Engine combustion prediction model construction method, application method and electronic device

Abstract

The application discloses an engine combustion prediction model construction method, an application method and electronic equipment, relates to the technical field of engines, and comprises the following steps: acquiring working condition parameters and cylinder pressure data of an engine; determining a target heat release rate curve based on the cylinder pressure data; sequentially determining target model parameters of a post-combustion stage, a main combustion stage and a pre-combustion stage based on the target heat release rate curve; and establishing a prediction parameter mapping model based on the target model parameters of each stage and the working condition parameters, so that the engine determines prediction model parameters based on the prediction parameter mapping model and real-time working condition parameters of the engine, and generates a predicted heat release rate curve. The application can realize the transformation from physical driving modeling to parameterized rapid prediction of the combustion process by converting the cylinder pressure data into the heat release rate curve, using a staged recursive identification model parameter, and establishing a mapping relationship in combination with the working condition parameters, so that the calculation complexity is reduced while the physical consistency and precision are ensured, and the real-time control demand is met.

Description

Technical Field

[0001] This application relates to the field of engine technology, and more specifically, to a method for constructing an engine combustion prediction model, an application method, and an electronic device. Background Technology

[0002] With increasingly stringent energy conservation and emission reduction regulations and the rapid development of hybrid and range-extended powertrains, the accurate prediction and control of engine combustion processes have become increasingly crucial for optimizing vehicle performance. The combustion process directly affects engine power, fuel economy, and emissions levels, while also impacting vehicle ride comfort and noise, vibration, and harshness (NVH) performance. Therefore, accurately and rapidly predicting the in-cylinder combustion process under different operating conditions has become an important research direction in engine control.

[0003] In related technologies, predictions of engine combustion processes typically rely on models based on empirical formulas or simplified zero-dimensional combustion models. These methods often depend on empirical fitting or complex optimization algorithms in determining model parameters. On the one hand, empirical fitting methods lack unified physical constraints, easily leading to model parameters deviating from the actual combustion mechanism, thus affecting prediction accuracy. On the other hand, parameter identification methods based on optimization algorithms involve large computational loads, making it difficult to meet the real-time requirements of control systems. Furthermore, some methods fail to effectively distinguish between different combustion stages when processing the combustion process, making it difficult to eliminate the coupling effects between stages, further reducing the model's robustness and generalization ability. In other words, related technologies suffer from insufficient prediction accuracy of engine combustion processes, complex model parameter acquisition, poor real-time performance, and insufficient adaptability to varying operating conditions. Summary of the Invention

[0004] In the summary section of this application, the relevant technical solutions are described in general terms, and a series of simplified concepts are introduced. These concepts will be further elaborated in the detailed embodiments section. This summary section should not be construed as limiting the key or essential technical features of the claimed solutions, nor is it intended to limit the scope of protection of the claimed solutions.

[0005] The engine combustion prediction model construction method, application method, and electronic equipment provided in this application can transform the combustion process from physical-driven modeling to parameterized rapid prediction by converting cylinder pressure data into heat release rate curves and identifying model parameters through staged recursive analysis, and establishing a mapping relationship with operating condition parameters. This reduces computational complexity and meets real-time control requirements while ensuring physical consistency and accuracy.

[0006] In a first aspect, this application provides a method for constructing an engine combustion prediction model, comprising: acquiring engine operating parameters and cylinder pressure data; determining a target heat release rate curve based on the cylinder pressure data; determining target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage sequentially based on the target heat release rate curve; and establishing a prediction parameter mapping model based on the target model parameters for each stage and the operating parameters, so that the engine determines the prediction model parameters and generates a prediction heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating parameters.

[0007] In some embodiments, the cylinder pressure data includes crankshaft angle, cylinder pressure varying with the crankshaft angle, and cylinder volume; determining the target heat release rate curve based on the cylinder pressure data includes: determining the volume change rate based on the change in cylinder volume with the crankshaft angle; determining the pressure change rate based on the change in cylinder pressure with the crankshaft angle; and determining the target heat release rate curve based on a preset adiabatic index, the cylinder pressure, the cylinder volume, the volume change rate, and the pressure change rate, using a preset thermodynamic relationship, wherein the preset adiabatic index includes a first preset adiabatic index value corresponding to the compression process and a second preset adiabatic index value corresponding to the expansion process.

[0008] In some embodiments, determining the target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage sequentially based on the target heat release rate curve includes: determining the tail combustion characteristic points and tail combustion shape factor of the tail combustion stage based on the target heat release rate curve; subtracting a first fitting curve fitted based on the tail combustion characteristic points and tail combustion shape factor from the target heat release rate curve to obtain a first intermediate curve; determining the main combustion characteristic points of the main combustion stage based on the first intermediate curve; subtracting a second fitting curve fitted based on the main combustion characteristic points and a preset main combustion shape factor from the first intermediate curve to obtain a second intermediate curve; and determining the pre-combustion characteristic points and pre-combustion shape factor of the pre-combustion stage based on the second intermediate curve.

[0009] In some implementations, determining the tail combustion characteristic point and tail combustion shape factor based on the target heat release rate curve includes: performing second-order differentiation on the tail segment of the target heat release rate curve, and determining the point with the largest absolute value of the second derivative as the tail combustion characteristic point; determining a first horizontal coordinate search interval according to a preset tail combustion stage proportion constraint range; constructing a first error function within the first horizontal coordinate search interval, wherein the first error function is used to characterize the cumulative difference between the target heat release rate curve and a first candidate fitting curve within a preset integration interval from the combustion end point to the combustion start point, the first candidate fitting curve being generated based on the tail combustion characteristic point and tail combustion shape factor candidate values; and determining the tail combustion shape factor candidate value that minimizes the value of the first error function as the tail combustion shape factor by traversing multiple tail combustion shape factor candidate values.

[0010] In some implementations, determining the main combustion characteristic point of the main combustion stage based on the first intermediate curve includes: obtaining a preset main combustion shape factor and a first abscissa corresponding to the peak point of the maximum heat release rate of the first intermediate curve; determining a second abscissa search interval as the interval between the first abscissa and the second abscissa of the tail combustion characteristic point; constructing a second error function within the second abscissa search interval, wherein the second error function is used to characterize the cumulative difference between the first intermediate curve and the second candidate fitting curve, the second candidate fitting curve being generated based on the main combustion shape factor and the combustion characteristic candidate point; and determining the combustion characteristic candidate point that minimizes the value of the second error function as the main combustion characteristic point by traversing multiple combustion characteristic candidate points within the second abscissa search interval.

[0011] In some embodiments, determining the pre-combustion characteristic point and pre-combustion shape factor of the pre-combustion stage based on the second intermediate curve includes: determining the point where the maximum heat release rate peak of the second intermediate curve is located as the pre-combustion characteristic point; determining the interval between the abscissa of the preset combustion start point and the main combustion characteristic point as a third abscissa search interval; constructing a third error function within the third abscissa search interval, wherein the third error function is used to characterize the cumulative difference between the second intermediate curve and the third candidate fitting curve, the third candidate fitting curve being generated based on the pre-combustion characteristic point and the candidate value of the main combustion shape factor; and determining the pre-combustion shape factor candidate value that minimizes the value of the third error function as the pre-combustion shape factor by traversing multiple candidate values ​​of the pre-combustion shape factor.

[0012] In some implementations, establishing a predictive parameter mapping model based on the target model parameters and operating condition parameters at each stage includes: acquiring multiple sets of operating condition parameters corresponding to offline operating conditions, and the tail combustion characteristic point, main combustion characteristic point, pre-combustion characteristic point, tail combustion shape factor, main combustion shape factor, and pre-combustion shape factor corresponding to each set of operating condition parameters, wherein the operating condition parameters include at least multiple of engine speed, fuel injection quantity, intake manifold pressure, exhaust manifold pressure, intake manifold temperature, exhaust manifold temperature, and exhaust gas recirculation rate; establishing a first mapping relationship matrix based on the operating condition parameters and the corresponding tail combustion characteristic point, main combustion characteristic point, and pre-combustion characteristic point under the multiple sets of offline operating conditions; establishing a second mapping relationship matrix based on the operating condition parameters and the corresponding tail combustion shape factor, main combustion shape factor, and pre-combustion shape factor under the multiple sets of offline operating conditions; and combining the first mapping relationship matrix and the second mapping relationship matrix as the predictive parameter mapping model.

[0013] In some embodiments, the step of enabling the engine to determine prediction model parameters and generate a predicted heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating parameters includes: inputting the real-time operating parameters into the first mapping matrix to obtain target exhaust combustion feature points, target main combustion feature points, and target pre-combustion feature points; inputting the real-time operating parameters into the second mapping matrix to obtain target exhaust combustion shape factors, target main combustion shape factors, and target pre-combustion shape factors; and generating the predicted heat release rate curve based on the target exhaust combustion feature points, target main combustion feature points, target pre-combustion feature points, target exhaust combustion shape factors, target main combustion shape factors, and target pre-combustion shape factors.

[0014] Secondly, this application also provides a method for applying an engine combustion prediction model, comprising: determining prediction model parameters based on the prediction parameter mapping model described in the first aspect and the real-time operating parameters of the engine, and generating a prediction heat release rate curve.

[0015] Thirdly, this application also provides an engine combustion prediction model construction device, comprising: a data acquisition unit for acquiring engine operating parameters and cylinder pressure data; a curve determination unit for determining a target heat release rate curve based on the cylinder pressure data; a parameter determination unit for sequentially determining target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage based on the target heat release rate curve; a model establishment unit for establishing a prediction parameter mapping model based on the target model parameters and operating parameters for each stage; and a curve prediction unit for enabling the engine to determine prediction model parameters based on the prediction parameter mapping model and the engine's real-time operating parameters, and generating a prediction heat release rate curve.

