An octane rating method and device with pressure oscillation integral quantity and characteristic fusion

Abstract

The present disclosure provides an octane rating method and device with pressure oscillation integral quantity and characteristic fusion. The method comprises the following steps: using a reference fuel, establishing an IMPO threshold prediction model; for the fuel to be tested, substituting the initial anti-knock performance and the test machine compression ratio into the threshold prediction model to obtain an IMPO prediction threshold, screening valid cycles based on the prediction threshold and outputting a new anti-knock performance prediction result; using the latest anti-knock performance prediction result to obtain an updated prediction threshold, further screening valid cycles based on the updated prediction threshold, and predicting a new anti-knock performance prediction result; returning to perform the updating step until the error of the anti-knock performance prediction result converges to a predetermined threshold range or the stability of the anti-knock performance prediction result reaches a predetermined threshold; and finally predicting the anti-knock performance. The present disclosure can adapt to various fuel types and compression ratio conditions, and improve the prediction accuracy and robustness of the anti-knock performance.

Description

Technical Field

[0001] This disclosure relates to the field of energy and power engineering technology, and in particular to a method and apparatus for determining octane number by integrating pressure oscillation integral and characteristic features. Background Technology

[0002] The anti-knock properties of fuel are one of the key factors affecting the further improvement of the thermal efficiency of spark-ignition engines. As engines develop towards higher compression ratios, higher boost pressures, and higher loads, traditional methods directly limit the development and application of high-performance fuels, and also constrain further improvements in engine thermal efficiency. Therefore, there is an urgent need for a new octane number determination method and device that can adapt to different fuel systems and various engine operating conditions, achieving accurate and reliable measurement of fuel anti-knock properties, thereby truly reflecting the fuel's anti-knock capability in actual engines. Summary of the Invention

[0003] This disclosure provides a method for determining octane number by fusing pressure oscillation integral intensity and characteristics. The method includes the following steps: Calibration step: Using a reference fuel, a pressure oscillation integral intensity threshold prediction model M1 is established; Measurement step: For the fuel to be tested, the initial anti-knock performance of the fuel to be tested and the compression ratio of the testing machine are substituted into the pressure oscillation integral intensity threshold prediction model M1 to obtain a pressure oscillation integral intensity prediction threshold. Based on the prediction threshold, effective cycles are selected, and new anti-knock performance prediction results are output based on the results obtained in the effective cycles; Update step: The latest anti-knock performance prediction results are fed back to the pressure oscillation integral intensity threshold prediction model M1 to obtain an updated prediction threshold. Further, effective cycles are selected based on the updated prediction threshold, and new anti-knock performance prediction results are predicted based on the results obtained in the updated effective cycles; Iteration step: The update step is repeated until one of the following two conditions is met: the error of the obtained anti-knock performance prediction result converges to a predetermined threshold range and the stability of the output anti-knock performance prediction result reaches a predetermined threshold; thereby finally predicting the anti-knock performance.

[0004] In some embodiments, during the calibration step, a test machine is run at different compression ratios to acquire high-frequency pressure signals and calculate the original pressure oscillation integral intensity value within the knock window for each cycle. Based on the correlation between the original pressure oscillation integral intensity value, the anti-knock performance of the reference fuel, and the compression ratio, a pressure oscillation integral intensity threshold prediction model M1 is established.

[0005] In some embodiments, the original pressure oscillation integral intensity value The calculation formula is: P(t) is the pressure oscillation component obtained after filtering the cylinder pressure signal. and These are the starting and ending crankshaft rotation angles (°CA) of the knock window, respectively. This indicates the crankshaft rotation angle increment.

[0006] In some embodiments, the pressure oscillation integral intensity threshold prediction model M1 maps the fuel's octane number and compression ratio to the prediction threshold: The prediction threshold ,in, The standard deviation of the octane number of the fuel is given. The compression ratio of the test machine is given.

[0007] In some embodiments, when predicting the fuel to be tested in the determination step, the condition for the effective screening cycle is: , Where i represents the i-th cycle. The integral intensity value of the pressure oscillation in the i-th cycle. The predicted threshold is the threshold obtained from the pressure oscillation integral intensity threshold prediction model M1 in the measurement step.

[0008] In some embodiments, when updating the prediction of the anti-knock performance of the fuel under test in the update step, the condition for the effective screening cycle is: , Where i represents the i-th cycle. The integral intensity value of the pressure oscillation in the i-th cycle. The updated prediction threshold is the one derived in the update step based on the updated pressure oscillation integral intensity threshold prediction model M1.

[0009] In some embodiments, during the measurement step and the update step, after selecting valid cycles, a multidimensional feature vector is extracted based on the results obtained in each of the valid cycles. The multidimensional feature vector is then input into the blast resistance performance prediction model M2 to obtain a new blast resistance performance prediction result.

[0010] In some embodiments, the multidimensional feature vector It includes at least pressure oscillation integral intensity characteristics and auxiliary knock characteristics, wherein the pressure oscillation integral intensity characteristics include standardized pressure oscillation integral intensity values ​​and their statistics; the auxiliary knock characteristics include two or more of the following: pressure oscillation frequency domain energy ratio, pressure oscillation dominant frequency, knock initiation crankshaft rotation angle, pressure oscillation integral intensity statistical characteristics and their relative characteristics.

