A Calibration Method for Catalyst Reaction Model in a Vehicle Aftertreatment Control Strategy

Abstract

The present invention provides a calibration method for a catalyst reaction model in an automotive aftertreatment control strategy, the calibration method comprising the following steps: (1) determining the chemical reaction rate model, i.e., the approximate range of the chemical reaction pre-exponential factor K and activation energy E in the Arrhenius chemical reaction rate equation, through DOE calculation, with the goal of minimizing the deviation between the chemical reaction rate model value and the experimental value; and (2) based on a detailed slice model, with the goal of minimizing the deviation of the catalyst outlet component concentration, meticulously calibrating the pre-exponential factor K, activation energy E, including the correction term coefficient θ, one by one through an optimization solver. The method of the present invention reduces the scientific literacy requirements of R&D personnel for model calibration, significantly saves catalyst model development time, and improves the model's prediction accuracy, better meeting the development needs of afterprocessors for National VI and even higher emission levels.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle control systems and relates to a calibration method for a catalyst reaction model in a vehicle after-treatment control strategy. Background Art

[0002] With the continuous upgrading of emission regulations, automotive post-processor control strategies have evolved from open-loop and closed-loop control to adaptive control and model-based predictive control, significantly improving the overall efficiency and stability of the system. Model-based predictive control, in particular, requires establishing a slicing model based on chemical reaction kinetics based on gas flow, heat transfer, diffusion, and chemical reactions in the post-processor to accurately calculate the components of the exhaust gas after passing through the catalyst.

[0003] CN106837495A provides a method for estimating NO2% after DOC based on a detailed slice model. The DOC is divided into a certain number of unit modules along the axial direction. The energy conservation equation, chemical equilibrium equation, and mass conservation equation are applied to each unit to calculate the NO storage capacity and NO2 storage capacity in each unit, as well as the NO2 and NO concentrations in the exhaust gas. The model considers six chemical reaction equations related to the calculation of NOx concentration in the DOC and involves 16 calibration parameters. Although it introduces the model prediction mechanism, it does not involve the calibration method of the catalyst chemical reaction model parameters.

[0004] CN111379636A discloses a method for regulating the filling level of an exhaust gas component storage device of a catalytic converter of an internal combustion engine. The filling level is regulated using a route model that includes a catalytic converter model, and uncertainties in measured or model variables influencing the filling level regulation are corrected by adaptation based on signals from an exhaust gas sensor disposed on the output side of the catalytic converter. The method employs a method for compensating for model parameter inaccuracies using sensor measurements from the exhaust gas sensor.

[0005] Whether the control strategy can play its role to the maximum extent depends heavily on the calibration accuracy of the catalyst model parameters. Since it involves too many chemical reactions and calibration parameters, the calibration of the catalyst model is a highly professional and time-consuming task. At present, the calibrated models are all output as "black cores" and are only used for post-processor performance calculations in their own commercial software. The models cannot be modified, nor can they be directly added to the independently developed control model. Therefore, their significance for the development of independent control strategies is reduced. In addition, R&D personnel can also use general optimization tools to calibrate the model parameters. However, in actual use, when there are more than three parameters to be calibrated, the optimization solver tends to converge on a local target.

[0006] In general, using a general optimization solver, it is urgent to build a targeted optimization process to calibrate the catalyst physical model with significantly improved optimization time and accuracy. Summary of the Invention

[0007] The purpose of the present invention is to provide a calibration method for a catalyst reaction model in an automotive aftertreatment control strategy. The present invention generates K and E random numbers for initialization to determine in which design intervals the catalyst may have the optimal solution. On this basis, detailed optimization is performed in steps. The method of the present invention is generally applicable to chemical reaction models established based on the Arrhenius equation, reduces the scientific literacy requirements of R&D personnel for model calibration, greatly saves catalyst model development time and improves the model's prediction accuracy, and better meets the development needs of post-processors for National VI and even higher emission levels.