[0016] Fourthly, this application also provides an electronic device, including: a memory and a processor, wherein the processor is configured to execute a computer program stored in the memory to implement the steps of the engine combustion prediction model construction method of the first aspect or the engine combustion prediction model application method of the second aspect.

[0017] Fifthly, this application also provides a computer-readable storage medium storing computer-executable instructions or a computer program, wherein when the computer-executable instructions or the computer program are executed by a processor, the steps of the engine combustion prediction model construction method described in the first aspect or the engine combustion prediction model application method described in the second aspect are implemented.

[0018] Sixthly, this application also provides a computer program product, including a computer program or computer executable instructions, which, when executed by a processor, implement the steps of the engine combustion prediction model construction method described in the first aspect or the engine combustion prediction model application method described in the second aspect.

[0019] In summary, this application derives the heat release rate curve based on cylinder pressure data, realizing the conversion from measurable physical quantities to combustion process state quantities. Cylinder pressure data is a fundamental signal that can be directly obtained from engine benches or vehicle systems. By converting it into a heat release rate curve through thermodynamic relationships, it essentially transforms the combustion process, which cannot be directly observed, into a calculable and analyzable form. This ensures that subsequent model parameter identification is based on the actual combustion physical process, rather than relying on empirical assumptions, thereby improving the physical consistency and reliability of the prediction from the source. Determining the model parameters for each stage in the order of exhaust combustion, main combustion, and pre-combustion is equivalent to decomposing the overall heat release process and solving it recursively. Since different combustion stages are distributed sequentially in time and have an additive effect on the heat release rate curve, fitting the exhaust combustion stage first can avoid its interference with the fitting of the previous stage. Then, the main combustion and pre-combustion parts are gradually separated, so that the parameter identification of each stage is performed on clean data, which can reduce the error propagation caused by parameter coupling. This approach makes the model parameters at each stage more independent and stable, thereby improving the overall fitting accuracy. Establishing a mapping relationship between the target model parameters at each stage and the operating condition parameters enables a structural transformation from offline identification to online invocation. Since model parameters (such as feature points and shape parameters at each stage) are essentially functions of operating condition variables, establishing a mapping model can compress the complex combustion process into a set of parameter expressions that vary with the operating conditions. This transforms the problem, which originally required iterative calculations or optimizations, into a function query problem, thereby reducing computational complexity and providing a feasible path for real-time applications. Under real-time operating condition input, the mapping model directly determines the prediction model parameters and generates the heat release rate curve, achieving rapid forward prediction of the combustion process. This eliminates the reliance on real-time iterative calculations for combustion prediction, allowing results to be obtained in one step through parameterized expression. This significantly improves computational efficiency and response speed while maintaining prediction accuracy, thus meeting the real-time and stability requirements of the engine control system. In summary, the engine combustion prediction model construction method provided in this application transforms cylinder pressure data into heat release rate curves and identifies model parameters through a phased recursive approach. By combining these parameters with operating conditions to establish a mapping relationship, the method enables the transformation of the combustion process from physical-driven modeling to parameterized rapid prediction. This reduces computational complexity and meets real-time control requirements while ensuring physical consistency and accuracy. Attached Figure Description

[0020] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating an engine combustion prediction model construction method provided in this application embodiment; Figure 2 This is a schematic diagram of the composition structure of an engine combustion prediction model construction device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the composition structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] The terms used in the specification, claims, and drawings of this application, such as "first," "second," "third," "fourth," etc. (if any), are used to distinguish similar objects and not to describe a specific order or sequence. Therefore, it is to be understood that these terms can be used interchangeably where appropriate, allowing the described embodiments to be used in different orders, unless specifically required by the illustrations or description. Furthermore, the terms "is" and "has," and any variations thereof, are intended to cover, non-exclusively, all possible constituent elements. For example, a process, method, system, product, or apparatus comprising several steps or units is not necessarily limited to the steps or units explicitly listed, but may also include other steps or units not explicitly listed, or steps or units inherent to the process, method, product, or apparatus.

[0022] In this application, a "module" or "unit" refers to a computer program or part of a computer program that has a specific function and works in conjunction with other related parts to achieve a predetermined goal. These modules or units can be implemented by software, hardware (e.g., processing circuitry or memory), or a combination of both. One or more processors or memories can implement one or more modules or units. Furthermore, each module or unit can also be part of a larger module or unit.

[0023] The technical solutions of this application will be described in detail below with reference to the accompanying drawings of the embodiments. It should be noted that the described embodiments are only a part of this application, and not all embodiments. In the following description, the "some embodiments" mentioned are only a subset of all possible embodiments, which may be the same or different subsets, and different embodiments can be combined with each other without conflict.

[0024] Figure 1 This is a schematic flowchart illustrating an engine combustion prediction model construction method provided in an embodiment of this application. For example, see [link to example]. Figure 1 The engine combustion prediction model construction method provided in this application embodiment may include the following steps 101 to 104: Step 101: Obtain engine operating parameters and cylinder pressure data.

[0025] In some examples, operating condition parameters are fundamental operational data used to quantitatively characterize the real-time operating status of the engine. They can intuitively reflect the engine's power output, intake and exhaust operation, fuel injection, and other operating states. They are key inputs for the combustion model to adapt to changing operating conditions and achieve adaptive parameter adjustment. These parameters can be collected by the engine control unit (ECU). The ECU, in conjunction with various high-precision sensors mounted on the engine, samples various state data during engine operation in real time. For example, operating condition parameters may include real-time engine speed, single-cycle fuel injection quantity, intake manifold operating pressure, exhaust manifold operating pressure, intake manifold ambient temperature, exhaust manifold ambient temperature, exhaust gas recirculation (EGR), and other operating parameters, comprehensively covering different dimensions of engine operation.

[0026] Cylinder pressure data is physical data that matches the engine crankshaft rotation cycle and characterizes the thermodynamic state inside the cylinder. It can accurately reflect the state changes of the entire process of compression, combustion, and expansion inside the cylinder. It is the raw data for analyzing the thermodynamic laws of combustion in the cylinder and calculating the heat release characteristics of combustion. It can be acquired by relying on high-precision embedded cylinder pressure sensors. The sensors are synchronously linked with the engine crankshaft rotation system and complete high-frequency data sampling in accordance with the crankshaft rotation rhythm. For example, cylinder pressure data can include the engine crankshaft angle, the cylinder pressure value at each crankshaft angle, and the cylinder volume value, which completely records the physical state changes of the cylinder in a single working cycle of the engine.

[0027] By implementing step 101, engine operating parameters and cylinder pressure data are obtained. On the one hand, boundary conditions describing the engine's operating state (such as speed, fuel injection quantity, etc.) are obtained. On the other hand, raw measurement signals that can truly reflect the in-cylinder combustion process are obtained, which can provide a complete data foundation for subsequent combustion process analysis.

[0028] Step 102: Determine the target heat release rate curve based on cylinder pressure data.

[0029] In some examples, the target heat release rate curve is a characteristic curve representing the in-cylinder combustion characteristics of an engine. It is composed of multiple sets of heat release rate data points corresponding to different crankshaft angles, continuously fitted together. This curve can intuitively reflect the rate and trend of heat release during in-cylinder fuel combustion in the engine's working cycle, fully reflecting the entire combustion process from pre-combustion to main combustion and exhaust combustion. It serves as a quantitative basis for determining in-cylinder combustion stability, combustion rate, and combustion progress. By analyzing the dynamic relationship between in-cylinder pressure and cylinder volume at different crankshaft angles, the heat release rate of the fuel at each sampling moment can be derived. By integrating the heat release rate data corresponding to all crankshaft angles, a continuous and complete target heat release rate curve can be generated.

[0030] By implementing step 102, the target heat release rate curve is determined based on cylinder pressure data, realizing the transformation from measurable pressure signals to combustion process characterization quantities. This transforms the combustion heat release process, which was originally difficult to observe directly, into a quantifiable function form. This process shifts combustion analysis from indirect observation to direct characterization, which not only improves the precision of the combustion process description but also provides a data foundation with clear physical meaning for subsequent model parameter extraction, thereby enhancing the physical consistency of the model.

[0031] Step 103: Based on the target heat release rate curve, determine the target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage in sequence.

[0032] In some examples, the pre-combustion stage is the initial start-up stage of in-cylinder combustion in the engine, corresponding to the operating range from the start of combustion to the beginning of intense combustion. During this stage, the atomized fuel in the cylinder is gradually activated and oxidized, with a low and gradual overall heat release rate, reflecting the fuel activation characteristics at the initial stage of engine combustion. It is an important combustion range characterizing the speed of combustion start-up and combustion stability. The main combustion stage is the power-operating stage of in-cylinder combustion in the engine, corresponding to the operating range from the end of the pre-combustion stage to the beginning of the combustion tail. During this stage, a large amount of fuel in the cylinder is concentratedly oxidized and burned, and the heat release rate reaches the peak of the entire combustion cycle. Most of the engine's effective power and heat energy output comes from this stage, directly determining the engine's power output and combustion thermal efficiency. The tail combustion stage is the final stage of in-cylinder combustion in the engine, corresponding to the operating range from the end of the main combustion stage to the end of overall combustion. During this stage, only a small amount of residual fuel and incompletely oxidized mixture remain in the cylinder and continue to burn, with the heat release rate gradually decreasing until it reaches zero, reflecting the residual heat energy release characteristics in the later stages of engine combustion.

