Catalyst active site optimization method and system based on adaptive simulated annealing algorithm

Abstract

The invention discloses a catalyst active site optimization method and system based on an adaptive simulated annealing algorithm, and the method comprises the following steps: S1, building an initial catalyst model according to the initial structure configuration of a catalyst active site; s2, setting initial parameters of a simulated annealing algorithm; s3, at the initial temperature T0, performing random fine adjustment on the catalyst model, and simulating variation of active sites of the catalyst; s4, dynamically adjusting parameters of the simulated annealing algorithm according to the search efficiency in the optimization process and the proximity of the globally optimal solution by adopting a self-adaptive adjustment mechanism; s5, repeating the steps S3 and S4 until the temperature is reduced to the final temperature Tend or the convergence condition of the globally optimal solution is met; and S6, determining the optimal structural configuration of the active sites of the catalyst according to the obtained global optimal solution. According to the method, the principles of computational chemistry and simulated annealing algorithms are fully utilized, and the process of optimizing the active sites of the catalyst through a self-adaptive mechanism is described in detail.

Description

Technical Field

[0001] The present invention relates to the field of catalyst technology, and in particular to a catalyst active site optimization method and system based on an adaptive simulated annealing algorithm. Background Art

[0002] In chemistry and materials science, catalyst design and optimization are key to achieving efficient chemical reactions. The performance of a catalyst depends significantly on the structural configuration of its active sites, including the atomic type, spatial position, and electronic structure of the active sites. Traditionally, optimization of catalyst active sites has relied on trial-and-error methods or intuitive design, which is time-consuming and inefficient. With the development of computational chemistry and computational physics, researchers have begun using various algorithms to simulate and optimize catalysts in order to find the optimal active site configuration.

[0003] The simulated annealing algorithm is a global optimization algorithm inspired by the physical annealing process. It simulates the heating and gradual cooling process to avoid being trapped in local optimal solutions and seek a global optimal solution. However, traditional simulated annealing algorithms face several core difficulties when applied to catalyst active site optimization:

[0004] 1. Fixed parameter setting problem: Traditional simulated annealing algorithms usually use fixed parameters, such as cooling rate, which makes the algorithm lack flexibility and adaptability when facing complex catalyst optimization problems, especially in high-dimensional and multi-variable optimization scenarios.

[0005] 2. Inefficient global search capability: Although the simulated annealing algorithm itself is designed to avoid local optimality, in practical applications, fixed parameter settings and the lack of an effective adaptive mechanism may still cause the algorithm to fall into the local optimal solution prematurely, especially in the complex optimization problems of catalyst active sites.

[0006] 3. The contradiction between optimization efficiency and accuracy: In traditional simulated annealing algorithms, achieving high-precision global optimal solution search often requires sacrificing a large amount of computing resources. This is not desirable in catalyst design because it is necessary to improve the efficiency of the optimization process while ensuring the depth and breadth of the search.

[0007] Therefore, a method and system for optimizing catalyst active sites that can adaptively adjust optimization parameters, improve global search capabilities, and achieve accurate optimization results while ensuring high efficiency is an urgent problem for those skilled in the art. This invention aims to address these challenges by introducing an adaptive simulated annealing algorithm and combining expertise in the fields of catalysts and active sites to propose a novel method and system for optimizing catalyst active sites. Summary of the Invention

[0008] One purpose of the present invention is to propose a catalyst active site optimization method and system based on an adaptive simulated annealing algorithm. The present invention fully utilizes the principles of computational chemistry and simulated annealing algorithm, and describes in detail the process of optimizing catalyst active sites through an adaptive mechanism.