[0008] In order to achieve the purpose of the invention, the present invention adopts the following technical solutions:

[0009] The present invention provides a method for calibrating a catalyst reaction model in a vehicle aftertreatment control strategy. The calibration method comprises the following steps:

[0010] (1) Based on the lumped parameter model, the temperature and reactant concentration changes along the catalyst axis are not considered. The calculated value of the chemical reaction rate r is expressed based on the standard Arrhenius equation as follows:

[0011]

[0012] Among them, T cat is the catalyst temperature, C rec is the reactant concentration, K, E T are the pre-exponential factor and activation energy of chemical reaction respectively. T Perform DOE calculations and determine K and E based on the chemical reaction rate r model calculation value and the experimental value, which should not deviate by more than 20%. T preliminary scope;

[0013] (2) Based on a detailed slice model, that is, considering the changes in the temperature and reactant concentration of the catalyst along the axial direction, the catalyst is divided into a certain number of unit modules along the axial direction. The energy conservation equation, chemical equilibrium equation and mass conservation equation are applied to each unit to calculate the reactant concentration in each unit and at the catalyst outlet. In the detailed slice model, the chemical reaction rate r det It can be roughly expressed as:

[0014]

[0015] Where θ is the chemical reaction rate correction term, and T is the chemical reaction temperature cator concentration C rec The relationship between

[0016] With the goal of ensuring that the deviation between the calculated and experimental values of the reactant concentration at the catalyst outlet does not exceed 5%, the correction term coefficient θ, the pre-exponential factor K, and the activation energy E are determined by the optimization solver. T The precise range.

[0017] The present invention first determines the chemical reaction pre-exponential factor K and activation energy E by DOE calculation. T The approximate range of the correction term coefficient θ and the pre-exponential factor K and activation energy E are determined by the optimization solver. T The precise range of DOE calculation and chemical reaction rate calculation adopts the lumped parameter model, and temporarily does not consider the change of reactant concentration along the flow direction of the catalyst. While shortening the DOE calculation time, the pre-exponential factor K and activation energy E are established. T approximate range.

[0018] The method of the present invention reduces the scientific literacy requirements of R&D personnel for model calibration, greatly saves catalyst model development time and improves the prediction accuracy of the model, and better meets the development needs of post-processors for National VI and even higher emission levels.

[0019] Preferably, an experimental design is performed before the DOE calculation in step (1).

[0020] Preferably, the experimental design includes designing at least three catalyst-to-reactant conversion efficiency tests under different temperature and space velocity conditions, and recording the reactant concentration before the catalyst Rec ppm bef, the reactant concentration after the catalyst Rec ppmaft, the temperature before the catalyst CAT Temp mono inl, the temperature after the catalyst CAT Temp mono outl and the space velocity SV.

[0021] Preferably, the experimental data is pre-processed after the experimental design to calculate the experimental chemical reaction rate.

[0022] Preferably, the experimental chemical reaction rate is recorded as Rec rate exp.

[0023] Preferably, the calculation formula for the experimental chemical reaction rate is Rec rate exp=(Rec ppm bef-Rec ppm aft)*1e-6*Mol air C / CAT Volu mono C, wherein Mol air C is the number of moles of air and CAT Volumono C is the volume of the catalyst.

[0024] Preferably, Mol air C=mass air C / MolM air C.

[0025] Preferably, mass air C=pace velocity*CAT Volu mono C*Dens air C / 3600.

[0026] Preferably, MolM air C=Dens air C*8.315*273.15 / 1e5.

[0027] Preferably, CAT Volu mono C=CAT AreCros mono C*CAT Leng mono C*25.4*1e-3.

[0028] Preferably, CAT AreCros mono C=3.14*CAT Diam mono C*CAT Diam mono C / 4*25.4*25.4*1e-6.

[0029] Preferably, step (1) specifically includes using a random generation function to generate random numbers of K and E, and using the basic Arrhenius equation to determine the calculated chemical reaction rate values corresponding to K and E.

[0030] Preferably, the calculation formula of the chemical reaction rate is Wherein, Rec rate sim is the calculated chemical reaction rate, R is the gas molar constant, C is the molar concentration of the chemical reactant, C is the experimental measurement value, and CAT Temp mono is the catalyst reaction temperature.

[0031] Preferably, CAT Temp mono=(CAT Temp mono inl+CAT Temp mono outl).

[0032] Preferably, the error between the experimental chemical reaction rate and the calculated rate is observed through DOE calculation results, and the calculation formula is Rec rate err = abs (Rec rate sim - Rec rate exp) / Rate exp * 100.

[0033] Preferably, the Rec rate err not exceeding 20% is used as a criterion for determining the appropriate interval of the calibration parameters K and E.

