A chemical reaction mechanism optimization method and system based on coordinate descent method

Abstract

The application discloses a chemical reaction mechanism optimization method and system based on a coordinate descent method, and the method comprises the following steps: inputting a chemical reaction mechanism file to be optimized, a detailed reaction mechanism file, and set optimization working conditions and important components; judging the reaction type of each base reaction in the reaction mechanism and determining the number of optimization variables; automatically generating a target function according to the optimization working conditions and the important components; carrying out sensitivity analysis on the target function and excluding unimportant optimization variables; determining the coordinate order according to the sensitivity analysis result and generating a search direction sequence; carrying out line search on the search direction sequence in sequence based on a spline interpolation line search strategy until a termination condition is met, and saving the search result to a history file; and outputting an optimized mechanism file. The line search strategy is applied to the coordinate descent method to carry out chemical reaction mechanism optimization, which is beneficial to making the calculation result closer to the detailed mechanism, and meanwhile, the method has faster speed, better calculation precision and robustness.

Description

Technical Field

[0001] This invention relates to the field of chemical reaction mechanism optimization technology, and in particular to a chemical reaction mechanism optimization method and system based on coordinate descent. Background Technology

[0002] Currently, numerical simulations of chemical reaction flow processes are frequently required in the engineering design and fundamental research of combustion devices such as gas turbines, rocket engines, and piston engines. In Computational Fluid Dynamics (CFD) simulations of reaction flow, the accuracy of the chemical reaction mechanism model used directly affects the overall computational accuracy of the simulation.

[0003] Specifically, in reaction flow simulation, a chemical reaction mechanism file needs to be provided to the simulation program. This file specifies all the elementary reactions involved in the combustion process and their corresponding Arrhenius constants. Based on this information, the calculation program will use the Arrhenius formula to calculate the chemical reaction source term in the governing equation.

[0004] The process of optimizing chemical reaction mechanisms refers to altering the Arrhenius constant in the mechanism to increase its precision. The precision of a chemical reaction mechanism is reflected in its error compared to experimental data or simulation data from a detailed mechanism; the smaller the error, the higher the precision. Sometimes there is a detailed mechanism with high precision, but its scale is too large for direct calculation; other times there is a simplified mechanism with a smaller scale but lower precision.

[0005] Therefore, how to optimize the simplified mechanism through optimization methods to make the calculation results closer to the detailed mechanism, while having faster speed, better calculation accuracy and robustness, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] In view of this, the present invention provides a chemical reaction mechanism optimization method and system based on coordinate descent that can solve the above-mentioned technical problems. By applying the line search strategy to the coordinate descent method for chemical reaction mechanism optimization, the calculation results are closer to the detailed mechanism. At the same time, it has faster speed, better calculation accuracy and robustness, and can better solve the complex problem of chemical reaction mechanism optimization.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] In a first aspect, embodiments of the present invention provide a method for optimizing chemical reaction mechanisms based on the coordinate descent method, the method comprising the following steps:

[0009] S1. Input the chemical reaction mechanism file and detailed reaction mechanism file to be optimized, as well as the set optimization conditions and important components;

[0010] S2. Determine the reaction type of each elementary reaction in the reaction mechanism and determine the number of optimization variables;

[0011] S3. Automatically generate the objective function based on the input optimization conditions and key components;

[0012] S4. Conduct sensitivity analysis on the objective function to eliminate unimportant optimization variables;

[0013] S5. Determine the coordinate order based on the sensitivity analysis results and generate a search direction sequence;

[0014] S6. A line search strategy based on spline interpolation is used to perform line searches sequentially on the search direction sequence until the termination condition is met, and the search results are saved to the history file.

[0015] S7. Optimization complete, output optimization mechanism file.

[0016] Preferably, in step S2, the number of variables required for optimization of the elementary reaction is determined according to the reaction type, the number of variables required for the entire mechanism is determined by summing the number of variables for all elementary reactions, and a mapping relationship is established.