[0011] This disclosure also provides an octane number measuring device that integrates pressure oscillations with characteristics, and uses the above method to predict the anti-knock performance of fuel. The octane number measuring device includes an in-cylinder pressure sensor and a data analysis system. The in-cylinder pressure sensor collects combustion pressure signals in real time, and the data analysis system performs signal processing.

[0012] In some embodiments, the device further includes an electronically controlled fuel injection system, which includes a pressure control switch and a pressure gauge disposed on the fuel line and fuel injectors connected to an electronic control unit, so as to realize real-time monitoring and precise adjustment of fuel pressure.

[0013] This disclosure, by employing an adaptive IMPO prediction threshold as the knock criterion, explicitly proposes for the first time that the IMPO threshold should not be considered a fixed constant, but rather a predictable variable that varies with fuel octane number and engine compression ratio. The threshold is adaptively determined through multiple iterations using a modeling approach. By introducing an adaptive knock criterion and a multi-feature fusion mechanism, this disclosure achieves adaptive determination for different fuel systems and octane number ranges, avoiding the limitations of traditional fixed-threshold methods that rely on empirical interpolation and have limited applicability. Therefore, this disclosure can adapt to various fuel types and compression ratios, improving the prediction accuracy and robustness of anti-knock performance. Attached Figure Description

[0014] Figure 1 This is a schematic block diagram of an octane number determination method based on the fusion of pressure oscillation integral and characteristics according to an embodiment of this disclosure.

[0015] Figure 2 This is a schematic block diagram of an octane number measuring device that integrates pressure oscillation integral and characteristic features according to an embodiment of this disclosure.

[0016] Figure 3 This is a schematic structural diagram of an octane number measuring device that integrates pressure oscillation integral and characteristic features according to an embodiment of this disclosure.

[0017] Explanation of reference numerals in the attached figures: 1. High-pressure nitrogen cylinder; 2. High-pressure gas line; 3. Fuel tank; 4. Pressure gauge; 5. Fuel line; 6. Pressure control switch; 7. Fuel injector; 8. Intake pipe; 9. Exhaust pipe; 10. Cooling device; 11. Intake heating jacket; 12. Thermocouple; 13. Electronic control unit (ECU); 14. Octane number tester; 15. Ignition device; 16. Cylinder pressure sensor; 17. Combustion analyzer; 18. Oxygen sensor; 19. Lambda analyzer; 20. Data analysis system; 21. Engine combustion chamber; 22. Intake valve; and 23. Cylinder structure. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions of this disclosure will be described in detail below with reference to the accompanying drawings.

[0019] Exemplary embodiments will be described more fully below with reference to the accompanying drawings; however, these exemplary embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will enable those skilled in the art to fully understand the scope of this disclosure.

[0020] Where there is no conflict, the various embodiments of this disclosure and the features thereof in the embodiments may be combined with each other.

[0021] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0022] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. As used herein, the singular forms “a” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded.

[0023] The embodiments described herein can be described with reference to plan views and / or cross-sectional views using the ideal schematic diagrams of this disclosure. Therefore, the example illustrations can be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to those shown in the drawings, but include modifications to configurations formed based on manufacturing processes. Therefore, the areas illustrated in the drawings are schematic in nature, and the shapes of the areas shown in the figures illustrate specific shapes of areas of an element, but are not intended to be limiting.

[0024] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this disclosure, and will not be interpreted as having an idealized or overly formal meaning, unless expressly so defined herein.

[0025] Technical Terminology Explanation Unless otherwise specified in this disclosure, the following technical terms shall be interpreted as follows: "Knock," also known as engine knock or combustion knocking, is a malfunction caused by abnormal combustion in an engine. Knock typically occurs when, after the fuel-air mixture ignites in the combustion chamber, the flame has not fully propagated, and unburned fuel-air mixture at a distance spontaneously combusts due to high temperature or pressure. The resulting flame collides with the normally burning flame, generating immense pressure. This phenomenon can easily damage the engine.

[0026] The "knock window" refers to the crankshaft rotation angle range (window) after the compression top dead center where knocking occurs. Knock is determined to have occurred only when the vibration intensity of the cylinder block exceeds the limit within this window.

[0027] Compression ratio is the ratio of the total volume of the engine cylinders to the volume of the combustion chamber. It indicates the degree to which the gas inside the cylinder is compressed as the piston moves from bottom dead center to top dead center. A higher compression ratio results in higher pressure and temperature of the gas inside the cylinder at the end of compression.

[0028] "Anti-knock performance" refers to the characteristic of fuel to resist knocking during engine combustion, and is a key indicator for measuring fuel combustion efficiency.

[0029] Octane rating is a core indicator for measuring fuel's anti-knock performance; a higher number indicates more stable combustion of the fuel in the engine. Engines with high compression ratios have higher combustion chamber pressures, making them more prone to knocking if fuel with low anti-knock properties is used.

[0030] "Reference fuel" is a key substance used for standardized evaluation of fuel performance, especially in the petroleum product sector, where it serves as a calibration reference to quantify key indicators such as octane number and cetane number.

[0031] The "fuel equivalence ratio" (φ) is the ratio of the proportion of fuel to air in an actual fuel-air mixture to the theoretical stoichiometric ratio. It is used to measure the degree to which combustion deviates from complete combustion.

[0032] The Electronic Control Unit (ECU) is primarily responsible for the engine's operation and control. It collects engine status information through sensors such as oxygen sensors and knock sensors, and precisely controls fuel injection quantity, ignition advance angle, and idle speed to ensure the engine's power output and fuel economy under different operating conditions.

