Hydrogen engine NOx emission prediction method and system based on air-fuel ratio dynamic correction

Abstract

The invention discloses a hydrogen engine NOx emission prediction method based on air-fuel ratio dynamic correction, which relates to post-processing and comprises the following steps: acquiring operating parameters of a hydrogen engine, and acquiring an original NOx emission model value according to the operating parameters; a real-time air-fuel ratio is obtained, and an air-fuel ratio correction factor is calculated through an interval correction strategy according to the difference between the real-time air-fuel ratio and a set reference air-fuel ratio; and according to the air-fuel ratio correction factor, correcting the original emission NOx model value to obtain the original emission NOx after the air-fuel ratio is corrected. The invention further discloses a hydrogen engine NOx emission prediction system based on air-fuel ratio dynamic correction. According to the method, the NOx prediction precision under the transient working condition is remarkably improved. And compared with the mode only using the steady-state original NOx emission basic model, the error of the original NOx emission is reduced by more than 60% after the air-fuel ratio and the delay are corrected.

Description

TECHNICAL FIELD

[0001] The present application relates to post-processing, more particularly, it relates to a hydrogen engine NOx emission prediction method and system based on air-fuel ratio dynamic correction. BACKGROUND

[0002] Under the background of policy-driven realization of the "double carbon" goal and global energy transformation, the use of hydrogen energy and the development of hydrogen internal combustion engines play an important and positive role in improving environmental benefits, reducing energy dependence on foreign countries, and promoting industrial technology upgrading. Because hydrogen is clean and carbon-free, compared with traditional fuel engines, hydrogen internal combustion engines do not generate a large amount of carbon dioxide and hydrocarbon combustion byproducts (carbon monoxide, hydrocarbons, particulate matter PM) during combustion, but still produce nitrogen oxides (NOx). In the era of increasingly stringent emission regulations for internal combustion engine vehicles (such as the latest Euro VII standard of the European Union and China's National Seven Emission Plan), the design of hydrogen internal combustion engines also needs to develop efficient emission control systems.

[0003] Due to the inherent characteristics of hydrogen fuel, hydrogen internal combustion engine development generally adopts ultra-lean combustion technology. NOx removal technology in an oxygen-rich environment requires the introduction of additional reducing agents (such as H2 in H2-SCR and NH3 in NH3-SCR) from the outside world. In order to precisely control the amount of reducing agent added to precisely control the tail NOx emissions, it is necessary to determine the original NOx concentration produced by internal combustion engine combustion.

[0004] The conventional method for determining NOx concentrations in current aftertreatment systems is through NOx sensor readings. NOx sensors offer high-precision, real-time monitoring, but their application in hydrogen internal combustion engine aftertreatment systems presents several challenges, including high sensor cost, response delays caused by the need to warm up the components to operating temperature during cold starts, and sensor stability and limited service life in high-water environments. Compared to diesel and natural gas, hydrogen has extremely rapid diffusion and combustion rates, resulting in a more concentrated combustion process. The air-fuel ratio during combustion directly affects the combustion chamber temperature. Because NOx generation in hydrogen engines primarily follows the Zeldovich mechanism and is primarily thermal NOx, hydrogen engine NOx emissions are significantly more sensitive to combustion rate or air-fuel ratio (AFR) than diesel and natural gas engines. Existing NOx emission models are mostly based on fixed or steady-state operating conditions and fail to consider the nonlinear effects of air-fuel ratio fluctuations on NOx generation under transient conditions. For example, the NOx prediction model proposed in public document CN 117418924 B, based on pressure differential and time delay, does not address the impact of dynamic changes in air-fuel ratio. Public document CN 118188119 A proposes a method for correcting NOx emissions based on the engine's current operating state and the NOx sensor model downstream of the SCR, but does not address the need for NOx emissions in scenarios without a NOx sensor. Public document CN 114297803 B discloses a method for correcting NOx emissions from diesel engines in standard and non-standard environments, but this method is not applicable to the NOx characteristics of hydrogen engines. Therefore, in order to reduce the cost of aftertreatment systems while meeting strict emissions regulations, it is necessary to design a prediction model that can accurately predict the NOx concentration in the exhaust of hydrogen internal combustion engines. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the deficiencies of the existing technology and provide a method and system for predicting NOx emissions from hydrogen engines based on dynamic correction of the air-fuel ratio, which solves the problem of insufficient NOx prediction accuracy under transient conditions of hydrogen engines and optimizes emissions.

