A Phase Equilibrium Calculation Method Based on Constrained Grey Prediction Evolutionary Algorithm
By applying a constrained gray prediction evolution algorithm in NVT flash calculation, it is independent of the initial value and avoids derivatives, and solves the problems of sensitivity and complex derivatives, and realizes stable phase equilibrium calculation and efficient iterative solution.
Patent Information
- Application Number
- CN202410143292.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-31
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-01-31
AI Technical Summary
The existing NVT flash calculation methods are sensitive to initial values and require complex derivative calculations, which easily converge to pseudo-solving or local minimum values.
Using a constraint-based gray prediction evolution algorithm, through population iterative optimization, the derivative of the objective function is avoided, independent of the initial value, the total Helmholtz free energy is used as the objective function, a mathematical model is constructed and constraints are added, and the iterative solution is solved. Obtain phase equilibrium state.
Phase balance calculation under NVT conditions is realized, pseudo-understanding and local minimum convergence are avoided, the stability and efficiency of calculation are improved, and the complexity of derivative calculation is reduced.
Smart Images

Figure CN118053512B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of phase equilibrium calculation, and more specifically, to a phase equilibrium calculation method at a given molar volume and temperature based on a constrained grey prediction evolutionary algorithm. Background Art
[0002] In order to study complex fluid behavior, it is often necessary to numerically solve multiphase and multicomponent fluids in geological structures. Phase equilibrium calculation, also known as flash calculation, is an important part of the numerical simulation of multicomponent and multiphase fluids in porous media and plays an important role in the field of petrochemicals. The purpose of phase equilibrium calculation is to calculate the physical properties of the mixture when it reaches equilibrium, including the composition, quantity and density of the phases after phase splitting.
[0003] At present, the flash calculation methods in engineering calculations are mainly divided into two categories according to different given conditions: NPT flash calculation and NVT flash calculation. NPT flash calculation is a phase equilibrium calculation under given moles (N), pressure (P) and temperature (T). This method was proposed very early and has been widely used. However, since a cubic state equation usually needs to be solved under NPT conditions, it is impossible to uniquely determine whether the component is a gas phase, a liquid phase, or a mixture of two phases. In addition, many solutions are required before reaching the equilibrium pressure, so the efficiency of NPT flash calculation is low. The phase equilibrium calculation problem (NVT flash calculation) at constant volume, temperature and moles proposed by Mikyska and Firoozabadi overcomes the defects of NPT flash calculation. This problem can be expressed as solving the minimum value of the total Helmholtz free energy function with highly nonlinear characteristics.
[0004] The existing methods for solving NVT flash calculations can be mainly divided into two categories. The first type of methods are direct solutions based on Newton's method. In 2011, Mikyska and Firoozabadi proposed a successive substitution iteration (SSI) method based on Newton's method, which solved the NVT flash calculation problem for the first time. Later, Jindrova and Mikyska proposed an algorithm for solving the minimum of the total Helmholtz free energy constraint, which used Newton's method, linear search technology and Cholesky decomposition of the Hessian matrix. Subsequently, they extended the previous work and applied it to multiphase NVT flash calculations. In addition, Lu et al. improved the SSI algorithm from different perspectives to enhance the robustness of SSI. Nichita, Castier, Mikyska et al. studied the phase stability analysis used to initialize the NVT phase splitting calculation. The second type of method is an indirect solution based on thermodynamic principles. This method uses thermodynamic principles to construct a kinetic model consisting of nonlinear equations to indirectly solve the minimum of the objective function. In 2016, Kou et al. proposed an energy-stable evolutionary method using Fick's diffusion law and the Peng-Robinson equation of state to model the dynamics. In 2018, Kou and Sun et al. established a kinetic model using the laws of thermodynamics and the Onsager reciprocal relation and developed a thermodynamically stable numerical algorithm to solve the NVT problem under the influence of capillary pressure. Building on previous work, Li and Kou proposed a thermodynamically stable numerical algorithm by introducing a generalized Onsager coefficient matrix. In 2020, Kou et al. proposed a new energy factorization (EF) method and proposed a linear semi-implicit numerical method that retains the energy decay property. In 2022, Feng et al. proposed a fully explicit and unconditional energy-stable method to deal with multi-component problems in NVT flash calculations.