[0033] A "Lambda analyzer" is an instrument used to accurately measure the air-fuel ratio (AFR) lambda value (λ, the ratio of the actual air-fuel ratio to the stoichiometric ratio) in a gas mixture.

[0034] The "knock sensor" is used to detect whether knocking occurs during engine operation and transmits the knock signal to the engine ECU. The ECU issues a command based on this signal to control the on / off state of the primary circuit of the ignition coil and adjust the ignition timing to prevent knocking from occurring.

[0035] The "detonation criterion" is mainly based on the physical phenomena and measurable parameters generated by detonation, and is used to identify and quantify the degree of detonation.

[0036] "Knock intensity" is an indicator for judging the severity of knocking, including, for example, the integral of the modulus of the pressure oscillation (IMPO) and the maximum amplitude of pressure oscillation (MAPO).

[0037] In this field, when predicting the anti-knock properties of fuels, the current ASTM standard method relies on an octane number tester equipped with a conventional knock sensor to determine the octane number by comparing the knock intensity with a reference fuel.

[0038] Existing research and patents on fuel octane rating determination can be broadly categorized into three technical approaches. The first is based on high-frequency in-cylinder pressure signals, using knock intensity indicators such as MAPO or IMPO, combined with artificially set fixed thresholds to establish a correlation between octane rating and the engine's performance. The second is based on knock sensors or characteristic compression ratios, using operating parameters when the engine enters a specific knock state to infer the octane rating. The third approach completely departs from the actual engine knock process, utilizing auto-ignition or pressure derivative characteristics in a constant-volume combustion chamber to predict the research octane rating. While these three methods differ in experimental techniques and signal formats, they share a common core assumption: treating knock criteria or key characteristics as fixed, one-time set parameters.

[0039] Taking the MAPO / IMPO method based on high-frequency cylinder pressure as an example, existing studies typically determine an empirical knock threshold through numerous experiments within a limited octane number range and under specific operating conditions, assuming that this threshold is universally applicable under different fuels and operating conditions. This approach can be considered an effective approximation within the traditional gasoline octane number range. However, when the fuel octane number increases significantly or the engine compression ratio changes, the preconditions implied by the fixed threshold no longer hold, easily leading to the failure of the knock criterion and thus causing systematic biases in the anti-knock performance prediction results. From a methodological perspective, this type of method essentially relies on the mapping relationship of existing calibration data within a limited range, and is closer to an empirical interpolation-type measurement method than a prediction method with adaptive or extrapolation capabilities.

[0040] Patented methods based on knock sensors or characteristic compression ratios also suffer from similar problems. These methods typically rely on a specific knock intensity level or a characteristic operating point to establish an octane number mapping relationship. However, due to the limited ability of knock sensors to capture high-frequency pressure oscillation information, their criteria reflect more the macroscopic knock trend than the actual anti-knock behavior details of the fuel. This means that such methods also need to be recalibrated when the fuel type changes, making it difficult to achieve universal application across fuel systems.

[0041] As for the method of predicting octane number using a constant-volume combustion chamber, its starting point is not the engine knock mechanism, but rather the auto-ignition and chemical reaction kinetics of the fuel. This approach is significant in basic research, but due to the lack of actual engine knock characteristics such as compression ratio changes, flame propagation, and end-gas auto-ignition, its prediction results cannot directly reflect the anti-knock performance of the fuel in a real spark-ignition engine. Therefore, it differs fundamentally from engine knock measurement methods in terms of engineering applicability.

[0042] The limitations of the aforementioned traditional testing systems have become increasingly apparent as engines have evolved towards higher compression ratios, higher boost pressures, and higher loads: the low-pass filter of the knock sensor severely weakens the high-frequency pressure oscillation signal, making it impossible to accurately reflect the knock behavior under modern enhanced engines; the fuel supply method lags behind modern electronically controlled injection systems, resulting in a serious disconnect between test results and actual engine operating conditions; for high-octane fuels, traditional methods rely on harmful additives such as tetraethyl lead, which endanger human health; a single knock index or fixed criterion cannot adapt to various fuel types and compression ratios, ignoring multi-dimensional information such as amplitude, frequency domain, and phase of the knock process, resulting in low accuracy and poor robustness in anti-knock performance prediction.

[0043] Figure 1 This is a schematic block diagram of an octane number determination method based on the fusion of pressure oscillation integral and characteristics according to an embodiment of this disclosure.

[0044] Firstly, referring to Figure 1This disclosure provides a method for determining octane number by fusing pressure oscillation integral intensity and characteristics. The method includes the following steps: Calibration step S102: Using a reference fuel, establish a pressure oscillation integral intensity threshold prediction model M1; Measurement step S104: For the fuel to be tested, substitute the initial anti-knock performance of the fuel to be tested and the compression ratio of the testing machine into the pressure oscillation integral intensity threshold prediction model M1 to obtain the pressure oscillation integral intensity prediction threshold, filter effective cycles based on the prediction threshold, and output a new anti-knock performance prediction result based on the results obtained in the effective cycles; Update step S106: Feed back the latest anti-knock performance prediction result to the pressure oscillation integral intensity threshold prediction model M1 to obtain an updated prediction threshold, and further filter effective cycles based on the updated prediction threshold to predict a new anti-knock performance prediction result based on the results obtained in the updated effective cycles; and Iteration step S108: Return to execute the update step until one of the following two conditions is met: the error of the obtained anti-knock performance prediction result converges to a predetermined threshold range and the stability of the output anti-knock performance prediction result reaches a predetermined threshold; thereby finally predicting the anti-knock performance.