[0017] Preferably, in step S3, the generated objective function is:

[0018]

[0019] Among them, A k b k E k These represent the pre-exponential factor, temperature exponent, and activation energy of the Arrhenius equation for the k-th reaction in the reaction mechanism to be optimized; m is the operating condition number; F 1，m ~F 4，m The four sources of error in the calculation results of the mechanism to be optimized and the detailed mechanism are used to characterize the specific working conditions under each condition:

[0020] (1)F 1，m The error in the result of detailed mechanism calculations, specifically the error in reactants or products, is represented by the following formula:

[0021]

[0022] Where, τ s τ is the moment when the reaction begins. e I is the time when the reaction terminates. I For the set of reactants or products to be considered, x i，mFor the m-th operating condition, the mole fraction of the i-th reactant or product at each moment in the calculation results, with the superscript D indicating the calculation results of the detailed mechanism;

[0023] (2)F 2，m The error of intermediate products, used to characterize the error between the results and detailed mechanism calculations, is expressed by the following formula:

[0024]

[0025] Among them, I p This indicates the intermediate product error that needs to be considered; τ p，j，m This refers to the time when the j-th intermediate product reaches its peak value or the time when the generation / consumption rate reaches its peak value under the m-th operating condition.

[0026] (3)F 3，m The error in temperature data, used to characterize the error between the results and detailed mechanistic calculations, is expressed by the following formula:

[0027]

[0028] Among them, T m For the m-th operating condition, the temperature at each moment in the calculation results is:

[0029] (4)F 4，m The error in the peak rate of temperature rise, used to characterize the error between the results and detailed mechanistic calculations, is expressed by the following formula:

[0030]

[0031] Where, τ T，m This represents the moment when the rate of temperature change reaches its peak under the m-th operating condition.

[0032] Preferably, in step S4, the sensitivity of the optimization variables is sorted. First, the optimization variables with a sensitivity of 0 to all targets are marked as variables that need to be skipped. Then, the root mean square of the sensitivity of all targets is taken and sorted again. Variables with lower sensitivity are marked according to the number of inputs.

[0033] Preferably, in step S5, the coordinate order is determined based on the sensitivity analysis results, and the order is either the positive sensitivity order or the negative sensitivity order.

[0034] Preferably, in step S6, the line search strategy based on spline interpolation specifically includes:

[0035] ① Enter the current position x (k) Search direction p (k) Initial number of points n init and maximum step size γ max ;

[0036] ② Take n uniformly within the search interval init From a set of points, we obtain a series of initial points X, with one endpoint of the search interval being... The other end is

[0037] ③ Calculate the objective function at a series of initial points X to obtain a series of objective function values ​​F;

[0038] ④ Take n uniformly within the search interval interp From these points, a series of interpolation points X are obtained. interp ;

[0039] ⑤ For the series of objective function values ​​F obtained in step ③ at this series of interpolation points X interp Spline interpolation is performed to obtain a series of interpolation objective function values ​​F. interp ;

[0040] ⑥ Obtain the initial point x corresponding to the minimum value among the series of objective function values ​​F obtained in step ③. t，min，0 ;

[0041] ⑦ Obtain the series of interpolation objective function values ​​F obtained in step ⑤. interp The interpolation point x corresponding to the minimum value t，min，1 And calculate x t，min，1 objective function value at ft ,min,1;

[0042] ⑧If x t，min，0 and x t，min，1 If the distance meets the preset requirement, the algorithm is considered to have converged, and x is returned. t，min，1 If it does not converge, then repeat the process of applying x. t，min，1 Add it to the initial point X, and set f t，min，1 Add it to the objective function value F, and jump back to step ④.

[0043] Secondly, embodiments of the present invention also provide a chemical reaction mechanism optimization system based on the coordinate descent method. This system applies the aforementioned chemical reaction mechanism optimization method based on the coordinate descent method to optimize chemical reaction mechanisms. The system includes:

[0044] The input module is used to input the chemical reaction mechanism file and detailed reaction mechanism file that need to be optimized, as well as the set optimization conditions and important components;

[0045] The judgment module is used to determine the reaction type of each elementary reaction in the reaction mechanism and to determine the number of optimization variables.