[0005] Although both of the above two methods can find satisfactory solutions to the NVT flash calculation problem, the existing technologies still have shortcomings: (1) They are sensitive to initial values. This is because they are highly dependent on the initial values provided by the phase stability analysis; otherwise, both methods may converge to a pseudo-solution, i.e., a single phase or a local minimum. (2) These traditional methods all require solving the derivative of the objective function, which is very complicated to calculate. Summary of the invention
[0006] The present invention aims at solving the technical problems existing in the prior art and provides a phase equilibrium calculation method based on a constrained grey prediction evolutionary algorithm, comprising:
[0007] Collecting data sets for NVT flash calculations, the data sets are pure substances and mixtures;
[0008] Taking the total Helmholtz free energy as the objective function and the molar vector and volume of one phase as the decision variables, a mathematical model for NVT flash calculation is constructed, and the constraints of the mathematical model for NVT flash calculation are defined.
[0009] The parameters of the constrained grey prediction evolutionary algorithm are set, and the mathematical model of NVT flash calculation is iteratively solved based on the constrained grey prediction evolutionary algorithm to obtain the phase diagram when phase equilibrium is reached, as well as the total Helmholtz free energy convergence diagram as the number of iterations increases. The phase diagram includes the mole fractions of the two phases, phase mole densities, saturation and pressure at different total molar densities.
[0010] The present invention provides a phase equilibrium calculation method based on a constrained grey prediction evolutionary algorithm, which adds the constraints in the NVT problem to the original operation steps of GPE, and proposes an NVT flash calculation method based on a constrained grey prediction evolutionary algorithm (CGPE). The present invention overcomes the shortcomings of the existing methods that they are highly dependent on the initial values provided by phase stability analysis and are sensitive to the initial values. In addition, unlike traditional optimization methods that require the derivation of the objective function, the present algorithm is an optimization method based on population iteration, which avoids complex derivation calculations. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] Figure 1 A flow chart of a phase equilibrium calculation method based on a constrained grey prediction evolutionary algorithm provided by the present invention;
[0012] Figure 2 Schematic diagram of the solution process of constrained grey prediction evolutionary algorithm. DETAILED DESCRIPTION
[0013] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not subject to the constraints of the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0014] Compared with the solution of two types of NVT flash calculation problems in the background technology, the meta-heuristic optimization algorithm can be independent of the problem information and does not depend on the initial value; at the same time, the meta-heuristic optimization algorithm is an algorithm based on population iteration and does not need to derive the objective function. Grey prediction evolution algorithm (GPE) is an emerging meta-heuristic optimization algorithm with strong global search capability and insensitivity to the initial population.
[0015] Based on this, the present invention provides a phase equilibrium calculation method based on a constrained grey prediction evolutionary algorithm, such as Figure 1 As shown, the method includes:
[0016] Step 1: Collect data sets for NVT flash calculation, which are pure substances and mixtures.
[0017] It is understandable that the present invention collects six data sets of pure n-butane, pure carbon dioxide, 35% methane and 65% pentane, 54.7413% methane and 45.2587% pentane, 55% methane and 45% decane, 50% methane, 30% pentane and 20% decane, and these six sets of experiments test the algorithm.
[0018] Step 2, taking the total Helmholtz free energy as the objective function and the molar vector and volume of one phase as the decision variables, a mathematical model for NVT flash calculation is constructed, and the constraints of the mathematical model for NVT flash calculation are defined.
[0019] It is understandable that obtaining the phase equilibrium calculation knowledge used in the present invention is conducive to the development of subsequent work. By reading relevant literature, we can understand the mathematical model of NVT flash calculation. From the literature, we know that this problem can be expressed as solving the minimum value of the total Helmholtz free energy H, with the molar vector N of the gas phase G and volume V G is the decision variable. It is necessary to satisfy the volume and mole constraints, and the prerequisites for correctly defining the Helmholtz free energy f(n), chemical potential μ and pressure P, and use the direct search method and the exterior point method to handle the constraints. At the same time, it is necessary to satisfy the convergence condition of phase equilibrium, namely ||Δμ|| rel <10 -6 and |ΔP|<10 -6 .
[0020] For a mixture system containing M (M ≥ 1) components, if the mixture system is unstable under given conditions, it will split into gas and liquid phases. Let n k and N k is the molar density and molar number of the kth component, k = 1, 2, ..., M, and the molar density vector and molar vector are expressed as n = [n1, n2, ..., n M ] Tand N=[N1,N2,…,N M ] T , the total number of moles of a given substance Recorded as The total Helmholtz free energy is denoted by H, and f(n) is the Helmholtz free energy density of the uniform fluid. The total volume V calculated by NVT flash evaporation is total Total molar N total and temperature T are both constant; G and L represent the gas phase and liquid phase, respectively.
[0021] Therefore, for a two-phase system, the volumes of the gas phase and the liquid phase are denoted by V and G and V L , the moles of gas phase and liquid phase are expressed as N G and N L , the molar density of the gas phase and the liquid phase are denoted by n L and n G .