[0045] Existing technologies for determining the anti-knock performance of fuels all employ fixed knock criteria. This disclosure, for the first time, explicitly proposes that the IMPO threshold should not be considered a fixed constant, but rather a predictable variable that varies with fuel octane number and engine compression ratio. This threshold is adaptively determined through multiple iterations of a model. By introducing an adaptive knock criterion and a multi-feature fusion mechanism, this disclosure achieves adaptive determination for different fuel systems and octane number ranges, avoiding the limitations of traditional fixed-threshold methods that rely on empirical interpolation and have limited applicability. Therefore, this disclosure can adapt to various fuel types and compression ratios, improving the prediction accuracy and robustness of anti-knock performance.

[0046] In this disclosure, the reference fuel is a reference fuel with known anti-knock properties, and the fuel to be tested may be a fuel with anti-knock properties higher than the reference fuel range or an oxygen-containing additive fuel.

[0047] In some embodiments, during calibration step S102, the test machine is run at different compression ratios to acquire high-frequency pressure signals and calculate the original pressure oscillation integral intensity value within the knock window for each cycle. Therefore, based on the original pressure oscillation integral intensity value A pressure oscillation integral intensity threshold prediction model M1 was established to correlate the anti-knock performance of the reference fuel with the compression ratio of the test machine.

[0048] In some embodiments, the original pressure oscillation integral intensity value The calculation formula is: (1) P(t) is the pressure oscillation component obtained after filtering the cylinder pressure signal. and These represent the start and end crankshaft rotation angles (°CA) of the knock window, respectively. This indicates the crankshaft rotation angle increment.

[0049] In some embodiments, for different fuels and compression ratios Statistical analysis of the distribution was performed, and statistical characteristics were calculated for significant detonation cycles to obtain the effective IMPO threshold. .

[0050] (2) in, and These are the integral intensity values ​​of pressure oscillations in significant detonation cycles ( The mean and standard deviation are given, and k is an empirical coefficient.

[0051] In some embodiments of this disclosure, the IMPO threshold prediction model M1 is based on the effective IMPO threshold obtained above. The anti-knock properties of the reference fuel and the compression ratio of the test machine are used to establish the system.

[0052] Specifically, the IMPO threshold prediction model M1 can be constructed using various methods, such as multiple linear regression, support vector machine (SVM), and neural network.

[0053] Multiple linear regression uses a linear model to fit the relationship between the standard deviation of octane number and compression ratio and the IMPO threshold. This model assumes that the IMPO threshold is a linear combination of the standard deviation of octane number and compression ratio, and is suitable for cases where there is a linear relationship between the input features.

[0054] SVM is a supervised learning model used for regression and classification tasks. In this case, SVM maps octane number and compression ratio to a threshold prediction space, attempting to find an optimal hyperplane in a high-dimensional space to minimize prediction error. SVM has a stronger ability to model nonlinear relationships, thus handling the complex relationships between octane number standard deviation and compression ratio with the IMPO threshold.

[0055] Neural networks are models capable of learning complex nonlinear relationships. Through multi-layered neural networks, complex patterns can be learned from octane ratings and compression ratios, and the IMPO threshold can be predicted. Neural networks excel when dealing with highly complex feature relationships.

[0056] Those skilled in the art can choose appropriate methods to construct the IMPO threshold prediction model M1 as needed.

[0057] In some embodiments, the constructed IMPO threshold prediction model M1 maps the fuel's octane number and compression ratio to a prediction threshold: Predicted threshold ,in, The standard deviation of the octane number of the fuel. This is the compression ratio of the test machine.

[0058] For example, the following regression model can be used to construct the IMPO threshold prediction model M1.

[0059] (3) in, The intercept term of the regression model represents the standard deviation of the octane number ( ) and compression ratio ( The IMPO prediction threshold when ) is 0 The baseline value. It is a constant term in the model used to adjust the base level of predictions; The coefficients of the regression model represent the standard deviation of the octane number ( ). The extent to which this affects the IMPO prediction threshold. Specifically, It describes the contribution of octane number variation to IMPO (i.e., knock sensitivity) prediction; the greater the fluctuation in octane number, the greater the possible change in the predicted IMPO value. The coefficients of the regression model represent the compression ratio ( The degree of influence of compression ratio on the IMPO prediction threshold. Compression ratio has a direct impact on the sensitivity to detonation, therefore, This paper describes how the knock response of fuel at different compression ratios affects the prediction of IMPO.

[0060] In some embodiments, fuel knock response behavior is induced by progressively adjusting the engine compression ratio. This allows for a more realistic characterization of the fuel's anti-knock performance under actual engine operating conditions.

[0061] For example, the compression ratio can be changed by using a compression ratio adjustment mechanism (not shown, which is used to adjust the distance between the engine cylinder head and the cylinder block), and the compression ratio can be gradually increased by using this compression ratio adjustment mechanism. Tests can be conducted at different compression ratios to explore the knock response of the fuel.

[0062] In this disclosure, the IMPO threshold prediction model M1 constructed based on the above method is used to predict the IMPO threshold based on the fuel octane number and compression ratio, that is, under specific engine operating conditions, to predict the IMPO threshold required for an "effective cycle" or "significant knock" through these characteristics.