[0046] The objective function generation module is used to generate an objective function based on the input optimization conditions and key components;

[0047] The sensitivity analysis module is used to perform sensitivity analysis on the objective function and exclude unimportant optimization variables;

[0048] The coordinate order determination module is used to determine the coordinate order based on the sensitivity analysis results and generate a search direction sequence;

[0049] The line search module is used to perform line search on the search direction sequence in sequence according to the spline interpolation line search strategy until the termination condition is met, and save the search results to the history file.

[0050] The output module is used to output the optimization mechanism file after optimization is completed.

[0051] Compared with the prior art, the present invention has at least the following beneficial effects:

[0052] 1. This invention provides a chemical reaction mechanism optimization method and system based on coordinate descent. By applying the line search strategy to the coordinate descent method for chemical reaction mechanism optimization, the calculation results are closer to the detailed mechanism, while also having faster speed, better calculation accuracy and robustness.

[0053] 2. The objective function designed for optimizing chemical reaction mechanisms in this invention can better characterize the difference between the mechanism to be optimized and the detailed mechanism. The line search strategy based on spline interpolation adopted in this invention densifies the value at the expected minimum when the minimum value of the objective function becomes unapproximate. At the same time, when densification fails to approximate the value, the inherent characteristics of the strategy itself can ensure that the loop is exited, preventing excessive waste of computation time. This invention applies this line search strategy to the coordinate descent method and ultimately to the optimization of chemical reaction mechanisms. Due to the robustness of this line search strategy and the coordinate descent method itself, it is beneficial to better solve the complex problem of optimizing chemical reaction mechanisms.

[0054] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0055] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0058] Figure 1 This is a schematic flowchart of a chemical reaction mechanism optimization method based on coordinate descent, provided for an embodiment of the present invention.

[0059] Figure 2 This is a flowchart illustrating the line search strategy provided in an embodiment of the present invention.

[0060] Figure 3 This diagram illustrates the objective function descent process during the execution of different optimization methods provided in the embodiments of the present invention.

[0061] Figure 4 The diagram illustrates the component results of the final objective function obtained by different optimization methods provided in the embodiments of the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0063] In the description of this invention, it should be noted that some processes described in this application specification and drawings include multiple operations that appear in a specific order. However, it should be clearly understood that these operations may be performed in any order or in parallel. Furthermore, various numbers are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0064] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0065] See Figure 1 As shown, this invention provides a method for optimizing chemical reaction mechanisms based on the coordinate descent method. This method mainly includes the following steps:

[0066] 1. Based on the user's instructions, read in the chemical reaction mechanism file and detailed reaction mechanism file that need to be optimized, and read in the optimization conditions and important components given by the user;

[0067] 2. Read the reaction type of each elementary reaction in the reaction mechanism, determine the number of variables required to optimize this elementary reaction according to the reaction type, sum the number of variables of all elementary reactions to determine the number of variables required for the entire mechanism, and establish a mapping relationship;

[0068] 3. Automatically generate the calculation function for the objective function based on the important components and operating conditions input by the user;

[0069] 4. Conduct sensitivity analysis on the objective function and rank the sensitivity of the optimization variables. First, mark the optimization variables with a sensitivity of 0 to all objectives as variables that need to be skipped. Then, take the root mean square of the sensitivity of all objectives and rank them again. Mark the variables with lower sensitivity according to the number of user inputs.

[0070] 5. Determine the coordinate order based on the sensitivity analysis results. Two orders can be selected: sensitivity order or sensitivity reverse order. Sensitivity order can achieve a faster descent speed in the early stages. Generate a search direction sequence, where the k-th search direction contains the k-th coordinate. One component is 1, and the other directional components are 0;

[0071] 6. Perform line searches sequentially along the search direction sequence until the termination condition is met, then append and save the search results to the history file;

[0072] 7. Optimization complete, output optimization mechanism file.