[0009] According to an embodiment of the present invention, a catalyst active site optimization method based on an adaptive simulated annealing algorithm includes the following steps:

[0010] S1. Establishing an initial catalyst model based on the initial structural configuration of the catalyst active sites;

[0011] S2. Set the initial parameters of the simulated annealing algorithm, including the initial temperature T0, cooling rate r, and termination temperature T end ;

[0012] S3, at the initial temperature T0, the catalyst model is randomly fine-tuned to simulate the variation of the catalyst active sites;

[0013] S4, using an adaptive adjustment mechanism to dynamically adjust the parameters of the simulated annealing algorithm according to the search efficiency and the proximity to the global optimal solution during the optimization process;

[0014] S5, repeat steps S3 and S4 until the temperature drops to the end temperature T end , or meet the convergence conditions of the global optimal solution;

[0015] S6. Determine the optimal structural configuration of the catalyst active site based on the global optimal solution achieved.

[0016] Optionally, the S1 specifically includes:

[0017] S11. Collecting raw data of the catalyst active site, including atomic type, spatial position coordinates x, y, z, and electronic structure of the catalyst active site;

[0018] S12. Based on the collected raw data, use computational chemistry methods to calculate the initial energy state E of the catalyst model initial ;

[0019] S13. Using atomic position coordinates and electronic structure, an initial three-dimensional catalyst model is constructed using molecular simulation software.

[0020] Optionally, the S2 specifically includes:

[0021] S21, setting the initial temperature T0 of the simulated annealing algorithm. In order to explore a wide range of configuration spaces in the initial search phase, the initial temperature T0 should be higher than the critical temperature of the physical or chemical process in which the active sites of the catalyst undergo significant changes;

[0022] S22. Determine the initial value of the cooling rate r, where the cooling rate r is within the range of 0 < r < 1;

[0023] S23. Set the termination temperature T end , and when the system temperature drops below this value, terminate the optimization process.

[0024] Optionally, the S3 specifically includes:

[0025] S31. At the set initial temperature T0, randomly select one or more active sites in the catalyst model for fine-tuning, including replacing the atomic type of the active site, making a small offset in the spatial position, and adjusting the electronic structure;

[0026] S32. For the catalyst model after each fine-tuning, recalculate its energy state E using computational chemistry methods new [[ID=1...]];

[0027] S33. Adopt the Metropolis criterion. When ΔE = E new -E initial , then accept the new configuration; when ΔE > 0, then with a probability:

[0028]

[0029] where ΔE = E new -E initial represents the energy difference between the new configuration and the initial configuration, E new represents the energy state of the active site of the catalyst after fine-tuning, and E initial represents the energy state of the active site of the catalyst before fine-tuning, and kT·f(T,α,β) represents the relationship between temperature and energy, where T is the current simulation temperature and represents the "virtual" temperature during the optimization process. <00...]]

[0030] Optionally, the S4 specifically includes:

[0031] S41. Dynamically adjust the cooling rate r of the simulated annealing algorithm according to the data obtained during the optimization process, with the adaptive cooling rate adjustment function R(T, E var ): ...

[0032]

[0033] where T is the current temperature, E var is the variance of the energy change in the recent several iterations, r0 is the base cooling rate, and φ, γ, and μ are adjustment parameters used to control the sensitivity and speed of the adaptive adjustment;

[0034] S42. Evaluate the optimization efficiency at the current temperature through an efficiency evaluation function E(T,S), where S is the current exploration state of the search space, including the frequency of successfully finding new local optimal solutions and the magnitude of improvement:

[0035]

[0036] Where ΔE avg is the average energy improvement, N success is the number of attempts to successfully improve the solution at the current temperature, N total is the total number of attempts, while η and ζ are weight parameters;

[0037] S43. Based on the evaluation result of E(T,S), adaptively adjust the temperature drop rate r and the update strategy of the temperature T;

[0038] S44、Use feedback regulation mechanism (F(E(T,S),R(T,E var ), comprehensively consider the output of efficiency evaluation and cooling rate adjustment function, and dynamically adjust the search strategy and parameter settings:

[0039] F(E,R)=θ·E+(1-θ)·R;

[0040] Here, θ is a parameter between 0 and 1 that is used to balance the effects of optimization efficiency and cooling rate.