[0063] In some embodiments, when predicting the fuel to be tested in step S104, the criteria for screening valid cycles are: (4) Where i represents the i-th cycle. The IMPO value for the i-th cycle. To determine the predicted threshold obtained from the IMPO threshold prediction model M1 in step S104.

[0064] This serves as the initial IMPO prediction threshold, used to distinguish between valid and invalid cycles in the initial cycle set. A cycle is considered a "valid cycle" only if its IMPO value exceeds this threshold. This threshold is typically obtained through baseline calibration or preliminary experiments and is used to identify significant cycles associated with detonation.

[0065] In this disclosure, The calculation can refer to the original pressure oscillation integral intensity value. The calculation formula is obtained.

[0066] In some embodiments, when updating the prediction of the antiknock performance of the fuel to be tested in update step S106, the criteria for screening valid cycles are: (5) Where i represents the i-th cycle. The IMPO value for the i-th cycle. To update the predicted threshold obtained in step S106 based on the updated IMPO threshold prediction model M1.

[0067] It can be understood that in this disclosure, Compression ratio (CR) is used as input features and passed to the IMPO threshold prediction model M1. The IMPO threshold prediction model M1 is trained based on known relationships in the training data, learning how octane number standard deviation and compression ratio affect the IMPO threshold. After training is complete, the trained model is used with new input... and The predicted IMPO threshold is obtained. Set the IMPO threshold. Used to determine valid cycles, filtering out those with IMPO values ​​greater than the IMPO threshold. The cycle.

[0068] In some embodiments, in the determination step S104 and the update step S106, after selecting valid cycles, a multidimensional feature vector is extracted based on the results obtained in each valid cycle. and multidimensional feature vectors Input the data into the blast resistance performance prediction model M2, and obtain new blast resistance performance prediction results.

[0069] This disclosure employs a multi-feature fusion method, integrating IMPO features and other knock-related features (such as pressure oscillation frequency domain features, knock initiation crankshaft angle, etc.) to reduce the randomness and instability caused by single features; it avoids predicting anti-knock performance based solely on a single pressure amplitude, thereby improving the stability and generalization ability of the prediction results; it is applicable to high-octane fuels and oxygenated additive fuel systems, and has strong engineering applicability.

[0070] Furthermore, this disclosure goes beyond this. It does not simply introduce a new knock index into existing methods, but fundamentally reconstructs the role of the knock criterion at the methodological level. Specifically, this application abandons the traditional approach of directly mapping octane number to a single knock index through multi-dimensional feature vectors. Instead, it comprehensively characterizes the knock response behavior of fuel in the engine by integrating multi-dimensional information (such as knock energy, frequency domain characteristics, and phase characteristics). More importantly, it constructs an iterative feedback mechanism between an adaptive threshold model and an anti-knock performance prediction model, enabling the prediction results to correct the knock criterion in reverse, achieving self-consistent optimization of the measurement process. This "dual-model + feedback" measurement framework is unprecedented in existing technologies, proposing a new testing method at the system level.

[0071] Therefore, this disclosure is not a simple improvement on existing MAPO / IMPO methods or constant-volume combustion chamber methods, but rather, by introducing an adaptive knock criterion and a multi-feature fusion mechanism, it achieves adaptive determination for different fuel systems and different octane number ranges, avoiding various limitations of traditional methods. A new octane number determination method that more closely approximates the real engine knock mechanism and possesses cross-fuel and cross-operating-condition adaptability has been established.

[0072] In some embodiments, multidimensional feature vectors It includes at least IMPO features and auxiliary knock features. The IMPO features include the standardized pressure oscillation integral intensity value and its statistics. The auxiliary knock features include two or more of the following: pressure oscillation frequency domain energy ratio, pressure oscillation dominant frequency, knock initiation crankshaft rotation angle, IMPO statistical features and their relative features.

[0073] In some embodiments, multidimensional feature vectors Not limited to the features mentioned above, it may also include features such as vectors. In addition to the parameters already listed, other parameters may include: the standard deviation of IMPO (which measures the magnitude of fluctuations in IMPO values), the mean of IMPO (which represents the average level of detonation intensity), the duration of detonation (which measures the persistence of detonation), and the inter-cycle variance (COV) (which describes the change in detonation in each cycle).

[0074] Those skilled in the art can select parameter combinations to evaluate performance based on the specific needs of the situation.

[0075] For example, in some embodiments, a combination of IMPO normalized values, detonation initiation angle, and maximum pressure rise rate can be selected to provide detonation intensity, timing of occurrence, and pressure rise rate to comprehensively assess the severity of detonation.

[0076] For example, in some embodiments, the IMPO peak ratio, IMPO standard deviation, and spectral entropy can be selected and combined with a combination of extreme intensity, volatility, and frequency characteristics to gain a more comprehensive understanding of the complexity of detonation.

[0077] In some embodiments, multidimensional feature vectors For example, constructed as (6) in, The standardized value of IMPO is represented as (7), Where N represents the number of iterations, The initial IMPO prediction threshold, This represents the IMPO value of the i-th cycle; (8); The crankshaft rotation angle at the onset of knock. The main frequency of pressure oscillation. This represents the peak ratio of IMPO.

[0078] IMPO peak ratio It refers to the ratio of the maximum peak value of IMPO (i.e. the maximum amplitude of the detonation signal) to a certain reference value. It usually reflects the difference in detonation intensity relative to the standard level. A higher peak value ratio indicates that the detonation intensity is greater.