[0073] The specific implementation of the objective function used in the method of the present invention will be described in detail below:

[0074] In optimizing chemical reaction mechanisms, existing methods involve performing zero-dimensional auto-ignition or one-dimensional laminar flame simulations, comparing the distribution data of physical quantities (such as component content and temperature) in the calculation results with experimental values ​​or simulation values ​​of detailed mechanisms, and optimizing the following objective function:

[0075]

[0076] Among them, A k b k E k These are the pre-exponential factor, temperature exponent, and activation energy of the Arrhenius equation for the k-th reaction in the reaction mechanism to be optimized, respectively; these are also the variables to be optimized. ij，计算 (A k b k E k Y represents the data points calculated using this set of variables; ij，准确 These are accurate data points (usually the results calculated using detailed reaction mechanisms or experimental data).

[0077] Existing conventional methods involve using algorithms such as genetic algorithms, differential evolution, or particle swarm optimization for optimization. However, these algorithms suffer from significant randomness, making it impossible to theoretically prove that the obtained point is the optimal point.

[0078] The optimization of chemical reaction mechanisms is a typical non-convex problem, and the objective function varies greatly with variables. Furthermore, the problem size is often large, making the calculation of the objective function time-consuming and the derivative difficult to compute. In the field of non-convex optimization, coordinate descent based on line search is a commonly used algorithm, its advantage being that it can still find the minimum point for non-convex problems. For exact line search of non-convex problems, polynomial interpolation is a common method. However, these methods rely on the local approximation of the objective function within the search region, which is too complex for the objective function in chemical reaction optimization, failing to achieve ideal results.

[0079] In this embodiment, to better characterize the difference between the mechanism to be optimized and the detailed mechanism, the following objective function was designed for the optimization of the chemical reaction mechanism:

[0080]

[0081] Where m is the working condition number, if there are 6 working conditions to optimize, then the value range of m is 1, 2, 3, 4, 5, 6, meaning the target number is 6; F 1，m ~F 4，m The four sources of error used to characterize the result of the optimization mechanism and the detailed mechanism calculation for each operating condition are:

[0082] (1)F 1，m The error in the result of detailed mechanism calculations, specifically the error in reactants or products, is represented by the following formula:

[0083]

[0084] Where, τ s τ is the moment when the reaction begins. e I is the time when the reaction terminates. I This refers to the set of reactants or products to be considered here. For example, in the reaction of hydrogen and oxygen to produce water, the components include H, H2, O, O2, OH, H2O, H2O2, and HO2, which are numbered 1, 2, 3, 4, 5, 6, and 7 respectively. Generally, H2, O2, and H2O are considered the initial products and products, so the set I_I is {2, 4, 6}, because H2 is the second, O2 is the third, and H2O is the fifth; x i，mFor the m-th operating condition, the mole fraction of the i-th reactant or product at each moment in the calculation results is given. The superscript D indicates that the calculation results are from a detailed mechanism.

[0085] (2)F 2，m The error of intermediate products, used to characterize the error between the results and detailed mechanism calculations, is expressed by the following formula:

[0086]

[0087] Among them, I p This indicates the intermediate product error that needs to be considered here. For example, in the reaction mentioned above where hydrogen and oxygen produce water, OH, H₂O₂, and HO₂ can be considered important intermediate products. Therefore, set I... p It is {5,7,8}; This refers to the time when the j-th intermediate product reaches its peak value or the time when the generation / consumption rate reaches its peak value under the m-th operating condition.

[0088] (3)F 3，m The error in temperature data, used to characterize the error between the results and detailed mechanism calculations, is expressed by the following formula:

[0089]

[0090] Among them, T m Let be the temperature at each moment in the calculation results for the m-th operating condition.

[0091] (4)F 4，m The error in the peak rate of temperature rise, used to characterize the error between the results and detailed mechanistic calculations, is expressed by the following formula:

[0092]

[0093] Where, τ T，m This represents the moment when the rate of temperature change reaches its peak under the m-th operating condition.