[0041] Optionally, the S5 specifically includes:

[0042] S51. During each iteration, record the energy state E of the catalyst model. current and corresponding catalyst structural configurations;

[0043] S52. Compare the energy state difference between two consecutive iterations:

[0044] ΔE=E current -E previous ;

[0045] Among them, E previous is the energy state of the previous iteration;

[0046] S53, when ΔE is less than the set threshold ∈, and the current temperature T is lower than or equal to the termination temperature T end , then the convergence condition of the global optimal solution is considered to be met and the optimization process is terminated;

[0047] S54: When the convergence condition is not met, the temperature T and other related parameters are updated according to the adaptive adjustment mechanism in step S4, and the next round of iteration is performed;

[0048] S55. Repeat steps S51 to S54 until the convergence condition is met or the maximum number of iterations is reached.

[0049] Optionally, the S6 specifically includes:

[0050] S61. After the optimization process is terminated, the state with the lowest energy E is selected from all iterative processes. min The catalyst model is taken as the final optimization result;

[0051] S62. Analyze the structural configuration of the optimal catalyst model, including the atomic type of the active site, spatial position coordinates x, y, z, and electronic structure;

[0052] S63. Based on the optimal catalyst model, calculate the final catalytic performance indicators, catalytic efficiency η and selectivity σ, where η and σ are calculated using the following formulas:

[0053]

[0054] Among them, E reactants 、E products 、E desiredproducts and E allproducts represent the energy states of reactants, all products, desired products, and all products respectively;

[0055] S64. Compare the calculated catalytic performance indicators with known data or expected targets to verify the performance improvement of the optimized catalyst model.

[0056] Optionally, a catalyst active site optimization system based on an adaptive simulated annealing algorithm includes:

[0057] The data input unit is used to receive the raw data of the catalyst active site, including the atomic type, spatial position coordinates x, y, z and electronic structure information, as well as the environmental parameters of the catalytic reaction, the energy state E of the reactants and products reactants 、E products 、E desiredproducts and E allproducts ;

[0058] Parameter setting unit, used to set the initial parameters of the simulated annealing algorithm, including initial temperature T0, cooling rate r, termination temperature T end and the initial value of the adaptive adjustment factor λ0;

[0059] The optimization processing unit performs the following operations: constructing an initial catalyst model based on the input catalyst raw data and calculating the initial energy state E initial Adaptive simulated annealing algorithm is used for iterative optimization, including random fine-tuning of catalyst active sites and calculation of the energy state E of the new configuration. new,Use the MS5 tropolis criterion to decide whether to accept the new configuration, and ,dynamically adjust the cooling rate r and the temperature drop strategy according to the ,search efficiency and the proximity to the global optimal solution during the ,optimization process; monitor the changes in the energy state and the ,temperature drop until the convergence condition is met or the maximum number of ,iterations is reached, terminating the optimization process;

[0060] The performance evaluation unit calculates the catalytic efficiency based on the optimal catalyst model and compares it with known data or expected targets to verify the effectiveness of the optimization results;

[0061] The output unit is used to display the optimal structural configuration of the optimized catalyst active site, including the atomic type, position coordinates and electronic structure of the active site, as well as the calculated catalytic efficiency η and selectivity σ.

[0062] The beneficial effects of the present invention are:

[0063] This paper introduces an adaptive regulation mechanism to dynamically adjust key parameters in the simulated annealing algorithm (such as the cooling rate and temperature reduction strategy) to meet the search requirements at different stages, significantly improving search efficiency and the probability of finding the global optimal solution. Furthermore, the paper describes in detail how to utilize the global optimization technology of simulated annealing to significantly improve the efficiency of the optimization process and the accuracy of the results while maintaining the depth and breadth of the search. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0065] Figure 1 This is an overall flow chart of a catalyst active site optimization method based on an adaptive simulated annealing algorithm proposed in the present invention. DETAILED DESCRIPTION

[0066] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0067] refer to Figure 1 , a catalyst active site optimization method based on an adaptive simulated annealing algorithm, comprising the following steps:

[0068] S1. Establishing an initial catalyst model based on the initial structural configuration of the catalyst active sites;

[0069] In this embodiment, S1 specifically includes:

[0070] S11. Collect the original data of the catalyst active sites, including the atomic type, spatial position coordinates x, y, z, and the electronic structure of the catalyst active sites;

[0071] S12. Based on the collected original data, use computational chemistry methods to calculate the initial energy state E of the catalyst model initial ;

[0072] S13. Utilize the atomic position coordinates and electronic structure to construct an initial three-dimensional catalyst model through molecular simulation software.