[0079] Common benchmark values ​​include, for example: (Initial IMPO prediction threshold), the average or standardized value of IMPO. For example, the standardized value of IMPO can be used as a benchmark to calculate the peak-to-peak ratio.

[0080] The blast resistance performance prediction model M2 is used to extract multidimensional feature vectors from the IMPO calculation. Mapped to fuel octane number ( That is, predicting the octane number of fuel based on IMPO value and other knock-related characteristics: (9).

[0081] For example, partial least squares regression (PLSR) is used in some embodiments to predict octane numbers: (10).

[0082] In this example method, the octane number is predicted ( ) is calculated using a weighted sum, where, These are the weight coefficients in the regression model, representing the weights of each feature. The extent of its contribution to the final octane number prediction It is associated with feature vectors (such as the standardized value of IMPO, the statistics of IMPO, etc.) and reflects the influence of each feature on the final octane number prediction. This is the input feature vector, containing multiple features extracted from the IMPO analysis, such as the IMPO standardized value, IMPO statistics, and knock initiation crankshaft angle. ), the main frequency of pressure oscillation ( )wait; It is the bias term (or intercept term) in the regression model, which represents the baseline predicted octane number when all features are zero. It is used to adjust the predicted final octane number to make it more consistent with or better match the actual observation data.

[0083] In some embodiments of this disclosure, for example, a partial least squares regression (PLSR) model is used to learn the relationship between the feature vector and the fuel octane number to find the optimal weighting coefficients. ) and bias terms ( The new IMPO feature vector ( The input is fed into the trained model, and the output is the predicted octane number. ).

[0084] In some embodiments, in iterative step S108, the first prediction is... Feedback to model M1 to update the threshold: The feature vectors are re-extracted and the octane number is predicted to improve the stability of the determination.

[0085] In some embodiments, in iteration step S108, update step 3 to 5 are repeated.

[0086] This disclosure constructs a comprehensive knock response index based on the extracted multidimensional knock characteristics and their statistical properties, and classifies the knock sensitivity of fuel according to the numerical distribution characteristics of the index to form a fuel anti-knock performance criterion; further, it maps the criterion to a preset octane number correspondence to output the corresponding octane number or its numerical range for the fuel.

[0087] Compared with existing technologies, this disclosure has the following advantages: by progressively adjusting the engine compression ratio to excite fuel knock response behavior, it more realistically characterizes the anti-knock performance of fuel under actual engine operating conditions; by adopting a multi-feature fusion method, it integrates IMPO features and other knock-related features (such as pressure oscillation frequency domain features, knock initiation crankshaft angle, etc.) to reduce the randomness and instability caused by single features; it avoids predicting octane number based solely on a single pressure amplitude, thereby improving the stability and generalization ability of the prediction results; it is applicable to high-octane fuels and oxygenated additive fuel systems, and has strong engineering applicability.

[0088] More specifically, the engine employs a variable compression ratio structure, acquiring in-cylinder combustion pressure signals under different operating conditions by progressively increasing the compression ratio; and calculating for each cycle... And select effective loops to construct multidimensional feature vectors. This includes standardized IMPO values, IMPO statistics, and knock initiation crankshaft angle. Pressure oscillation main frequency The octane number is predicted using a multi-feature fusion model M2, which includes features such as IMPO peak ratio. The initial prediction result is fed back to model M1 to update the adaptive IMPO prediction threshold, and features are re-extracted for iterative optimization, thereby further improving the accuracy and stability of the anti-knock performance measurement. In this embodiment, the maximum amplitude of the knock pressure is only used as part of the multi-dimensional feature input, not the sole criterion, thus ensuring the reliability and generalization ability of the octane number prediction results.

[0089] This disclosure can be widely applied to fuel anti-knock performance evaluation, rapid octane number prediction, and the research and development of high-octane fuels and oxygenated additive fuels; it can be used for engine combustion optimization, ignition strategy adjustment, compression ratio matching, and high-performance engine tuning; it can be combined with engine simulation platforms, online monitoring systems, and fuel quality management platforms to achieve fuel-engine matching analysis, real-time combustion control, and emission optimization; it can be extended to the research, screening, and evaluation of alternative fuels, high-performance gasoline, aviation gasoline, and multi-fuel engines, as well as automated fuel research and development platforms, high-throughput fuel screening, intelligent vehicle fuel adaptation, and emission control, providing high-precision and stable technical support and decision-making basis for fuel research and development, engine design, and operation optimization.

[0090] Figure 2This is a schematic block diagram of an octane number measuring device that integrates pressure oscillation integral and characteristic features according to an embodiment of this disclosure.

[0091] Secondly, referring to Figure 2 This disclosure provides an octane number measuring device 100, which uses the above method to predict the anti-knock performance of fuel. The octane number measuring device 100 includes an in-cylinder pressure sensor 110 and a data analysis system 120.

[0092] In some embodiments, the in-cylinder pressure sensor 110 and the data analysis system 120 support at least 0.1° CA resolution and perform functions such as threshold prediction using model M1, extraction of multiple feature vectors, and anti-knock performance prediction using model M2.