[0033] The target model parameters are quantitative characteristic parameters adapted to each independent combustion stage. They can accurately characterize the unique heat release rate changes and combustion characteristics of the pre-combustion stage, main combustion stage, and tail combustion stage. Different combustion stages correspond to specific target model parameters, which can achieve independent characterization of segmented combustion characteristics, replacing the traditional overall unified combustion parameters. This can effectively improve the fitting and prediction accuracy of the engine's full-cycle combustion process. By extracting the heat release data features of each combustion stage separately, the target model parameters adapted to the corresponding combustion range can be solved separately, realizing the layered, independent, and refined parameter acquisition of the entire combustion process.

[0034] By implementing step 103, the model parameters for each stage are determined sequentially based on the target heat release rate curve in the order of tail combustion, main combustion, and pre-combustion. This is equivalent to breaking down and recursively solving the complex combustion process in stages. By processing the later stages first and then gradually pushing forward, the superposition interference between different stages can be reduced, and the parameter identification of each stage can be carried out under relatively independent data conditions. This can reduce the error propagation caused by parameter coupling and improve the stability of parameter identification and the overall fitting accuracy.

[0035] Step 104: Based on the target model parameters and operating condition parameters of each stage, establish a predictive parameter mapping model so that the engine can determine the predictive model parameters and generate the predicted heat release rate curve based on the predictive parameter mapping model and the engine's real-time operating condition parameters.

[0036] In some examples, the predictive parameter mapping model is a model that takes various engine operating conditions as input conditions and the target model parameters of the pre-combustion stage, main combustion stage, and exhaust combustion stage as output results. It can quantify the influence of different operating conditions on the combustion characteristics of each stage in the cylinder. This model can rely on multiple sets of measured data covering different engine operating states to explore the intrinsic relationship between operating condition parameters and target model parameters of each combustion stage. Through data statistics and fitting modeling, a stable correspondence between operating conditions and combustion characteristic parameters is established, thus completing the construction of the predictive parameter mapping model.

[0037] Real-time operating parameters are engine operating status data continuously sampled and updated by the engine vehicle controller throughout the entire process of vehicle driving and equipment operation. Unlike offline operating parameters used for model training during the modeling phase, real-time operating parameters are characterized by dynamic updates and instantaneous response. They accurately reflect the engine's current instantaneous load state, fuel injection state, and intake and exhaust operating state, serving as online input data for the adaptive adjustment of the combustion model. Predictive model parameters are combustion characteristic parameters adaptively matched and output by the predictive parameter mapping model based on the engine's real-time operating state. These parameters correspond to specific combustion characterization parameters for the engine's pre-combustion, main combustion, and exhaust combustion stages. They can adapt to the oxidation and combustion characteristics of the in-cylinder mixture under the current real-time operating conditions, accurately quantify the heat release rate changes in each combustion stage, and provide computational parameters for combustion curve reconstruction. The predicted heat release rate curve is a virtual combustion characteristic curve generated based on the parameters of the prediction model matched in real time. This curve can completely restore the trend of the change of the in-cylinder fuel heat release rate corresponding to each crankshaft angle in a single engine working cycle. It can characterize the entire combustion state of the engine pre-combustion, main combustion and tail combustion under the current working conditions in real time, and can be directly used for engine combustion process prediction and thermal efficiency evaluation.

[0038] The engine can call the prediction parameter mapping model that has been trained and solidified offline inside the controller, and use the dynamically collected real-time operating parameters as the model input; it can rely on the correlation between the operating conditions and combustion parameters built into the model to solve the prediction model parameters of each stage that are adapted to the current operating state, and then use standardized combustion calculation logic to fit and generate a complete prediction heat release rate curve, so as to realize real-time prediction of engine combustion without sensor assistance.

[0039] By implementing step 104, a mapping model is established between the target model parameters and operating condition parameters at each stage, enabling the transformation from offline analysis to online application. By expressing the combustion model parameters as functional relationships of operating condition variables, the process that originally required repeated calculations or optimization solutions can be transformed into parameter lookup or function calculation, thereby effectively reducing computational complexity and providing a feasible path for real-time invocation in the controller. Based on the prediction parameter mapping model and real-time operating condition parameters, the prediction model parameters are directly determined and the heat release rate curve is generated, transforming the combustion process prediction from a computationally intensive process to a rapid forward calculation process. This avoids the time overhead caused by real-time iterative solutions, significantly improving computational efficiency and response speed while ensuring prediction accuracy, thus meeting the real-time and stability requirements of the engine control system.

[0040] In summary, this application's embodiment derives the heat release rate curve based on cylinder pressure data, realizing the conversion from measurable physical quantities to combustion process state quantities. Cylinder pressure data is a fundamental signal that can be directly obtained from engine benches or vehicle systems. By converting it into a heat release rate curve through thermodynamic relationships, it essentially transforms the combustion process, which cannot be directly observed, into a calculable and analyzable form. This ensures that subsequent model parameter identification is based on the actual combustion physical process, rather than relying on empirical assumptions, thereby improving the physical consistency and reliability of the prediction from the source. Determining the model parameters for each stage in the order of tail combustion, main combustion, and pre-combustion is equivalent to decomposing the overall heat release process and solving it recursively. Since different combustion stages are distributed sequentially in time and have an additive relationship on the heat release rate curve, fitting the tail combustion stage first can avoid its interference with the fitting of the previous stage. Then, the main combustion and pre-combustion parts are gradually separated, so that the parameter identification of each stage is performed on clean data, which can reduce the error transmission caused by parameter coupling. By mapping the model parameters at each stage to make them more independent and stable, the overall fitting accuracy is improved. Establishing a mapping relationship between the target model parameters at each stage and the operating condition parameters enables a structural transformation from offline identification to online invocation. Since model parameters (such as feature points and shape parameters at each stage) are essentially functions of operating condition variables, establishing a mapping model can compress the complex combustion process into a set of parametric expressions that vary with operating conditions. This transforms the problem, which originally required iterative calculations or optimizations, into a function query problem, thereby reducing computational complexity and providing a feasible path for real-time applications. Under real-time operating condition input, the mapping model directly determines the prediction model parameters and generates the heat release rate curve, enabling rapid forward prediction of the combustion process. This eliminates the reliance on real-time iterative calculations for combustion prediction, allowing results to be obtained in one step through parameterized expression. This significantly improves computational efficiency and response speed while maintaining prediction accuracy, thus meeting the real-time and stability requirements of the engine control system. In summary, the engine combustion prediction model construction method provided in this application converts cylinder pressure data into heat release rate curves and uses a phased recursive identification of model parameters. By combining the model with operating condition parameters to establish a mapping relationship, the method can realize the transformation of the combustion process from physical-driven modeling to parameterized rapid prediction. This reduces computational complexity and meets real-time control requirements while ensuring physical consistency and accuracy.

[0041] In some embodiments, the aforementioned cylinder pressure data may include crankshaft angle, cylinder pressure varying with the aforementioned crankshaft angle, and cylinder volume; step 102 may include: determining the volume change rate based on the aforementioned cylinder volume varying with the crankshaft angle; determining the pressure change rate based on the cylinder pressure varying with the crankshaft angle; and determining the target heat release rate curve based on a preset adiabatic index, cylinder pressure, cylinder volume, volume change rate, and pressure change rate, using a preset thermodynamic relationship, wherein the preset adiabatic index may include a first adiabatic index preset value corresponding to the compression process and a second adiabatic index preset value corresponding to the expansion process.

[0042] In some examples, crankshaft angle is an angular parameter used to identify the real-time rotational position of the engine crankshaft and characterize the engine's working process. Its angular reference is set at the top dead center of compression. The crankshaft angle corresponds to the real-time displacement height of the engine piston inside the cylinder and is a benchmark variable relating to the dynamic changes in cylinder pressure and cylinder volume. Cylinder pressure is the physical quantity representing the gas pressure of the air-fuel mixture inside the closed cavity of the engine cylinder. Cylinder pressure continuously and dynamically changes with the piston's up-and-down movement, the compression of the air-fuel mixture, and the heat release from combustion. Cylinder pressure directly reflects the degree of gas compression and the intensity of heat release from combustion inside the cylinder. Cylinder volume is the physical quantity representing the volume of the closed cavity formed between the piston top and the cylinder head. It changes systematically only with the rotational displacement of the crankshaft angle. Under the fixed mechanical structure of the engine, any crankshaft angle corresponds to a unique cylinder volume value.

[0043] The volume change rate is the change in cylinder volume per unit crankshaft rotation angle. It quantifies the rate of expansion and contraction of the cylinder cavity during engine operation. The volume change rate can be obtained by differential calculation of continuously sampled cylinder volume data, accurately reflecting the dynamic change in cylinder volume caused by piston movement. It is a derived parameter for constructing in-cylinder thermodynamic calculation models. The pressure change rate is the change in cylinder pressure per unit crankshaft rotation angle. It quantifies the rate of rise and fall of gas pressure inside the cylinder. The pressure change rate can be obtained by differential calculation based on time-aligned in-cylinder pressure sampling data, directly reflecting the pressure fluctuation characteristics caused by in-cylinder gas compression and fuel combustion heat release.