[0094] Furthermore, in this embodiment, an optimization algorithm based on coordinate descent is used to minimize the aforementioned objective function. This invention extends the polynomial interpolation algorithm, proposing a line search strategy based on spline interpolation. When an approximation becomes unapproximate near the minimum of the objective function, the algorithm densifies the value at the expected minimum. Simultaneously, when densification fails to approximate the value, the algorithm's inherent characteristics ensure it exits the loop, preventing excessive computational waste. This line search strategy is applied to the coordinate descent method and ultimately to the optimization of chemical reaction mechanisms. Due to the robustness of this line search strategy and the coordinate descent method itself, it can better solve the complex problem of chemical reaction mechanism optimization. The flowchart of the spline interpolation-based line search strategy is described below. Figure 2 As shown, it specifically includes:

[0095] ① Enter the current position x (k) Search direction p (k) Initial number of points n init and maximum step size γ max ;

[0096] ② Take n uniformly within the search interval init From a set of points, we obtain a series of initial points X, with the left endpoint of the search interval being... The right endpoint is

[0097] ③ Calculate the objective function at a series of initial points X to obtain a series of objective function values ​​F;

[0098] ④ Take n uniformly within the search interval interp From these points, a series of interpolation points X are obtained. interp ;

[0099] ⑤ For the series of objective function values ​​F obtained in step ③, at this series of interpolation points X interp Spline interpolation is performed to obtain a series of interpolation objective function values ​​F. interp ;

[0100] ⑥ Find the initial point x corresponding to the minimum value among the series of objective function values ​​F obtained in step ③. t，min，0 ;

[0101] ⑦ Find the series of interpolation objective function values ​​F obtained in step ⑤ interp The interpolation point x corresponding to the minimum value t，min，1 And calculate x t，min，1 The objective function value f at the location t，min，1 ;

[0102] ⑧If x t，min，0 and x t，min，1 If the distance meets the preset requirement, the algorithm is considered to have converged, and x is returned. t，min，1 The process ends there; if it does not converge, the process is repeated with x. t，min，1 Add it to the initial point X, and set f t，min，1 Add it to the objective function value F, and jump back to step ④ to continue execution.

[0103] The experimental verification process of this invention is described below:

[0104] In this embodiment, the method of the present invention was experimentally compared with other conventional methods: a certain 45-step combustion reaction mechanism was optimized under 6 operating conditions using different methods. The 6 operating conditions covered lean, rich, and equivalence operating conditions under low and high pressure. The methods used included gradient descent, genetic algorithm, differential evolution, and the method of the present invention.

[0105] The optimization program runs on a personal computer; the process and results are as follows: Figure 3 and Figure 4 As shown; where: R45 represents the mechanism before optimization; GD represents gradient descent; GA represents genetic algorithm; DE represents differential evolution; CD represents the method of this invention; Figure 3 The diagram illustrates the descent process of the objective function during program execution. It can be seen that gradient descent converges to a relatively poor value very early on and cannot escape from the local optimum; while genetic algorithm and differential evolution method eventually converge to similar values; while the method of this invention (coordinate descent method) has the best optimization effect. Under the same running time, the objective function value is significantly lower than other methods, and the final objective function value obtained is an order of magnitude smaller than that of GA and DE methods.

[0106] In this embodiment, Figure 4 The diagram illustrates the components of the final objective function obtained from different optimization methods. The components of the objective function are defined as follows:

[0107] F m =F 1，m +F 2,m +F 3,m +F 4,m

[0108] Combination Figure 4 As shown, component 3 is the main reason why the objective function value of the other algorithms cannot continue to decrease. For components 3 and 6, the method of this invention is significantly better than other optimization methods; for component 2, it is slightly worse than other algorithms; the optimization effect of the remaining components is similar to other algorithms. The overall optimization effect is significantly better than other methods, and the order of magnitude difference between different components in the objective function is significantly smaller than that of other methods.