[0006] The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of the air-fuel ratio according to the present invention comprises:

[0007] Collecting operating parameters of the hydrogen engine and obtaining a raw NOx model value based on the operating parameters;

[0008] Obtaining a real-time air-fuel ratio, and calculating an air-fuel ratio correction factor using a partitioned correction strategy according to a difference between the real-time air-fuel ratio and a set reference air-fuel ratio;

[0009] The original exhaust NOx model value is corrected according to the air-fuel ratio correction factor to obtain the original exhaust NOx after the air-fuel ratio correction.

[0010] Preferably, the partition correction strategy is:

[0011] determining the size relationship between the real-time air-fuel ratio and the reference air-fuel ratio;

[0012] if the real-time air-fuel ratio is greater than or equal to the reference air-fuel ratio, then

[0013] if the real-time air-fuel ratio is less than the reference air-fuel ratio, then

[0014] wherein K AFR is an air-fuel ratio correction factor; both a1 and a2 are calibration coefficients, and a1 is less than a2; AFR 实时 is a real-time air-fuel ratio; AFR0 is a reference air-fuel ratio.

[0015] Preferably, the numerical relationship between the calibration coefficients a1 and a2 is: a2:a1=7:1-5:1.

[0016] Preferably, the reference air-fuel ratio is 78-80.

[0017] Preferably, the real-time air-fuel ratio is obtained in the following manner:

[0018] obtaining the working state of the wide-range oxygen sensor, determining whether the wide-range oxygen sensor is working normally, if yes, reading the signal value of the wide-range oxygen sensor and taking it as the real-time air-fuel ratio; otherwise, collecting the instantaneous air flow and the injection pulse width, and calculating the real-time air-fuel ratio according to the air flow and the injection pulse width.

[0019] Preferably, the operating parameters include engine speed, intake manifold pure air pressure, intake manifold temperature and intake manifold humidity.

[0020] Preferably, the original exhaust NOx model value is obtained according to the operating parameters, specifically in the following manner:

[0021] correcting the intake manifold pure air pressure according to the intake manifold humidity and the intake manifold temperature to obtain the effective pure air partial pressure of the intake manifold under the reference environmental condition;

[0022] obtaining the original exhaust NOx model value under the current working condition by looking up a 2D map table according to the engine speed and the effective pure air partial pressure of the intake manifold under the reference environmental condition.

[0023] Preferably, the intake manifold pure air pressure is corrected according to the intake manifold humidity, and is calculated by the following formula:

[0024] humidity-corrected intake manifold pure air pressure = measured pressure - relative humidity x saturated water vapor pressure;

[0025] According to the intake manifold temperature, the humidity-corrected intake manifold pure air pressure is calculated by the following formula:

[0026] The temperature-corrected intake manifold pure air pressure = humidity-corrected intake manifold pure air pressure * reference temperature / intake manifold temperature.

[0027] Preferably, a delay correction time is set, and after the delay correction time ends, the original exhaust NOx corrected by the air-fuel ratio is taken as the predicted value of the original exhaust NOx concentration.

[0028] A hydrogen engine NOx emission prediction system based on dynamic air-fuel ratio correction, comprising:

[0029] An original exhaust NOx base model is used to collect the operating parameters of the hydrogen engine, and the original exhaust NOx model value is obtained according to the operating parameters;

[0030] An air-fuel ratio correction module is used to obtain the real-time air-fuel ratio, and the air-fuel ratio correction factor is calculated by a partition correction strategy according to the difference between the real-time air-fuel ratio and the set reference air-fuel ratio; and the original exhaust NOx model value is corrected according to the air-fuel ratio correction factor to obtain the original exhaust NOx corrected by the air-fuel ratio.

[0031] A delay correction module is used to set a delay correction time, and after the delay correction time ends, the original exhaust NOx corrected by the air-fuel ratio is taken as the predicted value of the original exhaust NOx concentration.

[0032] Advantages

[0033] The advantages of the present application are:

[0034] 1. The coupling mechanism of the air-fuel ratio correction factor and the combustion parameter significantly improves the NOx prediction accuracy under transient operating conditions. Compared with only using the steady-state original exhaust NOx base model, after air-fuel ratio and delay correction, the original exhaust NOx error is reduced by more than 60%.