[0073] S2. Set the initial parameters of the simulated annealing algorithm, including the initial temperature T0, the cooling rate r, and the termination temperature T end ;

[0074] In this embodiment, S2 specifically includes:

[0075] S21. Set the initial temperature T0 of the simulated annealing algorithm. In order to explore a wide configuration space in the initial search stage, the initial temperature T0 should be higher than the critical temperature of the physical or chemical process in which significant changes occur in the catalyst active sites;

[0076] S22. Determine the initial value of the cooling rate r, where the cooling rate r is within the range of 0 < r < 1;

[0077] S23. Set the termination temperature T end , and when the system temperature drops below this value, terminate the optimization process.

[0078] S3. At the initial temperature T0, perform random fine-tuning on the catalyst model to simulate the mutation of the catalyst active sites;

[0079] In this embodiment, S3 specifically includes:

[0080] S31. At the set initial temperature T0, randomly select one or more active sites in the catalyst model for fine-tuning, including replacing the atomic type of the active site, making a small offset in the spatial position, and adjusting the electronic structure;

[0081] S32. For the catalyst model after each fine-tuning, use computational chemistry methods to recalculate its energy state E new ;

[0082] S33. Adopt the Metropolis criterion. When ΔE = E new - E initial , then accept the new configuration; when ΔE > 0, then with a probability:

[0083]

[0084] Where, ΔE = Enew -E initial , represents the energy difference between the new configuration and the initial configuration, E new represents the energy state of the active site of the catalyst after fine-tuning, E initial represents the energy state of the active site of the catalyst before fine-tuning, kT·f(T,α,β) represents the relationship between temperature and energy, and T is the current simulation temperature, which represents the “virtual” temperature during the optimization process.

[0085] S4, using an adaptive adjustment mechanism to dynamically adjust the parameters of the simulated annealing algorithm according to the search efficiency and the proximity to the global optimal solution during the optimization process;

[0086] In this embodiment, S4 specifically includes:

[0087] S41, dynamically adjust the cooling rate r of the simulated annealing algorithm according to the data obtained during the optimization process, and adaptively adjust the cooling rate function R(T,E var ):

[0088]

[0089] Where T is the current temperature, E var is the variance of energy changes in the last few iterations, r0 is the base cooling rate, φ, γ and μ are tuning parameters used to control the sensitivity and speed of adaptive adjustment;

[0090] S42. Evaluate the optimization efficiency at the current temperature through an efficiency evaluation function E(T,S), where S is the current exploration state of the search space, including the frequency of successfully finding new local optimal solutions and the magnitude of improvement:

[0091]

[0092] Where ΔE avg is the average energy improvement, N success is the number of attempts to successfully improve the solution at the current temperature, N total is the total number of attempts, while η and ζ are weight parameters;

[0093] S43. Based on the evaluation result of E(T,S), adaptively adjust the temperature drop rate r and the update strategy of the temperature T;

[0094] S44、Use feedback regulation mechanism (F(E(T,S),R(T,E var ), comprehensively consider the output of efficiency evaluation and cooling rate adjustment function, and dynamically adjust the search strategy and parameter settings:

[0095] F(E,R)=θ·E+(1-θ)·R;

[0096] Here, θ is a parameter between 0 and 1 that is used to balance the effects of optimization efficiency and cooling rate.