[0093] The octane number measuring device 100 disclosed herein executes the above-mentioned octane number measuring method by employing an in-cylinder pressure sensor 110 and a data analysis system 120, thereby realizing anti-knock performance prediction based on adaptive IMPO threshold and multi-feature fusion. This solves the problem of insufficient prediction accuracy of the fixed IMPO threshold method in the prior art under different fuel types and different compression ratios, and achieves high-precision prediction of fuel octane number, thereby more realistically evaluating the anti-knock performance of fuel in actual engines.

[0094] In some embodiments, the octane number measuring device 100 of this disclosure may further include: an electronically controlled fuel injection system, which includes a pressure control switch 6 and a pressure gauge 4 disposed on the fuel line 5 and a fuel injector 7 connected to an electronic control unit (ECU) 13, so as to realize real-time monitoring and precise adjustment of fuel pressure.

[0095] Figure 3 This is a schematic structural diagram of the octane number measuring device 100 according to an embodiment of the present disclosure.

[0096] The octane number measuring device 100 disclosed herein includes a high-pressure nitrogen cylinder 1, a high-pressure gas line 2, a fuel tank 3, a pressure gauge 4, a fuel line 5, a pressure control switch 6, a fuel injector 7, an intake pipe 8, an exhaust pipe 9, a cooling device 10, an intake heating jacket 11, a thermocouple 12, an electronic control unit (ECU) 13, an octane number tester 14, an igniter 15, an in-cylinder pressure sensor 16, a combustion analyzer 17, an oxygen sensor 18, a Lambda analyzer 19, and a data analysis system 20.

[0097] A high-pressure nitrogen cylinder 1 is connected to a fuel tank 3 via a high-pressure gas line 2, providing stable fuel pressure. Fuel is delivered to the fuel injector 7 via a pressure control switch 6, a pressure gauge 4, and a fuel line 5, enabling real-time monitoring and precise adjustment of fuel pressure. The fuel injector 7 is positioned upstream of the intake manifold 8, ensuring that the fuel is fully atomized and evenly mixed with air before entering the combustion chamber. An intake heating jacket 11 is installed outside the intake manifold 8, and a thermocouple 12 is installed inside the intake manifold 8 to monitor and regulate the intake air temperature, and to coordinate with the electronic control unit (ECU) 13 and the data analysis system 20 for control.

[0098] The octane rating tester 14 includes an engine combustion chamber 21, intake valves 22, and cylinder structure 23. An igniter 15 and an in-cylinder pressure sensor 16 are mounted on the cylinder head, respectively connected to an electronic control unit (ECU) 13 and a data analysis system 20, for ignition control and in-cylinder combustion pressure signal acquisition. An oxygen sensor 18 is mounted on the exhaust pipe 9, connected to a Lambda analyzer 19, the ECU 13, and the data analysis system 20, for measuring the post-combustion oxygen content and calculating the fuel equivalence ratio.

[0099] To ensure experimental accuracy, pressure gauge 4 and pressure control switch 6 can be configured as two or more, enabling multi-point monitoring and precise control of fuel pressure. A cooling device 10 is installed on the fuel injector 7 to reduce the impact of temperature fluctuations on injection characteristics. The fuel injector 7, in-cylinder pressure sensor 16, Lambda analyzer 17, and igniter 15 are all controlled by an electronic control unit (ECU) 13, achieving coordinated adjustment of fuel injection, ignition timing, and equivalence ratio.

[0100] In this disclosure, a high-pressure nitrogen cylinder 1 applies a stable pressure to the fuel tank 3, ensuring that the fuel is fully atomized and mixed with air in the intake manifold 8 before entering the combustion chamber. An in-cylinder pressure sensor 16 acquires combustion pressure signals in real time. A combustion analyzer 17 and a data analysis system 20 process the signals, focusing on extracting the integral pressure oscillation intensity (IMPO) and its statistical characteristics. Combined with other knock-related parameters, the knock sensitivity of the fuel at different compression ratios is quantified. The fuel equivalence ratio is measured by an oxygen sensor 18 and a Lambda analyzer 19 on the exhaust manifold 9 and is maintained stable by closed-loop control of the electronic control unit (ECU) 13.

[0101] In this disclosure, the octane rating tester 14 has electronic fuel injection functionality. The octane rating measuring device 100 of this disclosure also includes an input / output interface (not shown) for inputting fuel information and displaying octane rating prediction results.

[0102] The cylinder pressure sensor 16, electronic control unit (ECU) 13, combustion analyzer 17, and data analysis system 20 of the octane number measuring device 100 disclosed herein are used to control the compression ratio, fuel injection, and ignition, and to process pressure signals to extract knock-related amplitude characteristics and construct knock response behavior descriptors.

[0103] In some embodiments, the octane rating measuring device 100 of this disclosure can be obtained by modifying a conventional octane rating tester; the modification includes: replacing a conventional knock sensor with an in-cylinder pressure sensor 16; replacing the carburetor fuel supply system with an electronically controlled fuel injection system, the electronically controlled fuel injection system including a pressure control switch 6 and a pressure gauge 4 installed on the fuel line 5 and a fuel injector 7 connected to an electronic control unit (ECU) 13 to achieve real-time monitoring and precise adjustment of fuel pressure; optimizing the intake system structure, wherein the fuel injector 7 is arranged upstream of the intake manifold 8 to ensure that the fuel is fully atomized and uniformly mixed with air before entering the combustion chamber, and an intake heating jacket 11 is installed outside the intake manifold 8 for monitoring and adjusting the intake temperature; and installing an analysis device for measuring the fuel equivalence ratio in the exhaust system, installing an oxygen sensor 18 on the exhaust manifold 9, and connecting the oxygen sensor 18 to a Lambda analyzer 19, an electronic control unit (ECU) 13 and a data analysis system 20 for measuring the oxygen content after combustion and calculating the fuel equivalence ratio.