[0022] Two constraints define the mathematical model of NVT flash calculation, where:
[0023] Constraint 1: The following constraints on moles and volumes must be met in the NVT flash calculation:
[0024] VG+VL=Vtotal
[0025] NG+NL=Ntotal
[0026] VG≥0, VL≥0, NG≥0, NL≥0
[0027] In gas phase molar N G and volume V G ,but and V G ∈[0,V total ];
[0028] Based on the correct definition of Helmholtz free energy, chemical potential and pressure, constraint 2 is defined, and the exterior point method, i.e., penalty function method, is used to process constraint 2, where constraint 2 is:
[0029]
[0030]
[0031] Among them, parameter b k It can be obtained from the universal gas constant R, critical temperature T ck , and the critical pressure P ck The calculation results are:
[0032]
[0033] The following relative errors of chemical potential and pressure are required to meet the convergence criteria:
[0034]
[0035]
[0036] The mathematical model for NVT flash calculation is:
[0037]
[0038]
[0039] Step 3, setting the parameters of the constrained grey prediction evolutionary algorithm, iteratively solving the mathematical model of NVT flash calculation based on the constrained grey prediction evolutionary algorithm, obtaining the phase diagram when phase equilibrium is reached, and the total Helmholtz free energy convergence diagram as the number of iterations increases, the phase diagram including the mole fractions, phase mole densities, saturation and pressure of the two phases at different total molar densities.
[0040] Among them, the initialization parameters of the constrained grey prediction evolution algorithm are first set, including the population size of the constrained grey prediction evolution algorithm CGPE is 100, the initial value of the maximum number of iterations tmax is set to 100, and the step size λ is 20. Other parameters used in the present invention are set according to the parameters generally given in the existing literature. In order to reduce the error caused by randomness, each group of experiments is independently run 30 times.
[0041] The mathematical model of NVT flash calculation is solved based on the constrained grey prediction evolutionary algorithm. The solution process is as follows: Figure 2 As shown, it mainly includes:
[0042] 51. In the feasible region of mole and volume, randomly generate the first three generations of populations, process each individual in the first three generations of populations with constraint condition 2, and calculate the objective function value H for the individuals that satisfy constraint condition 2;
[0043] 52. Randomly select three individuals from three consecutive generations of population as a group of selected individuals, wherein an individual is randomly selected from each generation of population, a test individual is generated based on the randomly selected group of individuals, and a test population is generated based on the randomly selected multiple groups of individuals;
[0044] 53. Constraint condition 2 is processed for each individual in the test population, and the objective function value H is calculated for the test individuals that meet constraint condition 2;
[0045] 54. By comparing the objective function value H of the target individual and the test individual, the better individual is selected to enter the next generation;
[0046] 55. Determine whether the current number of iterations t is less than the maximum number of iterations t max If yes, then t=t+1, then execute 52, if no, then calculate the relative error of pressure and chemical potential energy, and judge whether the relative error of pressure and chemical potential energy converges to the standard, if yes, then output the global optimal value, if no, then t max =t max +λ, execute 51.
[0047] Specifically, the constrained grey prediction evolutionary algorithm solves the mathematical model of NVT flash calculation, which is mainly divided into three steps:
[0048] (1) Population initialization.
[0049] In the initialization process of the constrained grey prediction evolution algorithm CGPE, three generations of populations must be randomly generated in the feasible region. The jth dimension of the ith individual in the tth generation population is denoted as It can be generated by the following formula:
[0050]
[0051] Where D is the dimension of the problem; up j and low j are the upper and lower bounds of the jth dimension respectively; rand(0,1) represents a random number generated by a uniform distribution in the range [0,1]. The first three generations of populations generated need to be processed with constraint 2, and then the objective function value H is calculated for the individuals that meet constraint 2; for individuals that do not meet constraint 2, they are regenerated in the feasible region until all individuals in the three generations of populations generated meet constraint 2, and the objective function value H of each individual that meets constraint 2 is calculated.
[0052] (2) Reproduction operation
[0053] The reproduction operation of CGPE regards the population sequence as a time series, converts the three disordered original data into a sequence with the properties of an approximate exponential function, and predicts the offspring by constructing an exponential function. t-2 , X t -1 , X t (t≥2) is regarded as a population sequence, and from the population sequence X t-2 , X t-1 , X t (t≥2) randomly select three individuals and make Maxd r,j =max{d 12,j ,d13,j ,d 23,j},Mind r,j =min{d 12,j ,d 13,j ,d 23,j}. Test individuals It can be generated by the following formula:
[0054]
[0055] Among them, δ is a parameter for controlling the difference threshold of prediction, α, β and θ are the gray development coefficient, gray control parameter and disturbance coefficient respectively. The formulas for calculating α, β and θ are as follows:
[0056]
[0057]
[0058] θ=rand(-1,1)·δ.