[0034] Preferably, step (2) specifically includes K, E T As the initial value, and the change of ± 5% as K, E T Further optimization interval, substitute the detailed slice model, calculate the output catalyst outlet component concentration N, and use the optimization algorithm to adjust K, E T The other parameters in the θ term are calibrated simultaneously.

[0035] Preferably, in the slicing model of step (2), the chemical reaction rate r detail The formula is Where T cat is the temperature of the chemical reaction carrier, C rec is the molar concentration of the chemical reactant, and θ is the chemical reaction rate correction term.

[0036] The present invention is used to determine the chemical reaction pre-exponential factor K, activation energy E T Based on the approximate range of the catalyst, a detailed slicing model is used to minimize the deviation of the catalyst outlet component concentration. The pre-exponential factor K, activation energy E T , correction item coefficients, etc. are calibrated one by one in detail.

[0037] Preferably, step (2) further comprises using Emis out err to determine the accuracy of the detailed slice model.

[0038] Preferably, the calculation formula of Emis out err is: Emis out err=abs(Emis out sim-Emis outexp) / Emis out exp.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The method of the present invention reduces the scientific literacy requirements of R&D personnel for model calibration, greatly saves catalyst model development time and improves the prediction accuracy of the model, and better meets the development needs of post-processors for National VI and even higher emission levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a process flow chart of the calibration method described in Example 1 of the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention is further described below by way of specific embodiments. It should be understood by those skilled in the art that the embodiments are merely to help understand the present invention and should not be regarded as specific limitations of the present invention.

[0043] In the calibration method described in Example 1 of the present invention, the calculation parameter table is shown in Table 1:

[0044] Table 1

[0045]

[0046]

[0047] Example 1

[0048] This embodiment provides a method for calibrating a catalyst reaction model in a vehicle aftertreatment control strategy. The calibration method includes the following steps. The flow chart of the calibration method is shown in FIG. Figure 1 As shown:

[0049] Step 1: Experimental Design:

[0050] Design at least three tests for catalyst-to-reactant conversion efficiency under different temperature and space velocity conditions. Record the reactant concentration before the catalyst (Rec ppm bef), the reactant concentration after the catalyst (Rec ppm aft), the temperature before the catalyst (CAT Temp mono inl), the temperature after the catalyst (CAT Temp mono outl), and the space velocity (SV).

[0051] Step 2: Pre-process the experimental data and calculate the chemical reaction rate:

[0052] Chemical reaction rate of working condition 1 test:

[0053] Rec1 rate exp=(Rec1 ppm bef-Rec1 ppm aft)*1e-6*Mol air C / CAT Volumono C Test chemical reaction rate under condition 2:

[0054] Rec2 rate exp=(Rec2 ppm bef-Rec2 ppm aft)*1e-6*Mol air C / CAT Volumono C Working Condition 3 Test Chemical Reaction Rate:

[0055] Rec3 rate exp=(Rec3 ppm bef-Rec3 ppm aft)*1e-6*Mol air C / CAT Volumono C, where Mol air C is the number of moles of air and CAT Volumono C is the volume of the catalyst, which can be calculated using the following formula:

[0056] CAT AreCros mono C=3.14*CAT Diam mono C*CAT Diam mono C / 4*25.4*25.4*1e-6,

[0057] CAT Volu mono C=CAT AreCros mono C*CAT Leng mono C*25.4*1e-3,

[0058] mass air C=space velocity*CAT Volu mono C*Dens air C / 3600,

[0059] MolM air C=Dens air C*8.315*273.15 / 1e5,

[0060] Mol air C=mass air C / MolM air C,

[0061] CAT Temp mono C=(CAT Temp mono inlet+CAT Temp mono outlet) / 2(K);

[0062] Step 3: Set the parameters K and E to be calibrated T The basic Arrhenius equation can usually be used to

[0063]

[0064] , where r is the chemical reaction rate (MOL*M -3 S -1 ), R is the gas molar constant, C rec is the molar concentration of the chemical reactants, T cat is the catalyst temperature. K is the chemical reaction pre-exponential factor, E T is the activation energy of chemical reaction. T It is the parameter to be calibrated, and its value range is (0~10 N ). cat 、C rec is the experimental measurement value, T cat Take the average temperature of the catalyst inlet and outlet, C rec Taken as the concentration of reactants at the catalyst inlet;