[0044] The preset adiabatic index is a fixed coefficient used to describe the thermodynamic characteristics of gas compression and expansion inside the engine cylinder. During operation, the engine cylinder is approximately in an adiabatic closed state, with no significant heat loss during gas compression and expansion. The adiabatic index quantifies the thermodynamic state changes of the gas inside the cylinder and is a fundamental constant for in-cylinder thermodynamic calculations. It can include a first preset adiabatic index value corresponding to the compression process and a second preset adiabatic index value corresponding to the expansion process. The first preset adiabatic index value is a fixed value corresponding to the engine compression process. In this embodiment, the first preset adiabatic index value can be set to 1.34. This value is suitable for the thermodynamic characteristics of low-temperature mixture compression and energy storage during the engine compression stroke, accurately matching the working state where there is no intense fuel combustion and only gas compression and temperature rise during the compression stage. The second preset adiabatic index value is a fixed value corresponding to the engine expansion process. In this embodiment, the second preset adiabatic index value can be set to 1.28. This value is suitable for the thermodynamic characteristics of fuel combustion releasing heat and gas expansion doing work during the engine expansion stroke, and adapts to the thermodynamic changes of the high-temperature mixture after combustion.

[0045] Based on preset adiabatic index, in-cylinder pressure, in-cylinder volume, volume change rate, and pressure change rate, the process of determining the target heat release rate curve through preset thermodynamic relationships refers to relying on classic in-cylinder thermodynamic formulas for engines, using measured cylinder pressure-derived data and standardized thermodynamic coefficients as unified inputs, solving for the rate of change of real-time heat release in the cylinder from one angle, and finally fitting to generate a complete target heat release rate curve; where the basic in-cylinder pressure and volume satisfy the thermodynamic formula for in-cylinder state changes:

[0046] In the formula, This represents the cylinder pressure that varies with crankshaft angle. This represents the initial cylinder pressure corresponding to the compression top dead center at -180 degrees. This represents the initial cylinder volume corresponding to the compression top dead center at -180 degrees. This represents the cylinder volume that varies with the crankshaft angle. This represents the preset adiabatic index.

[0047] Using the formula for calculating in-cylinder heat release corresponding to the first law of thermodynamics, the heat release rate corresponding to each crankshaft rotation angle can be calculated. The specific formula is as follows:

[0048] In the formula, This represents the in-cylinder heat release rate as a function of crankshaft angle. Represents the rate of change of volume. This represents the rate of pressure change. By distinguishing between the compression and expansion processes and substituting the corresponding preset values ​​for the first and second adiabatic indices, the heat release rate of the entire working cycle is calculated point by point. By integrating all angle data, the target heat release rate curve can be generated.

[0049] For example, the engine vehicle controller synchronously collects the in-cylinder pressure and in-cylinder volume data corresponding to all crankshaft rotation angles within a single working cycle; it uses a differential algorithm to batch solve the volume change rate and pressure change rate corresponding to all sampling points, and automatically matches the corresponding preset adiabatic index according to the stroke range of the crankshaft rotation angle; it substitutes the thermodynamic relationship into the formula one by one to complete the calculation of the heat release rate of all angles, and simultaneously removes sampling abnormal points and calculation noise data, and finally fits and stitches all effective heat release rate data to generate a continuous, smooth target heat release rate curve that fits the actual combustion state in the cylinder.

[0050] By implementing the above embodiments, basic physical quantities such as crankshaft angle, cylinder pressure and cylinder volume are introduced, and thermodynamic relationships are constructed by combining volume change rate, pressure change rate and adiabatic index of sub-process. The heat release rate curve is directly derived from actual measurement data, so that combustion characterization is based on a strict physical mechanism, avoiding the deviation caused by empirical models. This can improve the physical consistency and accuracy of combustion prediction. At the same time, the calculation process is a deterministic analytical form, which is also conducive to controlling the computational complexity.

[0051] In some embodiments, step 103 may include: determining the tail combustion characteristic points and tail combustion shape factor of the tail combustion stage based on the target heat release rate curve; subtracting a first fitting curve based on the tail combustion characteristic points and tail combustion shape factor from the target heat release rate curve to obtain a first intermediate curve; determining the main combustion characteristic points of the main combustion stage based on the first intermediate curve; subtracting a second fitting curve based on the main combustion characteristic points and a preset main combustion shape factor from the first intermediate curve to obtain a second intermediate curve; and determining the pre-combustion characteristic points and pre-combustion shape factor of the pre-combustion stage based on the second intermediate curve.

[0052] In some examples, the exhaust combustion feature point is a reference coordinate point used to anchor the combustion characteristics of the engine's exhaust combustion stage. It is composed of the crankshaft angle corresponding to the exhaust combustion stage and the heat release rate at that angle. The exhaust combustion feature point is selected from the inflection point of the target heat release rate curve from the main combustion to the exhaust combustion. It can characterize the timing of exhaust combustion initiation and the basic heat release intensity. It is a benchmark parameter for constructing the exhaust combustion fitting model. It can be obtained by performing a second-order difference operation on the tail section of the target heat release rate curve and selecting the sampling point with the largest absolute value of the second derivative. The exhaust combustion shape factor is a model coefficient that quantifies the combustion rate and heat release profile characteristics of the engine's exhaust combustion stage. Its value determines the rate of heat release decay in the exhaust combustion stage. The larger the value, the stronger the cumulative heat release effect of the exhaust combustion and the smoother the combustion end. The smaller the value, the higher the instantaneous heat release rate of the exhaust combustion and the faster the combustion end. The preset value range of the exhaust combustion shape factor is 1.5 to 4. The optimal value can be selected by traversing the candidate values ​​within the range and matching the minimum error principle.

[0053] The first fitted curve is a fitting curve specifically characterizing the theoretical heat release law of the engine's exhaust combustion stage. It is generated based on the solved exhaust combustion feature points and the selected exhaust combustion shape factor. The first fitted curve completely replicates the variation law of the theoretical heat release rate of the exhaust combustion stage with the crankshaft angle under the current operating conditions. The corresponding calculation formula is as follows:

[0054] In the formula, This represents the theoretical heat release rate during the tail combustion stage. Represents crankshaft rotation angle. The crankshaft angle representing the start of combustion. This represents the crankshaft rotation angle corresponding to the tail combustion characteristic point. This represents the heat release rate corresponding to the tail combustion characteristic point. Represents the tail combustion shape factor.

[0055] The first intermediate curve is the residual heat release rate curve after removing the interference of tail combustion heat release. By subtracting the theoretical heat release data of tail combustion corresponding to the first fitted curve point by point from the original target heat release rate curve, the influence of the combustion characteristics of the tail combustion stage on the overall combustion curve is eliminated. The first intermediate curve only retains the real heat release information of the engine's main combustion stage and pre-combustion stage, which can provide a clean data source for subsequent accurate identification of main combustion parameters.

[0056] The main combustion feature point is a reference coordinate point used to anchor the combustion heat release characteristics of the engine's main combustion stage. It consists of the optimal crankshaft angle and the corresponding heat release rate within the main combustion range. The main combustion feature point is limited to the crankshaft angle corresponding to the peak of the overall heat release rate and the crankshaft angle of the tail combustion feature point. It can accurately characterize the peak heat release position and combustion progress of the engine's core power stage and serves as the benchmark for fitting the main combustion model. It can be obtained by traversing a preset angle range and selecting the sampling point corresponding to the minimum fitting error. The main combustion shape factor is a model coefficient that quantifies the stability of the combustion rate in the engine's main combustion stage. Based on the inherent engineering characteristics of concentrated heat release and stable combustion characteristics in the engine's main combustion stage, this application embodiment sets the main combustion shape factor as a fixed constant, eliminating the need for traversal optimization. This can significantly reduce the controller's computational load while ensuring fitting accuracy. In this application embodiment, the preset value of the main combustion shape factor can be 1.5.

[0057] The second fitting curve is a fitting curve specifically characterizing the theoretical heat release law of the engine's main combustion stage. It is generated based on the solved main combustion characteristic points and the preset main combustion shape factor. The second fitting curve completely restores the theoretical heat release change law of the main combustion power stage. The corresponding calculation formula is as follows:

[0058] In the formula, Represents the theoretical heat release rate during the main combustion stage. This represents the crankshaft rotation angle corresponding to the main combustion characteristic point. This represents the heat release rate corresponding to the main combustion characteristic point. Represents the main combustion shape factor.

[0059] The second intermediate curve is the residual heat release rate curve after simultaneously removing the interference of exhaust combustion heat release and main combustion heat release. By subtracting the theoretical heat release data of the main combustion corresponding to the second fitted curve from the first intermediate curve point by point, the combustion interference of the main combustion stage and exhaust combustion stage is completely eliminated. The second intermediate curve only retains the heat release data of the engine pre-combustion stage, which is used to accurately solve the model parameters corresponding to the pre-combustion stage.

[0060] Pre-combustion feature points are reference coordinate points used to anchor the combustion activation characteristics of the engine in the initial stage of pre-combustion. The selection rule for pre-combustion feature points is the sampling point corresponding to the maximum heat release rate of the second intermediate curve. This can accurately reflect the maximum heat release intensity and corresponding time of fuel activation and oxidation in the initial stage of engine combustion, and is the benchmark parameter for fitting the pre-combustion curve. The pre-combustion shape factor is a model coefficient that quantifies the fuel activation rate and combustion initiation intensity in the initial stage of engine pre-combustion. The pre-combustion shape factor is adapted to the characteristics of weak heat release and unstable combustion in the pre-combustion stage. The preset value range is 0.8 to 2.5. By traversing all candidate values ​​within the range, the value with the smallest fitting error can be selected as the final pre-combustion shape factor.

[0061] The formula for calculating the theoretical heat release rate during the pre-combustion stage is as follows:

[0062] In the formula, Represents the theoretical heat release rate during the pre-combustion stage. This represents the crankshaft rotation angle corresponding to the pre-ignition characteristic point. This represents the heat release rate corresponding to the pre-combustion characteristic point. Represents the pre-burning shape factor.