[0109] As described in the above embodiments, those skilled in the art will understand that the present invention provides a chemical reaction mechanism optimization method based on coordinate descent. In this invention, a line search strategy is applied to the coordinate descent method for chemical reaction mechanism optimization, which helps to make the calculation results closer to the detailed mechanism, while also offering faster speed, better computational accuracy, and robustness. The objective function designed for chemical reaction mechanism optimization in this invention can better characterize the difference between the mechanism to be optimized and the detailed mechanism. The line search strategy based on spline interpolation used in this invention, when an unapproximate value appears near the minimum value of the objective function, densifies the value at the expected minimum value. Furthermore, when densification fails to approximate the value, the strategy itself ensures that the loop will exit, preventing excessive computational time wastage. This invention applies this line search strategy to the coordinate descent method and ultimately to the optimization of chemical reaction mechanisms. Due to the robustness of this line search strategy and the coordinate descent method itself, it is beneficial to better solve the complex problem of chemical reaction mechanism optimization.

[0110] Furthermore, the present invention also provides a chemical reaction mechanism optimization system based on the coordinate descent method, applied to a chemical reaction mechanism optimization method based on the coordinate descent method in the above embodiments, for optimizing the chemical reaction mechanism. The system includes:

[0111] The input module is used to input the chemical reaction mechanism file and detailed reaction mechanism file that need to be optimized, as well as the set optimization conditions and important components;

[0112] The judgment module is used to determine the reaction type of each elementary reaction in the reaction mechanism and to determine the number of optimization variables.

[0113] The objective function generation module is used to generate an objective function based on the input optimization conditions and key components;

[0114] The sensitivity analysis module is used to perform sensitivity analysis on the objective function and exclude unimportant optimization variables;

[0115] The coordinate order determination module is used to determine the coordinate order based on the sensitivity analysis results and generate a search direction sequence, wherein the first search direction sequence is... One component is 1, and the other directional components are 0; where k represents the sequence number of the search direction sequence, and n represents the number of points;

[0116] The line search module is used to perform line search on the search direction sequence in sequence according to the spline interpolation line search strategy until the termination condition is met, and save the search results to the history file.

[0117] The output module is used to output the optimization mechanism file after optimization is completed.

[0118] The system provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned method embodiment, and will not be repeated here.

[0119] In addition, embodiments of the present invention also provide a storage medium storing one or more programs readable by a computing device. The one or more programs include instructions that, when executed by the computing device, cause the computing device to perform a chemical reaction mechanism optimization method based on coordinate descent as described in the above embodiments.

[0120] In embodiments of the present invention, the storage medium may be, for example, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples of storage media (a non-exhaustive list) include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, and any suitable combination thereof.

[0121] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products, etc. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects.

[0122] It should be noted that the word "comprising" does not exclude the presence of components or steps not listed in the claims. The words "a" or "an" preceding a component do not exclude the presence of a plurality of such components. This invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer.

[0123] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0124] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing chemical reaction mechanisms based on coordinate descent, characterized in that, The method includes the following steps: S1. Input the chemical reaction mechanism file and detailed reaction mechanism file to be optimized, as well as the set optimization conditions and important components; S2. Determine the reaction type of each elementary reaction in the reaction mechanism and determine the number of optimization variables; S3. Automatically generate an objective function based on the input optimization conditions and key components; the generated objective function is: in, These are the first of the reaction mechanisms that need to be optimized. k The Arrhenius equation for a reaction includes the pre-exponential factor, temperature index, and activation energy. m This is the operating condition number; The four sources of error in the calculation results of the mechanism to be optimized and the detailed mechanism are used to characterize the specific working conditions under each condition: (1) The error in the result of detailed mechanism calculations, specifically the error in reactants or products, is represented by the following formula: in, The moment when the reaction begins. The time when the reaction terminates. For the set of reactants or products to be considered, For the first m Under the first working condition, the calculation result is the first... i The mole fraction of each reactant or product at each moment, indicated by the superscript. D The calculation results show the detailed mechanism; (2) The error of intermediate products, used to characterize the error between the results and detailed mechanism calculations, is expressed by the following formula: in, This indicates the intermediate product error that needs to be considered; For the first m Under the first working condition, the first j The time when an intermediate product reaches its peak or the time when the production / consumption rate reaches its peak. (3) The error in temperature data, used to characterize the error between the results and detailed mechanistic calculations, is expressed by the following formula: in, For the first m Temperature at each moment in the calculation results under each working condition: (4) The error in the peak rate of temperature rise, used to characterize the error between the results and detailed mechanistic calculations, is expressed by the following formula: in, For the first m The moment when the rate of temperature change reaches its peak under each operating condition; S4. Conduct sensitivity analysis on the objective function to eliminate unimportant optimization variables; S5. Determine the coordinate order based on the sensitivity analysis results and generate a search direction sequence; S6. A line search strategy based on spline interpolation is used to perform line searches sequentially on the search direction sequence until the termination condition is met, and the search results are saved to the history file. S7. Optimization complete, output optimization mechanism file.