[0097] S5, repeat steps S3 and S4 until the temperature drops to the end temperature T end , or meet the convergence conditions of the global optimal solution;

[0098] In this embodiment, S1 specifically includes:

[0099] S51. During each iteration, record the energy state E of the catalyst model. current and corresponding catalyst structural configurations;

[0100] S52. Compare the energy state difference between two consecutive iterations:

[0101] ΔE=E current -E previous ;

[0102] Among them, E previous is the energy state of the previous iteration;

[0103] S53, when ΔE is less than the set threshold ∈, and the current temperature T is lower than or equal to the termination temperature T end , then the convergence condition of the global optimal solution is considered to be met and the optimization process is terminated;

[0104] S54: When the convergence condition is not met, the temperature T and other related parameters are updated according to the adaptive adjustment mechanism in step S4, and the next round of iteration is performed;

[0105] S55. Repeat steps S51 to S54 until the convergence condition is met or the maximum number of iterations is reached.

[0106] S6. Determine the optimal structural configuration of the catalyst active site based on the global optimal solution achieved.

[0107] In this embodiment, S6 specifically includes:

[0108] S61. After the optimization process is terminated, the state with the lowest energy E is selected from all iterative processes. min The catalyst model is taken as the final optimization result;

[0109] S62. Analyze the structural configuration of the optimal catalyst model, including the atomic type of the active site, spatial position coordinates x, y, z, and electronic structure;

[0110] S63. Based on the optimal catalyst model, calculate the final catalytic performance indicators, catalytic efficiency η and selectivity σ, where η and σ are calculated using the following formulas:

[0111]

[0112] Among them, E reactants 、E products 、E desiredproducts and E allproducts represent the energy states of reactants, all products, desired products, and all products respectively;

[0113] S64. Compare the calculated catalytic performance indicators with known data or expected targets to verify the performance improvement of the optimized catalyst model.

[0114] In this embodiment, a catalyst active site optimization system based on an adaptive simulated annealing algorithm includes:

[0115] The data input unit is used to receive the raw data of the catalyst active site, including the atomic type, spatial position coordinates x, y, z and electronic structure information, as well as the environmental parameters of the catalytic reaction, the energy state E of the reactants and products reactants 、E products 、E desiredproducts and E allproducts ;

[0116] Parameter setting unit, used to set the initial parameters of the simulated annealing algorithm, including initial temperature T0, cooling rate r, termination temperature T end and the initial value of the adaptive adjustment factor λ0;

[0117] The optimization processing unit performs the following operations: constructing an initial catalyst model based on the input catalyst raw data and calculating the initial energy state E initial Adaptive simulated annealing algorithm is used for iterative optimization, including random fine-tuning of catalyst active sites and calculation of the energy state E of the new configuration. new ,Use the MS5 tropolis criterion to decide whether to accept the new configuration, and ,dynamically adjust the cooling rate r and the temperature drop strategy according to the ,search efficiency and the proximity to the global optimal solution during the ,optimization process; monitor the changes in the energy state and the ,temperature drop until the convergence condition is met or the maximum number of ,iterations is reached, terminating the optimization process;

[0118] The performance evaluation unit calculates the catalytic efficiency based on the optimal catalyst model and compares it with known data or expected targets to verify the effectiveness of the optimization results;

[0119] The output unit is used to display the optimal structural configuration of the optimized catalyst active site, including the atomic type, position coordinates and electronic structure of the active site, as well as the calculated catalytic efficiency η and selectivity σ.

[0120] Example 1:

[0121] In early April 2023, the R&D team began collecting detailed data on the active sites of the currently used catalysts, including atomic types, spatial coordinates, and electronic structure information. Based on this information, the team constructed an initial 3D catalyst model and set the initial parameters for the simulated annealing algorithm: an initial temperature of 1000K, a cooling rate of 0.95, and an end temperature of 300K.

[0122] Optimization process

[0123] Starting on April 5th, the R&D team ran an optimization program based on an adaptive simulated annealing algorithm. This program automatically fine-tuned the active sites and recalculated the energy state, while dynamically adjusting the cooling rate and temperature ramp strategy based on the results of each iteration. Throughout the optimization process, this adaptive adjustment mechanism effectively guided the algorithm to explore a broad configuration space in search of the optimal catalyst active site configuration.