[0063] Through the implementation of the above embodiments, the heat release rate curve is recursively identified in stages of tail combustion, main combustion, and pre-combustion. The fitted part is gradually stripped away in each stage, realizing the layer-by-layer decomposition of the complex combustion process. The parameters of each stage are identified on a relatively independent data basis, which effectively reduces the coupling interference and error transmission between different stages. Thus, the stability of model parameter identification and the overall prediction accuracy can be improved without increasing the computational complexity.

[0064] In some embodiments, the aforementioned determination of the tail combustion characteristic point and tail combustion shape factor based on the target heat release rate curve may include: performing second-order derivative processing on the tail segment of the target heat release rate curve, and determining the point with the largest absolute value of the second derivative as the tail combustion characteristic point; determining a first horizontal coordinate search interval according to a preset tail combustion stage proportion constraint range; constructing a first error function within the first horizontal coordinate search interval, wherein the first error function is used to characterize the cumulative difference between the target heat release rate curve and a first candidate fitting curve within a preset integration interval from the combustion end point to the combustion start point, the first candidate fitting curve being generated based on the tail combustion characteristic point and tail combustion shape factor candidate values; and determining the tail combustion shape factor candidate value that minimizes the value of the first error function as the tail combustion shape factor by traversing multiple tail combustion shape factor candidate values.

[0065] In some examples, the tail end of the target heat release rate curve is the curve range corresponding to the engine combustion end process. The crankshaft angle range corresponding to this range is between the combustion end point and the main combustion end position. Only the slow oxidation heat release behavior of the residual mixture in the cylinder remains in the range, without large-scale concentrated fuel combustion phenomenon, fully bearing all the heat release characteristics of the engine tail combustion stage.

[0066] The second-order derivative processing is a second-order numerical difference operation method for the sampled data of the target heat release rate curve in discrete states. This processing method is used to identify the location of the curvature change of the heat release rate curve and can accurately distinguish the boundary characteristics between the heat release decay in the main combustion stage and the continuous heat release in the tail combustion stage. The corresponding discrete second-order derivative calculation formula is as follows:

[0067] In the formula, The second derivative represents the heat release rate during the tail combustion stage. This represents the tail combustion heat release rate corresponding to the current sampling angle. This represents the tail combustion heat release rate corresponding to the previous adjacent sampling angle. Represents the current crankshaft rotation angle being sampled. This represents the crankshaft rotation angle of the previous adjacent sample.

[0068] The point where the absolute value of the second derivative is the largest is the sampling point where the curvature of the tail section of the target heat release rate curve changes most significantly after the second derivative operation. This point corresponds to the critical inflection point when the main combustion stage transitions to the tail combustion stage during engine combustion. It can accurately mark the starting position of tail combustion. In this embodiment, this inflection point is directly defined as the tail combustion characteristic point and used as the reference coordinate for tail combustion curve fitting.

[0069] The exhaust combustion phase proportion constraint range is a pre-defined threshold range for the proportion of exhaust combustion heat release in the engine vehicle controller. This range is set to 0.05 to 0.15 to limit the maximum and minimum proportion of heat release during the exhaust combustion phase to the total heat release in a single engine cycle. This avoids feature point failure caused by an excessively large or small exhaust combustion proportion, ensuring that the selected feature points closely match the actual combustion patterns of the engine. The first horizontal axis search interval is the effective parameter solution interval defined in the crankshaft angle dimension based on the preset exhaust combustion phase proportion constraint range. This interval starts from the engine combustion endpoint and integrates backwards towards the combustion start point to determine the boundary, filtering out all crankshaft angle sampling points that meet the exhaust combustion heat release proportion requirements. This constitutes a dedicated search interval for exhaust combustion feature points, effectively reducing the parameter traversal range and lowering the controller's computational cost.

[0070] The preset integration interval is a dedicated angular interval for calculating the tail combustion fitting error. The interval ranges from the starting position of the first horizontal coordinate search interval to the end point of engine combustion. This interval completely covers the complete tail combustion process of the engine. All the cumulative calculations of fitting deviations are completed within this interval, ensuring that the error calculation is only for the tail combustion region and is not affected by the heat release data of the main combustion stage and pre-combustion stage.

[0071] The first candidate fitting curve is a theoretical heat release curve generated based on the determined tail combustion characteristic points and any set of tail combustion shape factor candidate values, using a formula specific to the heat release rate of the tail combustion stage. Different tail combustion shape factor candidate values ​​correspond to different first candidate fitting curves, which are used for deviation comparison with the measured target heat release rate curve. Its calculation formula follows the three-segment Vibe function fitting formula for the tail combustion stage.

[0072] The first error function is a cumulative deviation calculation function used to quantify the degree of matching between the first candidate fitted curve and the measured target heat release rate curve. It is used to objectively determine the fitting accuracy of a single candidate parameter. The corresponding calculation formula for the first error function is as follows:

[0073] In the formula, This represents the result of the first error function calculation. This represents the starting crankshaft angle of the first horizontal coordinate search interval. This represents the crankshaft angle corresponding to the end of combustion. This represents the heat release rate corresponding to the first candidate fitted curve. The measured heat release rate corresponds to the target heat release rate curve; the smaller the calculated value of the error function, the higher the fit between the current candidate fitting curve and the measured combustion data.

[0074] Multiple candidate values ​​for the exhaust combustion shape factor are generated by uniformly sampling multiple sets of discrete values ​​within the standard range of the exhaust combustion shape factor. The effective value range of the exhaust combustion shape factor is 1.5 to 4. The engine and vehicle controller can select continuous discrete values ​​in batches within this range as traversal candidate parameters to perform curve fitting and error comparison one by one. Determining the candidate value of the exhaust combustion shape factor that minimizes the first error function value is the optimal solution selection logic for the exhaust combustion model parameters. It can traverse all candidate values ​​of the exhaust combustion shape factor, calculate the corresponding fitting error for each one, and select the candidate parameter corresponding to the smallest cumulative fitting deviation as the final exhaust combustion shape factor, thereby ensuring that the exhaust combustion fitting model fits the actual exhaust combustion characteristics of the engine to the greatest extent.

[0075] By implementing the above embodiments, the exhaust feature points are located using the second derivative extremum, and the search interval is limited by the exhaust proportion constraint. Then, the exhaust shape factor is determined by minimizing the error function. This transforms the exhaust stage parameter identification from empirical selection to optimization based on data features. This not only improves the accuracy and robustness of parameter determination but also avoids the computational burden caused by large-scale blind search, thus balancing accuracy and efficiency.

[0076] In some embodiments, the aforementioned determination of the main combustion characteristic point based on the first intermediate curve may include: obtaining a preset main combustion shape factor and a first abscissa corresponding to the peak point of the maximum heat release rate of the first intermediate curve; determining the interval between the first abscissa and the second abscissa of the tail combustion characteristic point as a second abscissa search interval; constructing a second error function within the second abscissa search interval, wherein the second error function is used to characterize the cumulative difference between the first intermediate curve and the second candidate fitting curve, the second candidate fitting curve being generated based on the main combustion shape factor and combustion characteristic candidate points; and determining the combustion characteristic candidate point that minimizes the value of the second error function as the main combustion characteristic point by traversing multiple combustion characteristic candidate points within the second abscissa search interval.

[0077] In some examples, the peak point of the maximum heat release rate of the first intermediate curve is the feature sampling point with the largest heat release rate value among all discrete sampling points of the first intermediate curve. The theoretical heat release data of the tail combustion stage has been removed from the first intermediate curve, and only the real heat release information of the engine pre-combustion stage and main combustion stage is retained. This peak point corresponds to the combustion moment when the fuel in the engine cylinder is concentrated, the heat energy release rate is the highest, and the power contribution is the greatest. It is the benchmark for defining the starting boundary of the main combustion range.

[0078] The first horizontal axis is the crankshaft angle of the engine corresponding to the peak point of the maximum heat release rate of the first intermediate curve. This angle parameter locks the starting boundary of the intense combustion of the engine's main combustion, represents the starting position of large-scale fuel oxidation and heat release in the cylinder, and is used to limit the lower limit angle of the search range of the main combustion characteristic point.

[0079] The second horizontal axis is the crankshaft angle of the engine corresponding to the exhaust combustion feature point solved in the aforementioned exhaust combustion stage. This angle parameter corresponds to the critical transition position between the end of the main combustion and the start of exhaust combustion, representing the termination boundary of the main combustion heat release, and is used to limit the upper limit angle of the main combustion feature point search range.

[0080] The second horizontal coordinate search interval is a continuous crankshaft angle interval formed by the first horizontal coordinate and the second horizontal coordinate. This interval accurately corresponds to the pure main combustion range of the engine, completely isolating the invalid data interference of the pre-combustion range and the exhaust combustion range. It provides a clean and effective parameter search range for the traversal calculation of main combustion feature candidate points and the fitting of main combustion curve, which can significantly reduce the invalid calculation amount of the vehicle controller.

[0081] The second candidate fitting curve is a theoretical heat release fitting curve generated based on a predetermined fixed main combustion shape factor and any combustion feature candidate point within the second horizontal axis search interval, using a formula specific to the heat release rate of the main combustion stage. Each combustion feature candidate point corresponds to a unique set of second candidate fitting curves, used to simulate the theoretical heat release variation law of the engine's main combustion stage under different candidate points. Its calculation formula is as follows:

[0082] In the formula, This represents the theoretical heat release rate of the main combustion corresponding to the second candidate fitted curve. Represents the real-time crankshaft angle. The crankshaft angle representing the start of combustion. The crankshaft angle representing the candidate combustion feature point. The heat release rate corresponding to the candidate combustion feature point, This represents the preset main combustion shape factor.