[0104] In some embodiments, the octane number measuring device 100 of this disclosure is applicable to fuels with an octane number higher than 100 and oxygenated additive fuels.

[0105] This disclosure, through the aforementioned apparatus and method, enables the adjustment of the compression ratio and analysis of the IMPO response law of fuel under different compression ratio conditions, while keeping other engine operating conditions essentially unchanged. It also enables the construction of a fuel knock sensitivity criterion and, based on this criterion, the realization of high-precision prediction of fuel anti-knock performance, thereby more realistically and comprehensively reflecting the fuel's anti-knock performance in actual engines.

[0106] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in connection with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in connection with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of this disclosure as set forth by the appended claims.

Claims

1. A method for determining octane number by fusing pressure oscillation integral and characteristic features, characterized in that, The method includes the following steps: Calibration steps: Using reference fuel, establish a pressure oscillation integral intensity threshold prediction model M1; Measurement steps: For the fuel to be tested, the initial anti-knock performance of the fuel to be tested and the compression ratio of the test machine are substituted into the pressure oscillation integral intensity threshold prediction model M1 to obtain the pressure oscillation integral intensity prediction threshold. Based on the prediction threshold, effective cycles are screened, and new anti-knock performance prediction results are output based on the results obtained in the effective cycles. Update steps: Feed the latest blast resistance performance prediction results back to the pressure oscillation integral intensity threshold prediction model M1 to obtain the updated prediction threshold, and further filter the effective loops based on the updated prediction thresholds, and predict the new blast resistance performance prediction results based on the results obtained in the updated effective loops. Iteration steps: Return to execute the update step until one of the following two conditions is met: the error of the obtained explosion-proof performance prediction result converges to a predetermined threshold range and the stability of the output explosion-proof performance prediction result reaches a predetermined threshold; thereby finally predicting the explosion-proof performance.

2. The method for determining octane number according to claim 1, characterized in that, In the calibration step, The test machine was run at different compression ratios to collect high-frequency pressure signals and calculate the original pressure oscillation integral intensity value within the knock window for each cycle. The pressure oscillation integral intensity threshold prediction model M1 is established based on the correlation between the original pressure oscillation integral intensity value, the anti-knock performance of the reference fuel, and the compression ratio.

3. The method for determining octane number according to claim 2, characterized in that, The original pressure oscillation integral intensity value The calculation formula is: P(t) is the pressure oscillation component obtained after filtering the cylinder pressure signal. and These are the starting and ending crankshaft rotation angles (°CA) of the knock window, respectively. This indicates the crankshaft rotation angle increment.

4. The method for determining octane number according to claim 1, characterized in that, The pressure oscillation integral intensity threshold prediction model M1 maps the fuel's octane number and compression ratio to the prediction threshold: The prediction threshold ,in, The standard deviation of the octane number of the fuel is given. The compression ratio of the test machine is denoted as .

5. The method for determining octane number according to claim 1, characterized in that, When predicting the fuel to be tested in the determination step, the conditions for the effective screening cycle are as follows: ， Where i represents the i-th cycle. The integral intensity value of the pressure oscillation in the i-th cycle. The predicted threshold is the threshold obtained from the pressure oscillation integral intensity threshold prediction model M1 in the measurement step.

6. The method for determining octane number according to claim 1, characterized in that, When updating the prediction of the anti-knock performance of the fuel under test in the update step, the condition for the effective screening cycle is: ， Where i represents the i-th cycle. The integral intensity value of the pressure oscillation in the i-th cycle. The updated prediction threshold is the one derived in the update step based on the updated pressure oscillation integral intensity threshold prediction model M1.

7. The method for determining octane number according to claim 1, characterized in that, In the measurement step and the update step, after selecting the effective cycles, a multidimensional feature vector is extracted based on the results obtained in each of the effective cycles. The multidimensional feature vector is then input into the blast resistance performance prediction model M2 to obtain a new blast resistance performance prediction result.

8. The method for determining octane number according to claim 7, characterized in that, The multidimensional feature vector It includes at least pressure oscillation integral intensity characteristics and auxiliary knock characteristics, wherein the pressure oscillation integral intensity characteristics include standardized pressure oscillation integral intensity values ​​and their statistics; the auxiliary knock characteristics include two or more of the following: pressure oscillation frequency domain energy ratio, pressure oscillation dominant frequency, knock initiation crankshaft rotation angle, pressure oscillation integral intensity statistical characteristics and their relative characteristics.

9. An octane number determination device that integrates pressure oscillation integral and characteristics, which uses the octane number determination method according to any one of claims 1 to 8 to predict the anti-knock performance of fuel, characterized in that, The octane number measuring device includes an in-cylinder pressure sensor and a data analysis system. The in-cylinder pressure sensor collects combustion pressure signals in real time, and the data analysis system performs signal processing.

10. The octane number measuring device according to claim 9, characterized in that, The device also includes an electronically controlled fuel injection system, which includes a pressure control switch and a pressure gauge installed on the fuel line, as well as fuel injectors connected to the electronic control unit, to achieve real-time monitoring and precise adjustment of fuel pressure.

Images