[0035] 2. The partition correction strategy used in the present application designs different air-fuel ratio correction factors for high air-fuel ratio lean combustion and lower air-fuel ratio lean combustion modes, which not only avoids the problem of transition correction of the air-fuel ratio correction factor, but also ensures the strong correction of the air-fuel ratio correction factor to the air-fuel ratio. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The flowchart of the hydrogen engine NOx emission prediction method based on dynamic air-fuel ratio correction of the present application;

[0037] Figure 2 The NOx value distribution diagram collected in the WHTC city operating condition section of the present application;

[0038] Figure 3 The original NOx emission device of the present application is compared with the transient data of the calculated values of the base model;

[0039] Figure 4 The original NOx emission device of the present application is compared with the transient data of the calculated values of the base model after air-fuel ratio correction;

[0040] Figure 5 The original NOx emission device of the present application is compared with the transient data of the calculated values of the base model after air-fuel ratio and delay correction;

[0041] Figure 6 The integral values of the original NOx emission device of the present application are compared with the integral values of the model. DETAILED DESCRIPTION

[0042] The present application is further described below in conjunction with embodiments, but does not constitute any limitation on the present application, and any limited number of modifications made by anyone within the scope of the claims of the present application is still within the scope of the claims of the present application.

[0043] Reference Figure 1 The NOx emission prediction method for a hydrogen engine based on dynamic correction of air-fuel ratio, the method comprising:

[0044] S10, collecting the operating parameters of the hydrogen engine, and obtaining the original NOx model value according to the operating parameters. The step specifically comprises:

[0045] S101: Collecting the operating parameters of the hydrogen engine, including engine speed, intake manifold pure air pressure, intake manifold temperature and intake manifold humidity.

[0046] S102: Correcting the intake manifold pure air pressure according to the intake manifold humidity and the intake manifold temperature to obtain the effective intake manifold pure air partial pressure under the reference environmental state.

[0047] The humidity correction formula is: humidity-corrected intake manifold pure air pressure = measured pressure - relative humidity x saturated water vapor pressure. The relative humidity is measured by a humidity sensor, and the saturated water vapor pressure is calculated by the Antoine formula.

[0048] The temperature correction formula is: temperature-corrected intake manifold pure air pressure = humidity-corrected intake manifold pure air pressure x reference temperature / intake manifold temperature. The temperature-corrected intake manifold pure air pressure is the effective intake manifold pure air partial pressure under the reference environmental state.

[0049] S103: According to the engine speed and the effective intake manifold pure air partial pressure under the reference environmental state, the original NOx model value under the steady state condition is obtained by searching the 2DMap table.

[0050] S20: Obtain a real-time air-fuel ratio, calculate an air-fuel ratio correction factor based on the difference between the real-time air-fuel ratio and a set reference air-fuel ratio using a partitioned correction strategy, and correct the original exhaust NOx model value based on the air-fuel ratio correction factor to obtain an original exhaust NOx value after the air-fuel ratio correction.

[0051] Figure 2 The relationship between the original NOx and power and air-fuel ratio is shown in the second-collected data of the WHTC cycle urban and suburban conditions. Under the same power, the original NOx has a strong correlation with the air-fuel ratio. The smaller the air-fuel ratio, the higher the original NOx. Therefore, in order to accurately predict the original NOx concentration in transient scenarios, it is necessary to consider the influence of the air-fuel ratio. At the same time, according to Figure 2 At the same power level, when the air-fuel ratio is above 79.1, changes in the air-fuel ratio have little impact on the original NOx emissions. However, when the air-fuel ratio is below 79.1, the further the air-fuel ratio deviates from 79.1, the higher the original NOx concentration. Based on this phenomenon, the air-fuel ratio correction module is partitioned and a baseline air-fuel ratio data (AFR0) is defined. The partitioning rule is based on the comparison between the instantaneous air-fuel ratio and the baseline air-fuel ratio value. The air-fuel ratio correction method is described below.

[0052] S201: Determine the real-time air-fuel ratio.

[0053] Before the wide-range oxygen sensor dew point is released, the oxygen sensor cannot function properly. Therefore, the real-time air-fuel ratio is calculated by collecting instantaneous air flow (such as oxygen mass flow and hydrogen mass flow) and injection pulse width. After the wide-range oxygen sensor dew point is released, the oxygen sensor can function normally, and the engine air-fuel ratio (AFR) signal can be obtained by reading the wide-range oxygen sensor.

[0054] S202: Calculate the air-fuel ratio correction factor and correct the real-time air-fuel ratio.

[0055] According to the degree of deviation of the real-time air-fuel ratio from the reference air-fuel ratio (such as AFR0 = 79.1), the air-fuel ratio correction factor (K AFR ), which is calculated as follows:

[0056]

[0057] Among them, α is the calibration coefficient, which is determined by bench test.