[0083] The second error function is a formula for calculating the cumulative error between the second candidate fitted curve and the measured first intermediate curve. The second error function can statistically analyze the overall fitting deviation across the entire main combustion range, objectively evaluating the model fitting accuracy corresponding to the current combustion feature candidate point. Its calculation formula is as follows:

[0084] In the formula, This represents the result of the second error function calculation. Represents the first horizontal coordinate. Represents the second horizontal coordinate. This represents the measured primary combustion heat release rate corresponding to the first intermediate curve. The value represents the theoretical main combustion heat release rate corresponding to the second candidate fitting curve; the smaller the calculated value of the error function, the higher the degree of matching between the candidate fitting curve and the actual combustion heat release law of the engine.

[0085] Multiple combustion feature candidate points are all discrete crankshaft angle sampling points uniformly intercepted within the second horizontal coordinate search interval according to the engine's fixed sampling step size. Each discrete sampling point is independently used as a candidate point for the main combustion feature. These points are then substituted one by one into the main combustion fitting formula to generate the corresponding second candidate fitting curve, used for full-domain traversal error comparison. The combustion feature candidate point that minimizes the second error function value is determined as the main combustion feature point, representing the optimal selection logic for the main combustion feature point. After the traversal calculation of all combustion feature candidate points is completed, the discrete sampling point corresponding to the minimum second error function calculation result is selected as the final main combustion feature point, ensuring that the solved main combustion feature point can best adapt to the engine's actual main combustion heat release characteristics under the current operating conditions.

[0086] By implementing the above embodiments, a search interval is constructed between the peak heat release rate and the tail combustion feature point by fixing the main combustion shape factor and minimizing the error function to determine the main combustion feature point. The parameter identification problem of the main combustion stage is simplified from a multi-parameter optimization problem to a single-parameter search problem. While ensuring the ability to capture the main features of the main combustion stage, the computational complexity is greatly reduced, which can improve the real-time performance of the algorithm and reduce the impact of parameter uncertainty on the prediction results.

[0087] In some embodiments, the aforementioned determination of the pre-combustion characteristic point and pre-combustion shape factor based on the second intermediate curve may include: determining the point where the maximum heat release rate peak of the second intermediate curve is located as the pre-combustion characteristic point; determining the interval between the abscissa of the preset combustion start point and the main combustion characteristic point as the third abscissa search interval; constructing a third error function within the third abscissa search interval, wherein the third error function is used to characterize the cumulative difference between the second intermediate curve and the third candidate fitting curve, the third candidate fitting curve being generated based on the pre-combustion characteristic point and the candidate value of the main combustion shape factor; and determining the pre-combustion shape factor candidate value that minimizes the value of the third error function as the pre-combustion shape factor by traversing multiple pre-combustion shape factor candidate values.

[0088] In some examples, the point where the maximum heat release rate peak of the second intermediate curve is located is the feature point with the largest heat release rate value among all discrete sampling points of the second intermediate curve. The second intermediate curve has completed the double-layer subtraction processing of theoretical heat release data of the tail combustion stage and theoretical heat release data of the main combustion stage, and only retains the original heat release data of the engine pre-combustion stage, without data interference from other combustion stages. This peak point accurately corresponds to the moment when the initial activation oxidation of fuel in the engine cylinder is most intense and the heat release rate is highest, and can intuitively characterize the combustion characteristics of the pre-combustion stage. In the embodiments of this application, this point is directly defined as the pre-combustion feature point.

[0089] The preset combustion start point is a combustion start reference angle that is pre-fixed inside the engine vehicle controller. This angle corresponds to the critical crankshaft angle at which the fuel in the engine cylinder begins to undergo oxidation and release a small amount of heat energy. It is the starting boundary of a single combustion cycle of the engine and is used to limit the lower limit angle of the pre-combustion parameter search range, ensuring that all calculated data belong to the effective pre-combustion range.

[0090] The third horizontal coordinate search interval is a continuous crankshaft angle interval formed by the horizontal coordinates corresponding to the preset combustion start point and the main combustion characteristic point. This interval completely covers the entire pre-combustion combustion process of the engine, completely isolates the main combustion interval from the tail combustion interval, and is specifically used for error traversal calculation of pre-combustion shape factor candidate values. It can effectively reduce the invalid calculation range of the controller and improve the solution efficiency of pre-combustion model parameters.

[0091] The third candidate fitting curve is a theoretical heat release fitting curve generated based on fixed pre-combustion feature points and any set of pre-combustion shape factor candidate values, using a formula specific to the heat release rate in the pre-combustion stage. Each set of pre-combustion shape factor candidate values ​​can generate a unique third candidate fitting curve, used to simulate the theoretical heat release variation law of the engine pre-combustion stage under different parameter conditions. Its calculation formula is as follows:

[0092] In the formula, This represents the theoretical heat release rate of pre-combustion corresponding to the third candidate fitted curve. Represents the real-time crankshaft angle. This represents the crankshaft angle corresponding to the preset combustion start point. This represents the crankshaft rotation angle corresponding to the pre-ignition characteristic point. This represents the heat release rate corresponding to the pre-combustion characteristic point. This represents the candidate value of the pre-burning shape factor.

[0093] The third error function is a formula for calculating the cumulative error between the third candidate fitted curve and the measured second intermediate curve. The third error function can statistically analyze the overall fitting deviation within the complete pre-burning interval, accurately assessing the model fitting accuracy corresponding to the current pre-burning shape factor candidate value. Its calculation formula is as follows:

[0094] In the formula, This represents the result of the third error function calculation. Represents the preset combustion start point. This represents the crankshaft rotation angle corresponding to the main combustion characteristic point. This represents the measured pre-combustion heat release rate corresponding to the second intermediate curve. The theoretical pre-combustion heat release rate corresponding to the third candidate fitting curve is represented; the smaller the calculated value of the error function, the higher the matching degree between the fitting model corresponding to the current candidate parameters and the actual pre-combustion characteristics of the engine.

[0095] Multiple candidate values ​​for the pre-combustion shape factor are generated by discrete sampling within the standard effective range of the pre-combustion shape factor. Based on the combustion characteristics of the engine pre-combustion stage, the preset effective value range for the pre-combustion shape factor is 0.8 to 2.5. The engine and vehicle controller selects discrete values ​​in batches within this range at a fixed step size, using these as candidate values ​​for the pre-combustion shape factor through a process of optimization. The candidate value that minimizes the third error function is determined as the pre-combustion shape factor, representing the optimal solution selection logic for the pre-combustion model parameters. The controller can iterate through all candidate values, calculate the cumulative fitting error of the corresponding third candidate fitting curve for each, and select the candidate value with the smallest error calculation result as the final pre-combustion shape factor. This ensures that the pre-combustion fitting model best matches the heat release characteristics of the engine's initial combustion.

[0096] By implementing the above embodiments, the peak value of the second intermediate curve is directly selected as the pre-combustion feature point, and the error minimization search of the pre-combustion shape factor is performed within a limited range. This allows the identification of parameters in the pre-combustion stage to utilize obvious physical features while avoiding complex multivariate coupling optimization processes. This ensures the accuracy of characterizing the rapid changes in the initial stage of combustion while improving computational efficiency and method stability.

[0097] In some embodiments, step 104 may include: obtaining operating parameters corresponding to multiple sets of offline operating conditions, and tail combustion characteristic points, main combustion characteristic points, pre-combustion characteristic points, tail combustion shape factors, main combustion shape factors, and pre-combustion shape factors corresponding to each set of operating parameters, wherein the operating parameters may include at least multiple of engine speed, fuel injection quantity, intake manifold pressure, exhaust manifold pressure, intake manifold temperature, exhaust manifold temperature, and exhaust gas recirculation rate; establishing a first mapping relationship matrix based on the operating parameters of multiple sets of offline operating conditions and the corresponding tail combustion characteristic points, main combustion characteristic points, and pre-combustion characteristic points; establishing a second mapping relationship matrix based on the operating parameters of multiple sets of offline operating conditions and the corresponding tail combustion shape factors, main combustion shape factors, and pre-combustion shape factors; and combining the first mapping relationship matrix and the second mapping relationship matrix as a prediction parameter mapping model.

[0098] In some examples, multiple offline operating conditions represent various steady-state operating conditions of the engine under bench calibration. These offline conditions differ from the real-time dynamic operating conditions of a vehicle on the road; they occur in a controlled laboratory environment, allowing for precise control of the independent changes in various engine operating parameters. These offline conditions cover combinations of low, medium, and high loads, as well as low, medium, and high speeds, achieving full coverage of the engine's normal operating range and providing sufficient and comprehensive sample data for training the mapping model. The operating parameters corresponding to these offline conditions are a set of multi-dimensional fundamental parameters characterizing the overall engine operating state, collected synchronously during the operation of each offline steady-state condition. These parameters are input variables to the mapping model, directly determining the in-cylinder intake state, mixture atomization state, residual exhaust gas state, and fuel combustion reaction rate, thus distinguishing different engine combustion characteristics.