[0065] Step 4: Use random generation function to generate K and E T Random numbers, and use the basic Arrhenius equation to calculate different K, E T The corresponding chemical reaction rate:

[0066] Calculate the chemical reaction rate for case 1:

[0067]

[0068] Calculate the chemical reaction rate for Case 2:

[0069]

[0070] Calculate the chemical reaction rate for working case 3:

[0071]

[0072] Step 5: Observe the DOE results and calculate the error between the experimental chemical reaction rate and the calculated rate:

[0073] Rec1 rate err=abs(Rec1 rate sim-Rec1 rate exp) / Rate1 exp

[0074] Rec2 rate err=abs(Rec1 rate sim-Rec1 rate exp) / Rate1 exp

[0075] Rec3 rate err=abs(Rec1 rate sim-Rec1 rate exp) / Rate1 exprate err=(Rec1 rate err+Rec2 rate err+Rec3 rate err) / 3;

[0076] Step 6: Based on whether the rate err is less than 20%, determine whether the set calibration parameters K and E range are appropriate. If there is no minimum value of rate err in the entire K and E range, or the minimum value appears on the boundary of the K and E set range, return to step 3; otherwise, go to step 7;

[0077] Step 7: Based on the optimal values of K and E obtained in step 6, take them as the initial values, and use ±5% as the further optimization interval of K and E, substitute them into the detailed slice model, and calculate the output catalyst outlet component concentration N. In the slice model, the chemical reaction rate r detail Expressed as:

[0078]

[0079] Where K: chemical reaction pre-exponential factor;

[0080] E T : activation energy of chemical reaction;

[0081] T cat : Chemical reaction carrier temperature;

[0082] C rec : chemical reactant molar concentration;

[0083] θ: Chemical reaction rate correction term, usually the chemical reaction temperature T cat or concentration C rec The relationship generally contains several parameters to be calibrated;

[0084] Step 8: Use optimization algorithms such as genetic algorithms to optimize K and E. T, and other parameters in the θ term are calibrated simultaneously. Obtain Emis out 1sim, Emis out2sim, and Emis out3sim under three working conditions calculated according to the detailed slicing equation;

[0085] Step 9: Based on whether Emis out err rete err is less than 10%, determine whether the accuracy of the detailed slice model is sufficient. If the accuracy is poor, return to step 6 and increase K and E appropriately. T The value range of:

[0086] Emis out 1err=abs(Emis out 1sim-Emis out 1exp) / Emis out 1exp

[0087] Emis out 2err=abs(Emis out 2sim-Emis out 2exp) / Emis out 2exp

[0088] Emis out 3err=abs(Emis out 3sim-Emis out 3exp) / Emis out 3exp

[0089] Emis out err=(Emis out 1err+Emis out 2err+Emis out 3err) / 3;

[0090] Step 10: Output K and E T and other parameters in the θ term.

[0091] The applicant declares that the above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Those skilled in the art should understand that any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention fall within the scope of protection and disclosure of the present invention.

Claims

1. A method for calibrating a catalyst reaction model in a vehicle aftertreatment control strategy, characterized in that: The calibration method comprises the following steps: (1) Based on the lumped parameter model, the temperature and reactant concentration changes along the catalyst axis are not considered. The calculated value of the chemical reaction rate r is expressed based on the standard Arrhenius equation as follows: Among them, T cat is the catalyst temperature, C rec is the reactant concentration, K, E T They are the chemical reaction pre-exponential factor and activation energy, and the pre-exponential factor K and activation energy E are T Perform DOE calculations and determine K and E based on the chemical reaction rate r model calculation value and the experimental value, which should not deviate by more than 20%. T preliminary scope; (2) Based on a detailed slice model, that is, considering the changes in the temperature and reactant concentration of the catalyst along the axial direction, the catalyst is divided into a certain number of unit modules along the axial direction. The energy conservation equation, chemical equilibrium equation and mass conservation equation are applied to each unit to calculate the reactant concentration in each unit and at the catalyst outlet. In the detailed slice model, the chemical reaction rate r det Expressed as: Where θ is the chemical reaction rate correction term, and T is the chemical reaction temperature cat or concentration C rec The relationship between With the goal of ensuring that the deviation between the calculated and experimental values of the reactant concentration at the catalyst outlet does not exceed 5%, the correction term coefficient θ, the pre-exponential factor K, and the activation energy E are determined by the optimization solver. T The precise range of Step (1) specifically includes using a random generation function to generate K, E T The random number and the basic Arrhenius equation are used to determine K, E T The corresponding chemical reaction rate calculation value, the chemical reaction rate calculation formula is Wherein, Rec rate sim is the calculated chemical reaction rate, R is the gas molar constant, C is the molar concentration of the chemical reactant, C is the experimental measurement value, and CAT Temp mono is the catalyst reaction temperature.