[0058] As mentioned above, the effect of air-fuel ratio on original exhaust NOx depends on the deviation between air-fuel ratio and reference air-fuel ratio. There are two influence modes. Therefore, in order to ensure the accuracy of air-fuel ratio correction, K AFR Perform partition correction (using AFR0 as the benchmark, such as AFR0 = 79.1):

[0059] Correction mode 1: like AFR 实时 ≥AFR0:

[0060] Correction mode 2: like AFR 实时 <AFR0:

[0061] Therefore, depending on the real-time air-fuel ratio, use α1 or α2 to calculate K AFR Correction value.

[0062] like Figure 2 As shown, in AFR 实时 ≥AFR0 range, since the original exhaust NOx is not very sensitive to AFR and AFR0-AFR 实时 ≤0, to avoid K in this area AFR Overcorrection, α1 value is small. 实时 <AFR0区间,由于原排NOx对AFR敏感性很强且AFR0-AFR 实时 >0, to ensure that K AFR The stronger the correction of AFR, the larger the value of α2. In this study, the numerical relationship between α1 and α2 determined based on bench test results is approximately α2 / α1≈6.

[0063] S203: Calculate the original exhaust NOx concentration after the air-fuel ratio correction.

[0064] According to the result obtained in step S2, the original exhaust NOx concentration after the air-fuel ratio correction is calculated. The calculation formula is: Original exhaust NOx after air-fuel ratio correction = Original exhaust NOx model value × K AFR .

[0065] S30, setting a delay correction time, and after the delay correction time ends, using the original exhaust NOx after the air-fuel ratio correction as the predicted value of the original exhaust NOx concentration.

[0066] Considering the signal transmission and NOx mass flow from the turbine to the aftertreatment system, signal time correction is required to simulate the time difference between NOx generation, emission, and signal reception. In this module, two methods are used for delay correction: Method 1 uses a fixed delay correction time (e.g., 3 seconds); Method 2 retrieves the transient correction time from a calibrated 2D map table for each transient moment. This calibrated 2D map is determined through bench testing. The input values ​​are the engine turbine temperature and the original exhaust mass flow rate, and the output value is the delay time (unit: seconds). For each transient moment, the corresponding delay correction time is retrieved from the table. This delay correction time is then used to apply delay correction to the original exhaust NOx model calculated value after air-fuel ratio correction. Method 1 is suitable for engine operating conditions with less transient characteristics, while Method 2 is suitable for engine operating conditions with more transient characteristics.

[0067] The hydrogen engine NOx emission prediction system based on air-fuel ratio dynamic correction comprises a raw NOx basic model, an air-fuel ratio correction module and a delay correction module. The three models / modules are used to realize the hydrogen engine NOx emission prediction method based on air-fuel ratio dynamic correction as described above, specifically,

[0068] The raw NOx basic model is used to collect the operating parameters of the hydrogen engine, and obtain a raw NOx model value according to the operating parameters;

[0069] The air-fuel ratio correction module is used to obtain a real-time air-fuel ratio, calculate an air-fuel ratio correction factor through a partition correction strategy according to the difference between the real-time air-fuel ratio and a set reference air-fuel ratio, and correct the raw NOx model value according to the air-fuel ratio correction factor to obtain a raw NOx after air-fuel ratio correction;

[0070] The delay correction module is used to set a delay correction time, and after the delay correction time ends, use the raw NOx after air-fuel ratio correction as a prediction value of the raw NOx concentration.

[0071] According to the simulation of the hydrogen engine NOx emission prediction system based on air-fuel ratio dynamic correction as described above, Figure 3 、 Figure 4 and Figure 5 respectively show the comparison of the raw NOx emission device collection value of the hydrogen engine, the raw NOx basic model calculation value, the air-fuel ratio corrected model calculation value, and the air-fuel ratio and delay corrected model calculation value in the WHTC cycle 930s-1100s. As shown in Figure 3 , the NOx value calculated using the raw NOx basic model based on the steady state condition has a large difference with the measured raw NOx data in the transient state condition, and the basic model calculation value cannot predict the NOx peak value due to the change of the air-fuel ratio in the transient state. As shown in Figure 4 , after the air-fuel ratio correction, the raw NOx model calculation value can accurately predict the peak shape of the NOx peak, but there is still a time sequence deviation between the model calculation value and the measured raw NOx data. As shown in Figure 5 , after the air-fuel ratio correction is used and the delay correction is added, the model calculation value can not only accurately predict the peak shape of the NOx peak, but also accurately predict the position of the NOx peak.