[0099] Engine speed is the number of crankshaft rotations per unit time, used to characterize the overall speed of engine operation, and directly determines the intensity of gas flow in the cylinder and the duration of combustion reaction; fuel injection quantity is the mass of fuel injected into the cylinder in a single working cycle, used to determine the total energy of combustion in the cylinder, and directly affects the engine's work capacity and heat release intensity; intake manifold pressure is the gas pressure inside the engine's intake manifold, used to characterize the amount of air intake, and determines the concentration of the air-fuel mixture in the cylinder; exhaust manifold pressure is the gas pressure inside the engine's exhaust manifold, used to affect cylinder scavenging efficiency and residual exhaust gas content; intake manifold temperature is the ambient temperature of the gas inside the engine's intake manifold, which can affect the activation temperature of the intake mixture and atomization effect; exhaust manifold temperature is the exhaust gas temperature inside the engine's exhaust manifold, which can indirectly reflect the degree of heat release during combustion in the cylinder; exhaust gas recirculation rate is the proportion of exhaust gas mass that recirculates back to the cylinder to participate in combustion to the total intake gas mass, which can dilute the concentration of the air-fuel mixture in the cylinder, reduce the combustion temperature, and significantly change the engine's combustion rate and heat release profile. The tail combustion characteristic points, main combustion characteristic points, pre-combustion characteristic points, tail combustion shape factors, main combustion shape factors, and pre-combustion shape factors corresponding to each set of operating condition parameters are unique combustion model parameters obtained independently for each set of stable offline operating conditions through the aforementioned step-by-step subtraction fitting and error optimization algorithms. Each set of operating condition parameters uniquely corresponds to a complete set of combustion characteristic point locations and combustion shape factor parameters, which can accurately adapt to the unique in-cylinder heat release law under the corresponding operating conditions and serve as the output label data for mapping model training.

[0100] The first mapping relationship matrix is ​​a mathematical matrix model used to establish the correspondence between multi-dimensional engine operating parameters and three-stage combustion characteristic point parameters. Taking various engine operating parameters as input and outputting tail combustion characteristic points, main combustion characteristic points, and pre-combustion characteristic points, the first mapping relationship matrix can quantify the influence of different operating conditions on the combustion start-up position and heat release peak position of each stage of the engine. The corresponding calculation formula is as follows:

[0101] In the formula, , , These represent the crankshaft rotation angles corresponding to the combustion characteristic points of tail combustion, main combustion, and pre-combustion, respectively. , , These represent the heat release rates corresponding to the three combustion characteristic points; Represents engine speed. Represents the amount of fuel injected in a cycle. This represents the intake manifold pressure. Represents exhaust manifold pressure. This represents the intake manifold temperature. Represents the exhaust manifold temperature. Represents the exhaust gas recirculation rate; Represents the first weight coefficient matrix. This represents the first deviation compensation matrix.

[0102] The second mapping matrix is ​​a mathematical matrix model used to establish the correspondence between multi-dimensional engine operating parameters and three-stage combustion shape factors. Taking various engine operating parameters as input and outputting the tail combustion shape factor, main combustion shape factor, and pre-combustion shape factor, the second mapping matrix can quantify the influence of different operating conditions on the combustion rate and heat release profile of each stage of the engine. The corresponding calculation formula is as follows:

[0103] In the formula, , , These represent the tail combustion shape factor, main combustion shape factor, and pre-combustion shape factor, respectively. This represents the second weighting coefficient matrix. This represents the second deviation compensation matrix, and the remaining parameters are defined in the same way as the first mapping matrix.

[0104] The first mapping matrix responsible for outputting combustion feature point parameters and the second mapping matrix responsible for outputting combustion shape factor parameters can be encapsulated and integrated. The two sets of matrices can be operated independently and output with complementary parameters, which can fully cover all model parameters required for the three-stage combustion of the engine, forming a standardized combustion parameter prediction model that can be adapted to all working conditions, and stored uniformly in the engine vehicle controller.

[0105] Through the implementation of the above embodiments, a mapping relationship matrix between operating parameters under multiple operating conditions and feature points and shape factors at each stage is constructed. The complex offline parameter identification results are transformed into a function model that can be called online, realizing the rapid prediction of combustion model parameters as operating conditions change. This not only solves the problem of complex parameter acquisition, but also improves the generalization ability and adaptability of the model under different speeds, loads and environmental conditions.

[0106] In some embodiments, the aforementioned method of enabling the engine to determine prediction model parameters and generate a predicted heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating parameters may include: enabling the engine to input real-time operating parameters into a first mapping matrix to obtain target exhaust combustion characteristic points, target main combustion characteristic points, and target pre-combustion characteristic points; enabling the engine to input real-time operating parameters into a second mapping matrix to obtain target exhaust combustion shape factors, target main combustion shape factors, and target pre-combustion shape factors; and enabling the engine to generate a predicted heat release rate curve based on the target exhaust combustion characteristic points, target main combustion characteristic points, target pre-combustion characteristic points, target exhaust combustion shape factors, target main combustion shape factors, and target pre-combustion shape factors.

[0107] In some examples, the target tail combustion feature point is the optimal combustion reference coordinate point of the tail combustion stage calculated and output by the first mapping relationship matrix based on the engine's real-time operating parameters. The target tail combustion feature point includes the tail combustion crankshaft angle and corresponding heat release rate corresponding to the real-time operating conditions. It can adapt to the critical combustion characteristics of the transition from the main combustion stage to the tail combustion stage during the engine's dynamic operation, accurately characterize the start-up timing and basic heat release intensity of tail combustion under real-time operating conditions, and serve as the reference parameters for online construction of the tail combustion stage combustion model. The engine vehicle controller can import the pre-processed real-time operating parameters into the first mapping relationship matrix and output the results through the correlation law between the operating conditions and combustion features built into the matrix. The target main combustion feature point is the optimal combustion reference coordinate point for the main combustion stage, calculated and output by the first mapping matrix based on the engine's real-time operating parameters. This target main combustion feature point corresponds to the interval point where the in-cylinder fuel concentrates combustion and performs work under the engine's real-time dynamic operating conditions. It accurately reflects the peak heat release position and intensity of the main combustion stage under the current operating state, directly determining the heat release profile and combustion progress of the engine's core power-performing stage. The target main combustion feature point adaptively updates with the dynamic changes of parameters such as engine speed, fuel injection quantity, and exhaust gas recirculation rate, adapting to the differences in main combustion characteristics under all operating conditions. The target pre-combustion feature point is the optimal combustion reference coordinate point for the pre-combustion stage, calculated and output by the first mapping matrix based on the engine's real-time operating parameters. This target pre-combustion feature point corresponds to the peak point where the in-cylinder fuel undergoes initial atomization activation and weak oxidation heat release under real-time operating conditions. It accurately characterizes the initial activation characteristics and heat release pattern of the engine's combustion start-up stage, adapting to the differences in pre-combustion combustion under different intake temperature and pressure and different exhaust gas recirculation ratios.

[0108] The target exhaust shape factor is a combustion morphology coefficient specific to the exhaust stage, calculated from the second mapping matrix based on real-time engine operating parameters. It quantifies the combustion decay rate of the residual mixture in the engine exhaust under real-time operating conditions, accurately characterizing the speed of heat release and the curvature of the heat release curve during the exhaust stage. It dynamically and adaptively adjusts according to engine load and intake / exhaust conditions, ensuring that the theoretical heat release law of exhaust combustion closely matches the actual combustion state. The target main combustion shape factor is a combustion morphology coefficient specific to the main combustion stage, calculated from the second mapping matrix based on real-time engine operating parameters. It characterizes the heat release stability and combustion intensity of the concentrated combustion stage under real-time engine operating conditions, accurately matching the in-cylinder mixture concentration and intake flow characteristics under different operating conditions, ensuring the stability and accuracy of the main combustion stage model fitting. The target pre-combustion shape factor is a combustion morphology coefficient specific to the pre-combustion stage, calculated and output by the second mapping relationship matrix based on the engine's real-time operating parameters. The target pre-combustion shape factor is used to quantify the initial activation rate and weak heat release fluctuation characteristics of in-cylinder fuel under real-time operating conditions, adapting to operating conditions such as low engine load and small fuel injection quantity, which are prone to unstable combustion, and effectively compensating for the fitting accuracy defects caused by weak heat release and fuzzy features in the pre-combustion stage.

[0109] The process of generating a predicted heat release rate curve based on target exhaust combustion characteristic points, target main combustion characteristic points, target pre-combustion characteristic points, target exhaust combustion shape factors, target main combustion shape factors, and target pre-combustion shape factors involves uniformly substituting all combustion characteristic point parameters and combustion shape factor parameters obtained from online real-time solutions into the three-stage Vibe combustion calculation formula to solve for the theoretical heat release rates of the pre-combustion stage, main combustion stage, and exhaust combustion stage, respectively. By superimposing the heat release rate data of the three independent combustion stages, a complete predicted heat release rate curve adapted to the current real-time operating state of the engine is fitted and generated. The corresponding overall calculation formula is as follows:

[0110] In the formula, Represents the predicted heat release rate of the engine throughout all stages. Represents the real-time crankshaft angle. The crankshaft angle representing the start of combustion. The crankshaft angle represents the target combustion characteristic point corresponding to each combustion stage. This represents the heat release rate corresponding to the target combustion characteristic point in each combustion stage. The target shape factor represents each combustion stage, and the subscript j corresponds to the pre-combustion stage p, the main combustion stage m, and the tail combustion stage t, respectively.

[0111] By implementing the above embodiments, real-time operating parameters are directly input into the mapping model to obtain feature points and shape factors at each stage, and further generate predicted heat release rate curves. This transforms the combustion prediction process from traditional iterative calculation to a fast forward calculation process, which can significantly reduce calculation latency and improve response speed while ensuring prediction accuracy, thus meeting the requirements of engine control systems for real-time performance and stability under dynamic operating conditions.