[0124] Optimization results and effect verification

[0125] By April 19th, the optimization process was complete. The resulting catalyst model showed that the optimized catalyst's methanol yield increased from 250 tons per ton of catalyst to 290 tons, and selectivity increased from 92% to 98%. Compared to traditional methods, which took an average of three months and had limited yield improvements, this method not only shortened the optimization cycle to less than a month, but also significantly improved both yield and selectivity.

[0126] Table 1 records the key parameters and performance indicators of the catalyst before and after optimization

[0127]

[0128] As shown in Table 1, time efficiency is significantly improved: the traditional method requires a three-month optimization cycle, while this method only takes less than one month, greatly accelerating the development and optimization of catalysts.

[0129] Substantial performance improvements: Through optimization, methanol yield and selectivity increased by 16% and 6%, respectively, directly improving production efficiency and product quality.

[0130] Improved economic benefits: Shortened R&D cycles and improved catalyst performance enable companies to respond to market demands more quickly, reduce R&D costs, and improve economic benefits.

[0131] The specific data and actual application results in this example fully demonstrate the effectiveness, practicality and efficiency of the method of the present invention, and provide a new technical path for the field of catalyst optimization.

[0132] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A catalyst active site optimization method based on an adaptive simulated annealing algorithm, characterized in that: It includes the following steps: S1. Establish an initial catalyst model according to the initial structural configuration of the catalyst active sites; S2. Set the initial parameters of the simulated annealing algorithm, including the initial temperature T0, cooling rate r, and termination temperature T end ; S3. At the initial temperature T0, randomly fine-tune the catalyst model to simulate the variation of the catalyst active sites; S4. Adopt an adaptive adjustment mechanism to dynamically adjust the parameters of the simulated annealing algorithm according to the search efficiency and proximity to the global optimal solution during the optimization process; S5, repeat steps S3 and S4 until the temperature drops to the end temperature T end , or meet the convergence conditions of the global optimal solution; S6. Determine the optimal structural configuration of the catalyst active sites according to the achieved global optimal solution.

2. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S1 includes: S11. Collect the original data of the catalyst active sites, including atomic type, spatial position coordinates x, y, z, and the electronic structure of the catalyst active sites; S12. Based on the collected raw data, use computational chemistry methods to calculate the initial energy state E of the catalyst model initial ; S13. Use the atomic position coordinates and electronic structure to construct an initial three-dimensional catalyst model through molecular simulation software.

3. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S2 includes: S21. Set the initial temperature T0 of the simulated annealing algorithm. In order to explore a wide configuration space in the initial search stage, the initial temperature T0 should be higher than the critical temperature of the physical or chemical process in which significant changes occur in the catalyst active sites; S22. Determine the initial value of the cooling rate r, where the cooling rate r is within the range of 0 < r < 1; S23, set the end temperature T end , when the system temperature drops below this value, the optimization process is terminated.

4. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S3 includes: S31. At the set initial temperature T0, randomly select one or more active sites in the catalyst model for fine-tuning, including the replacement of atomic type of the active sites, small offsets in spatial position, and adjustment of electronic structure; S32. For each fine-tuned catalyst model, recalculate its energy state E using computational chemistry methods. new ; S33, using the Metropolis criterion, when ΔE=E new -E initial , then accept the new configuration; when ΔE>0, then with probability: Where ΔE=E new -E initial , represents the energy difference between the new configuration and the initial configuration, E new represents the energy state of the active site of the catalyst after fine-tuning, E initial represents the energy state of the active site of the catalyst before fine-tuning, kT·f(T,α,β) represents the relationship between temperature and energy, and T is the current simulation temperature, which represents the "virtual" temperature during the optimization process.

5. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S4 includes: S41, dynamically adjust the cooling rate r of the simulated annealing algorithm according to the data obtained during the optimization process, and adaptively adjust the cooling rate function R(T,E var ): Where T is the current temperature, E var is the variance of energy changes in the last few iterations, r0 is the base cooling rate, φ, γ and μ are tuning parameters used to control the sensitivity and speed of adaptive adjustment; S42. Evaluate the optimization efficiency at the current temperature through an efficiency evaluation function E(T, S), where S is the current exploration state of the search space, including the frequency and improvement amplitude of successfully finding new local optimal solutions: Where ΔE avg is the average energy improvement, N success is the number of attempts to successfully improve the solution at the current temperature, N total is the total number of attempts, while η and ζ are weight parameters; S43. Based on the evaluation result of E(T, S), adaptively adjust the temperature decrease rate r and the update strategy of temperature T; S44、Use feedback regulation mechanism (F(E(T,S),R(T,E var ), comprehensively consider the output of efficiency evaluation and cooling rate adjustment function, and dynamically adjust the search strategy and parameter settings: F(E, R) = θ·E + (1 - θ)·R; Among them, θ is a parameter between 0 and 1, which is used to balance the influence of optimization efficiency and cooling rate.

6. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S5 includes: S51. During each iteration, record the energy state E of the catalyst model. current and corresponding catalyst structural configurations; S52. Compare the energy state difference between two consecutive iterations: ΔE=E current -E previous ; Among them, E previous is the energy state of the previous iteration; S53, when ΔE is less than the set threshold ∈, and the current temperature T is lower than or equal to the termination temperature T end , then the convergence condition of the global optimal solution is considered to be met and the optimization process is terminated; S54. When the convergence condition is not met, update the temperature T and other relevant parameters according to the adaptive adjustment mechanism in step S4, and perform the next round of iteration; S55. Repeat steps S51 to S54 until the convergence condition is met or the maximum iteration number limit is reached.

7. The catalyst active site optimization method based on the adaptive simulated annealing algorithm according to claim 1, characterized in that: The specific content of S6 includes: S61. After the optimization process is terminated, the state with the lowest energy E is selected from all iterative processes. min The catalyst model is taken as the final optimization result; S62. Analyze the structural configuration of the optimal catalyst model, including the atomic type, spatial position coordinates x, y, z, and electronic structure of the active sites; S63. Based on the optimal catalyst model, calculate the final catalytic performance indicators, catalytic efficiency η and selectivity σ, where η and σ are calculated through the following formulas: Among them, E reactants 、E products 、E desiredproducts and E allproducts represent the energy states of reactants, all products, desired products, and all products respectively; S64. Compare the calculated catalytic performance indicators with known data or expected targets to verify the performance improvement of the optimized catalyst model.

8. The catalyst active site optimization system based on the adaptive simulated annealing algorithm according to claim 1, which executes the catalyst active site optimization method based on the adaptive simulated annealing algorithm according to any one of claims 1 to 7, characterized in that: It includes: The data input unit is used to receive the raw data of the catalyst active site, including the atomic type, spatial position coordinates x, y, z and electronic structure information, as well as the environmental parameters of the catalytic reaction, the energy state E of the reactants and products reactants 、E products 、E desiredproducts and E allproducts ; Parameter setting unit, used to set the initial parameters of the simulated annealing algorithm, including initial temperature T0, cooling rate r, termination temperature T end and the initial value of the adaptive adjustment factor λ0; The optimization processing unit performs the following operations: constructing an initial catalyst model based on the input catalyst raw data and calculating the initial energy state E initial Adaptive simulated annealing algorithm is used for iterative optimization, including random fine-tuning of catalyst active sites and calculation of the energy state E of the new configuration. new ,Use the MS5 tropolis criterion to decide whether to accept the new configuration, and ,dynamically adjust the cooling rate r and the temperature drop strategy according to the ,search efficiency and the proximity to the global optimal solution during the ,optimization process; monitor the changes in the energy state and the ,temperature drop until the convergence condition is met or the maximum number of ,iterations is reached, terminating the optimization process; A performance evaluation unit, which calculates the catalytic efficiency based on the optimal catalyst model and compares it with known data or expected targets to verify the effectiveness of the optimization result; The output unit is used to display the optimal structural configuration of the optimized catalyst active site, including the atomic type, position coordinates and electronic structure of the active site, as well as the calculated catalytic efficiency η and selectivity σ.

Images