2. The calibration method according to claim 1, wherein: Before the DOE calculation in step (1), an experimental design is performed.

3. The calibration method according to claim 2, wherein: The test design includes designing at least three catalyst-to-reactant conversion efficiency tests under different temperature and space velocity conditions, and recording the reactant concentration before the catalyst Rec ppm bef, the reactant concentration after the catalyst Rec ppm aft, the temperature before the catalyst CAT Temp mono inl, the temperature after the catalyst CATTemp mono outl and the space velocity SV.

4. The calibration method according to claim 2, wherein: After the experimental design, the experimental data is pre-processed and the experimental chemical reaction rate is calculated.

5. The calibration method according to claim 4, wherein: The experimental chemical reaction rate is recorded as Rec rate exp.

6. The calibration method according to claim 5, wherein: The calculation formula of the experimental chemical reaction rate is Rec rate exp = (Rec ppm bef - Rec ppm aft) * 1e-6 * Mol air C / CAT Volu mono C, wherein Mol air C is the number of moles of air and CAT Volu mono C is the volume of the catalyst.

7. The calibration method according to claim 6, wherein: Mol air C = mass air C / MolM air C, where mass air C is the air flow rate and MolM air C is the molar mass of air.

8. The calibration method according to claim 7, wherein: mass air C = space velocity * CATVolu mono C * Dens air C / 3600, where space velocity is the air speed SV and Dens air C is the air density under standard conditions.

9. The calibration method according to claim 7, wherein: MolM air C=Dens air C*8.315*273.15 / 1e5.

10. The calibration method according to claim 8, wherein: CAT Volu mono C = CAT AreCros monoC * CAT Leng mono C * 25.4 * 1e-3, where CAT Leng mono C is the vector length; Among them, CAT AreCros mono C=3.14*CAT Diam mono C*CAT Diam mono C / 4*25.4*25.4*1e-6, where CAT Diam mono C is the carrier diameter.

11. The method according to claim 1, wherein CAT Temp mono=(CAT Temp mono inl+CAT Temp mono outl), where CAT Temp mono inl is the temperature before the catalyst, and CAT Temp mono outl is the temperature after the catalyst.

12. The method according to claim 1, wherein The error between the experimental chemical reaction rate and the calculated rate was observed through DOE calculation results. The calculation formula is Rec rate err = abs (Rec rate sim - Rec rate exp) / Rateexp, where Rec rate sim is the calculated chemical reaction rate, Rec rate exp is the experimental chemical reaction rate, and Rateexp is the experimental value of the reaction rate.

13. The method according to claim 12, wherein: According to the Rec rate err not exceeding 20%, it is used as the calibration parameters K and E T The standard for the appropriate interval.

14. The method according to claim 1, wherein Step (2) specifically includes K, E T As the initial value, and the change of ± 5% as K, E T Further optimization interval, substitute the detailed slice model, calculate the output catalyst outlet component concentration N, and use the optimization algorithm to adjust K, E T Other parameters in the θ term are calibrated simultaneously, and the optimization algorithm includes a genetic algorithm.

15. The method according to claim 1, wherein In the slice model of step (2), the chemical reaction rate r detail The formula is Where T cat is the chemical reaction carrier temperature, C rec is the molar concentration of the chemical reactant, and θ is the chemical reaction rate correction term.

16. The method according to claim 1, wherein Step (2) also includes using Emis out err to determine the accuracy of the detailed slice model.

17. The method according to claim 16, wherein The calculation formula of Emis out err is: Emis out err = abs(Emis out sim - Emis out exp) / Emis out exp, where Emis out sim is the calculated value of the system output accuracy, and Emis out exp is the tested value of the system output accuracy.

Images