[0112] Furthermore, as an implementation of the aforementioned method embodiments, this application also provides an engine combustion prediction model construction apparatus for implementing the aforementioned method embodiments. This apparatus embodiment corresponds to the aforementioned method embodiments. For ease of reading, this engine combustion prediction model construction apparatus embodiment will not repeat the details of the aforementioned method embodiments one by one, but it should be understood that the apparatus in this application embodiment can correspondingly implement all the contents of the aforementioned method embodiments. For example... Figure 2 As shown, the engine combustion prediction model construction device 20 includes: a data acquisition unit 201, a curve determination unit 202, a parameter determination unit 203, a model building unit 204, and a curve prediction unit 205. The data acquisition unit 201 is used to acquire the engine's operating parameters and cylinder pressure data; the curve determination unit 202 is used to determine the target heat release rate curve based on the cylinder pressure data; the parameter determination unit 203 is used to determine the target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage sequentially based on the target heat release rate curve; and the model building unit 204 is used to establish a prediction parameter mapping model based on the target model parameters and operating parameters for each stage, so that the engine can determine the prediction model parameters and generate the prediction heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating parameters.

[0113] This application also provides a method for applying an engine combustion prediction model, which can determine the prediction model parameters and generate a predicted heat release rate curve based on the aforementioned prediction parameter mapping model and the engine's real-time operating parameters.

[0114] This application also provides a computer-readable storage medium storing computer-executable instructions or computer programs, which, when executed by a processor, will cause the processor to perform any step of the engine combustion prediction model construction method or engine combustion prediction model application method provided in this application.

[0115] In some embodiments, the computer-readable storage medium may be a random access memory (RAM), a read-only memory (ROM), flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); or it may be a variety of devices that include one or any combination of the above-mentioned memories.

[0116] In some embodiments, computer-executable instructions may take the form of programs, software, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.

[0117] In some embodiments, computer-executable instructions may, but do not necessarily, correspond to files in a file system, and may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple co-located files (e.g., files that store one or more modules, subroutines, or code sections).

[0118] In some embodiments, computer-executable instructions may be deployed to execute on an electronic device, or on multiple electronic devices located at one location, or on multiple electronic devices distributed across multiple locations and interconnected via a communication network.

[0119] like Figure 3 As shown, this application also provides an electronic device 30, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements any step of the above-mentioned engine combustion prediction model construction method or engine combustion prediction model application method.

[0120] This application also provides a computer program product comprising a computer program or computer-executable instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer program or computer-executable instructions from the computer-readable storage medium and executes the computer program or computer-executable instructions, causing the electronic device to perform any step of the engine combustion prediction model construction method described above.

[0121] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for constructing an engine combustion prediction model, characterized in that, include: Obtain engine operating parameters and cylinder pressure data; Based on the cylinder pressure data, the target heat release rate curve is determined; Based on the target heat release rate curve, the target model parameters for the tail combustion stage, the main combustion stage, and the pre-combustion stage are determined sequentially. Based on the target model parameters and operating condition parameters at each stage, a prediction parameter mapping model is established so that the engine can determine the prediction model parameters and generate a prediction heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating condition parameters.

2. The method for constructing an engine combustion prediction model according to claim 1, characterized in that, The cylinder pressure data includes crankshaft angle, cylinder pressure varying with the crankshaft angle, and cylinder volume; determining the target heat release rate curve based on the cylinder pressure data includes: The volume change rate is determined based on the change in the cylinder internal volume with the crankshaft angle; The pressure change rate is determined based on the change in cylinder pressure with the crankshaft angle. Based on the preset adiabatic index, the cylinder pressure, the cylinder volume, the volume change rate, and the pressure change rate, the target heat release rate curve is determined through a preset thermodynamic relationship. The preset adiabatic index includes a first preset value for the compression process and a second preset value for the expansion process.

3. The method for constructing an engine combustion prediction model according to claim 1, characterized in that, The step of determining the target model parameters for the tail combustion stage, main combustion stage, and pre-combustion stage sequentially based on the target heat release rate curve includes: Based on the target heat release rate curve, determine the tail combustion characteristic points and tail combustion shape factor of the tail combustion stage; Subtract the first fitted curve based on the tail combustion characteristic points and the tail combustion shape factor from the target heat release rate curve to obtain the first intermediate curve; Based on the first intermediate curve, the main combustion characteristic points of the main combustion stage are determined; Subtract the second fitted curve, which is fitted based on the main combustion feature points and the preset main combustion shape factor, from the first intermediate curve to obtain the second intermediate curve; Based on the second intermediate curve, the pre-combustion characteristic points and pre-combustion shape factor of the pre-combustion stage are determined.

4. The method for constructing an engine combustion prediction model according to claim 3, characterized in that, The determination of the tail combustion characteristic points and tail combustion shape factor based on the target heat release rate curve includes: The tail section of the target heat release rate curve is subjected to second-order derivative processing, and the point where the absolute value of the second derivative is the largest is determined as the tail combustion characteristic point. The first horizontal axis search interval is determined based on the preset tail combustion stage proportion constraint range; Within the first horizontal axis search interval, a first error function is constructed, wherein the first error function is used to characterize the cumulative difference between the target heat release rate curve and the first candidate fitting curve within a preset integral interval from the combustion endpoint to the combustion start point, and the first candidate fitting curve is generated based on the tail combustion characteristic point and the tail combustion shape factor candidate value. By traversing multiple candidate values ​​of the tail combustion shape factor, the candidate value of the tail combustion shape factor that minimizes the value of the first error function is determined as the tail combustion shape factor.

5. The method for constructing an engine combustion prediction model according to claim 3, characterized in that, The step of determining the main combustion characteristic points of the main combustion stage based on the first intermediate curve includes: Obtain the first abscissa corresponding to the preset main combustion shape factor and the peak point of the maximum heat release rate of the first intermediate curve; The interval between the first horizontal coordinate and the second horizontal coordinate of the tail combustion feature point is determined as the second horizontal coordinate search interval; Within the second horizontal axis search interval, a second error function is constructed, wherein the second error function is used to characterize the cumulative difference between the first intermediate curve and the second candidate fitting curve, and the second candidate fitting curve is generated based on the main combustion shape factor and combustion feature candidate points; By traversing multiple combustion feature candidate points within the second horizontal coordinate search interval, the combustion feature candidate point that minimizes the value of the second error function is determined as the main combustion feature point.

6. The method for constructing an engine combustion prediction model according to claim 3, characterized in that, The determination of the pre-combustion characteristic points and pre-combustion shape factor of the pre-combustion stage based on the second intermediate curve includes: The point where the maximum heat release rate peak of the second intermediate curve is located is determined as the pre-combustion characteristic point; The interval between the preset combustion start point and the main combustion characteristic point is determined as the third horizontal coordinate search interval; Within the third horizontal coordinate search interval, a third error function is constructed, wherein the third error function is used to characterize the cumulative difference between the second intermediate curve and the third candidate fitting curve, and the third candidate fitting curve is generated based on the pre-combustion feature point and the candidate value of the main combustion shape factor; By traversing multiple candidate values ​​of the pre-burning shape factor, the candidate value of the pre-burning shape factor that minimizes the value of the third error function is determined as the pre-burning shape factor.

7. The method for constructing an engine combustion prediction model according to claim 3, characterized in that, The establishment of a prediction parameter mapping model based on the target model parameters and the operating condition parameters at each stage includes: The system acquires the operating parameters corresponding to multiple sets of offline operating conditions, as well as the tail combustion characteristic point, the main combustion characteristic point, the pre-combustion characteristic point, the tail combustion shape factor, the main combustion shape factor, and the pre-combustion shape factor corresponding to each set of operating parameters. The operating parameters include at least multiple parameters among engine speed, fuel injection quantity, intake manifold pressure, exhaust manifold pressure, intake manifold temperature, exhaust manifold temperature, and exhaust gas recirculation rate. Based on the operating parameters under the multiple sets of offline operating conditions and the corresponding tail combustion characteristic points, main combustion characteristic points and pre-combustion characteristic points, a first mapping relationship matrix is ​​established; Based on the operating parameters under the multiple sets of offline operating conditions and the corresponding tail combustion shape factor, main combustion shape factor and pre-combustion shape factor, a second mapping relationship matrix is ​​established; The first mapping matrix and the second mapping matrix are combined as the prediction parameter mapping model.

8. The method for constructing an engine combustion prediction model according to claim 7, characterized in that, The step of enabling the engine to determine prediction model parameters and generate a predicted heat release rate curve based on the prediction parameter mapping model and the engine's real-time operating parameters includes: So that the engine inputs the real-time operating parameters into the first mapping matrix to obtain the target tail combustion characteristic point, the target main combustion characteristic point and the target pre-combustion characteristic point; So that the engine inputs the real-time operating parameters into the second mapping matrix to obtain the target tail combustion shape factor, the target main combustion shape factor and the target pre-combustion shape factor; So that the engine generates the predicted heat release rate curve based on the target exhaust combustion characteristic point, the target main combustion characteristic point, the target pre-combustion characteristic point, the target exhaust combustion shape factor, the target main combustion shape factor, and the target pre-combustion shape factor.

9. A method for applying an engine combustion prediction model, characterized in that, include: Based on the prediction parameter mapping model according to any one of claims 1 to 8 and the real-time operating parameters of the engine, the prediction model parameters are determined and the prediction heat release rate curve is generated.

10. An electronic device, comprising: The memory and processor are characterized in that the processor is used to execute a computer program stored in the memory to implement the steps of the engine combustion prediction model construction method as described in any one of claims 1 to 8 or the engine combustion prediction model application method as described in claim 9.

Images