indexes, thereby obtaining the target replacement component.
3. The method according to claim 2, characterized in that The target physical and chemical indicators include: density, viscosity, surface tension, net calorific value, distillation point, saturated vapor pressure, carbon content, hydrogen content, and cetane number.
4. The method according to claim 2, characterized in that: The optimization algorithm includes one of the following: a genetic algorithm and a particle swarm optimization algorithm.
5. The method according to claim 1, characterized in that According to the target replacement component and the multiple groups of initial combustion influencing factors, numerical simulation is performed within the geometric model using the combustion model and the turbulence model to obtain multiple groups of initial emission concentrations, including: Determine the chemical reaction kinetic mechanism file corresponding to the target alternative component; The chemical reaction kinetic mechanism file is input into the combustion model, the multiple groups of initial combustion influencing factors are input into the combustion model and the turbulence model, and numerical simulation is performed using the combustion model and the turbulence model within the geometric model to obtain multiple groups of initial emission concentrations.
6. The method according to claim 1, characterized in that The initial machine learning model is trained and verified using the sample data set to obtain an emission prediction model, including: Randomly divide the sample data set into a training set and a test set; Using the training set to train the initial machine learning model to obtain an intermediate model; Using the test set to verify the accuracy of the intermediate model to obtain a verification result; Whether the test verification result is qualified; When the verification result is determined to be qualified, the intermediate model is used as the emission prediction model.
7. The method according to claim 1, characterized in that The method further comprises: Determine the delayed ignition time corresponding to the target replacement component through shock tube experiments; The delayed ignition time is used to detect whether the accuracy of the target alternative component meets the requirements; when it is determined that it meets the requirements, numerical simulation is performed within the geometric model using the combustion model and the turbulence model based on the target alternative component and the multiple groups of initial combustion influencing factors to obtain multiple groups of initial emission concentrations.
8. The method according to claim 1, characterized in that The initial machine learning model includes one of the following: a random forest model, a neural network model.
9. A method for predicting emissions, characterized in that: include: Obtaining target combustion influencing factors and emission prediction models of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio; The target combustion influencing factors are input into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by the method described in any one of claims 1 to 8.
10. A training device for an emission prediction model, characterized in that: include: A first acquisition module is used to obtain a target replacement component of the biofuel of the aircraft, multiple groups of initial combustion influencing factors, and an initial machine learning model; Construction module, used to build the combustion model, turbulence model, and geometric model of the aircraft; A calculation module, configured to perform numerical simulation using the combustion model and the turbulence model within the geometric model according to the target alternative component and the multiple groups of initial combustion influencing factors, to obtain multiple groups of initial emission concentrations; A combination module, used for combining the multiple groups of initial combustion influencing factors and the multiple groups of initial emission concentrations to obtain a sample data set; A training module is used to train and verify the initial machine learning model using the sample data set to obtain an emission prediction model.
11. An emission prediction device, characterized in that: include: The second acquisition module is used to obtain the target combustion influencing factors and emission prediction model of the aircraft; wherein the combustion influencing factors include: combustion chamber pressure, combustion chamber air inlet temperature, and fuel blending ratio; A prediction module is used to input the target combustion influencing factors into the emission prediction model to obtain a target emission concentration; the emission prediction model is trained by the method described in any one of claims 1 to 8.
12. An electronic device, comprising a processor and a memory for storing processor-executable instructions, wherein the processor implements any one of claims 1 to 8 or the steps of the method described in claim 9 when executing the instructions.
13. A storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed by a processor, the steps of any one of claims 1 to 8 or the method described in claim 9 are implemented.