2. The method for optimizing chemical reaction mechanisms based on coordinate descent as described in claim 1, characterized in that, In step S2, the number of variables required for optimization of the elementary reaction is determined according to the reaction type, the number of variables required for the entire mechanism is determined by summing the number of variables for all elementary reactions, and a mapping relationship is established.

3. The method for optimizing chemical reaction mechanisms based on coordinate descent as described in claim 1, characterized in that, In step S4, the sensitivity of the optimization variables is sorted. First, the optimization variables with a sensitivity of 0 to all objectives are marked as variables that need to be skipped. Then, the root mean square of the sensitivity of all objectives is taken and sorted again. Variables with lower sensitivity are marked according to the number of inputs.

4. The chemical reaction mechanism optimization method based on coordinate descent as described in claim 3, characterized in that, In step S5, the coordinate order is determined based on the sensitivity analysis results. The order can be either the positive sensitivity order or the negative sensitivity order.

5. The method for optimizing chemical reaction mechanisms based on coordinate descent according to claim 1, characterized in that, In step S6, the line search strategy based on spline interpolation specifically includes: ① Enter the current location Search direction Initial points and maximum step size ; ② Take uniformly within the search interval From these points, we obtain a series of initial points. One endpoint of the search interval is The other end is ; ③ After obtaining a series of initial points The objective function is calculated, and a series of objective function values ​​are obtained. ; ④ Take uniformly within the search interval From these points, a series of interpolation points are obtained. ; ⑤ The series of objective function values ​​obtained in step ③ In this series of interpolation points Spline interpolation is performed to obtain a series of interpolation objective function values. ; ⑥ Obtain the series of objective function values ​​obtained in step ③. The initial point corresponding to the minimum value ; ⑦ Obtain the series of interpolation objective function values ​​obtained in step ⑤. Interpolation point corresponding to the minimum value and calculate objective function value at ; ⑧If and If the distance meets the preset requirement, the algorithm is considered to have converged and returns. If it does not converge, repeat the process. Add to initial point In the middle, Add to the target function value Then, jump back to step ④.

6. A chemical reaction mechanism optimization system based on coordinate descent method, characterized in that, A chemical reaction mechanism optimization method based on coordinate descent, as described in any one of claims 1-5, is used to optimize the chemical reaction mechanism. The system comprises: The input module is used to input the chemical reaction mechanism file and detailed reaction mechanism file that need to be optimized, as well as the set optimization conditions and important components; The judgment module is used to determine the reaction type of each elementary reaction in the reaction mechanism and to determine the number of optimization variables. The objective function generation module is used to generate an objective function based on the input optimization conditions and key components; The sensitivity analysis module is used to perform sensitivity analysis on the objective function and exclude unimportant optimization variables; The coordinate order determination module is used to determine the coordinate order based on the sensitivity analysis results and generate a search direction sequence; The line search module is used to perform line search on the search direction sequence in sequence according to the spline interpolation line search strategy until the termination condition is met, and save the search results to the history file. The output module is used to output the optimization mechanism file after optimization is completed.

Images