[0072] Figure 6The comparison of the integral values of the NOx emission device collected in the whole WHTC cycle and the model integral values is shown. In the whole WHTC cycle, the integral value of the original NOx basic model is about 75% smaller than the integral value of the measured original NOx emission data, and the integral value of the model after the air-fuel ratio and delay correction is only about 7% smaller than the integral value of the measured original NOx emission data. Therefore, the data shows that after the comparison, the error of the original NOx is reduced by more than 60% compared with only using the steady-state original NOx basic model.

[0073] The above only describes the preferred embodiments of the present application, and it should be pointed out that for those skilled in the art, without departing from the structure of the present application, a number of modifications and improvements can be made, which will not affect the effect of the implementation of the present application and the practicability of the patent.

Claims

1. A hydrogen engine NOx emission prediction method based on dynamic correction of air-fuel ratio, characterized in that: The method includes: Collecting operating parameters of the hydrogen engine and obtaining a raw NOx model value based on the operating parameters; Obtaining a real-time air-fuel ratio, and calculating an air-fuel ratio correction factor using a partitioned correction strategy according to a difference between the real-time air-fuel ratio and a set reference air-fuel ratio; The original exhaust NOx model value is corrected according to the air-fuel ratio correction factor to obtain the original exhaust NOx after the air-fuel ratio correction.

2. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 1, characterized in that: The inter-partition correction strategy is: Determining a magnitude relationship between the real-time air-fuel ratio and a reference air-fuel ratio; If the real-time air-fuel ratio is greater than or equal to the reference air-fuel ratio, then If the real-time air-fuel ratio is less than the reference air-fuel ratio, then Where K AFR is the air-fuel ratio correction factor; α1 and α2 are both calibration coefficients, and α1 is less than α2; AFR 实时 is the real-time air-fuel ratio; AFR0 is the reference air-fuel ratio.

3. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 2, characterized in that: The numerical relationship between the calibration coefficients α1 and α2 is: α2:α1=7:1~5:

1.

4. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 1, characterized in that: The base air-fuel ratio is 78-80.

5. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 1, characterized in that: The real-time air-fuel ratio is obtained as follows: Obtain the working status of the margin oxygen sensor and determine whether the margin oxygen sensor is working normally. If so, read the signal value of the margin oxygen sensor and use it as the real-time air-fuel ratio; otherwise, collect the instantaneous air flow and injection pulse width, and calculate the real-time air-fuel ratio based on the air flow and injection pulse width.

6. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 1, characterized in that: The operating parameters include engine speed, intake manifold pure air pressure, intake manifold temperature and intake manifold humidity.

7. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 6, characterized in that: The original exhaust NOx model value is obtained according to the operating parameters, specifically: Correcting the intake manifold pure air pressure according to the intake manifold humidity and the intake manifold temperature to obtain an effective intake manifold pure air partial pressure under a reference environmental state; According to the engine speed and the effective pure air partial pressure of the intake manifold under the reference environmental conditions, the original exhaust NOx model value under the current working condition is obtained by searching the 2D Map table.

8. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 7, characterized in that: The intake manifold pure air pressure is corrected according to the intake manifold humidity and calculated by the following formula: Intake manifold pure air pressure after humidity correction = measured pressure - relative humidity × saturated water vapor pressure; The intake manifold pure air pressure corrected by the intake manifold temperature and humidity is calculated by the following formula: Temperature-corrected intake manifold pure air pressure = humidity-corrected intake manifold pure air pressure × reference temperature / intake manifold temperature.

9. The method for predicting NOx emissions from a hydrogen engine based on dynamic correction of air-fuel ratio according to claim 1, characterized in that: A delay correction time is set, and after the delay correction time ends, the original exhaust NOx after the air-fuel ratio correction is used as the predicted value of the original exhaust NOx concentration.

10. A hydrogen engine NOx emission prediction system based on dynamic correction of air-fuel ratio, characterized in that: include: A raw NOx basic model is used to collect operating parameters of the hydrogen engine and obtain raw NOx model values ​​based on the operating parameters; an air-fuel ratio correction module, configured to obtain a real-time air-fuel ratio and calculate an air-fuel ratio correction factor based on a difference between the real-time air-fuel ratio and a set reference air-fuel ratio using a partitioned correction strategy; and correcting the original exhaust NOx model value according to the air-fuel ratio correction factor to obtain the original exhaust NOx after the air-fuel ratio correction; The delay correction module is used to set a delay correction time, and after the delay correction time ends, use the original exhaust NOx after the air-fuel ratio correction as the predicted value of the original exhaust NOx concentration.

Images