[0059] For example, each generation of the population includes 100 individuals. For three consecutive generations of populations, one individual is randomly selected from the first generation population, one individual is randomly selected from the second generation population, and one individual is randomly selected from the third generation population. The three-body individuals are selected as a group of individuals, and a test individual is generated based on the group of individuals. Then, 100 groups of individuals are selected from the three-generation population to generate 100 test individuals.
[0060] Each individual in the generated experimental population must be processed with constraint 2. If constraint 2 is met, the individual's function objective value H is directly calculated. If constraint 2 is not met, the individual is regenerated in the feasible area until every individual in the experimental population meets constraint 2, and the objective function value H of each experimental individual in the experimental population is calculated.
[0061] (3)Select an operation.
[0062] The greedy selection mechanism is used as the selection operator of CGPE. and experimental individuals The objective function value is used to select the better individuals to enter the next generation. The selection operation of CGPE can be performed through the following formula:
[0063]
[0064] Keep looping and iterating to determine whether the current number of iterations t is less than the maximum number of iterations t max If yes, continue iterating. If the current number of iterations is greater than or equal to the maximum number of iterations t max, then the relative error of pressure and chemical potential energy is calculated ||Δμ|| rel and ||ΔP|| rel , and determine whether the relative errors of pressure and chemical potential meet the convergence criteria, that is, whether ||Δμ|| rel <10 -6 and |ΔP|<10 -6 If yes, then output the global optimal value, if no, then t max =t max +λ, and re-execute the steps of generating the initial three-generation population, re-iterate the solution, and finally output the optimal individual.
[0065] This invention is the first to apply a heuristic optimization algorithm to solve the NVT flash calculation problem. Therefore, six groups of experiments were selected from existing literature to test the effectiveness of this method in solving this problem. The experimental results show that CGPE has good convergence in all cases and is in good agreement with the experimental results obtained in existing literature, proving the effectiveness of this invention in solving the NVT flash calculation problem.
[0066] The present invention provides a method for calculating phase equilibrium based on a constrained grey prediction evolution algorithm, which has the following beneficial effects:
[0067] (1) For the first time, the heuristic optimization algorithm was applied to the field of phase equilibrium calculation. Numerical experiments demonstrated the effectiveness of the proposed algorithm in solving the NVT flash calculation problem.
[0068] (2) The constrained grey prediction evolutionary algorithm is applied to solve the NVT flash calculation problem, which overcomes the shortcomings of existing methods that are highly dependent on the initial values provided by phase stability analysis and are sensitive to initial values.
[0069] (3) Unlike traditional optimization methods that require the derivation of the objective function, the constrained grey prediction evolutionary algorithm is an optimization method based on population iteration, which avoids complex derivation calculations.
[0070] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0071] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0072] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0073] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0074] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0075] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
[0076] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A phase equilibrium calculation method based on constrained grey prediction evolution algorithm, It is characterized in that include: Collecting data sets for NVT flash calculations, the data sets are pure substances and mixtures; Taking the total Helmholtz free energy as the objective function and the molar vector and volume of one phase as the decision variables, a mathematical model for NVT flash calculation is constructed, and the constraints of the mathematical model for NVT flash calculation are defined. Setting parameters of the constrained grey prediction evolutionary algorithm, iteratively solving the mathematical model of NVT flash calculation based on the constrained grey prediction evolutionary algorithm, obtaining a phase diagram when phase equilibrium is reached, and obtaining a total Helmholtz free energy convergence diagram as the number of iterations increases, wherein the phase diagram includes mole fractions of two phases, phase mole densities, saturation and pressure at different total molar densities; The total Helmholtz free energy is used as the objective function, the molar vector and volume of one phase are used as decision variables, a mathematical model for NVT flash calculation is constructed, and the constraints of the mathematical model for NVT flash calculation are defined, including: For a mixture system containing M components, if the mixture system is unstable under given conditions, it will split into two phases, gas and liquid, M ≥ 1; Assume n k and N k is the molar density and molar number of the kth component, k = 1, 2, ..., M, the molar density vector and molar vector are expressed as n = [n1, n2, ..., nM]T and N = [N 1 ,N 2 ,…,N M ] T , the total number of moles of a given substance Denoted as S, The total Helmholtz free energy is denoted by H, and f(n) is the Helmholtz free energy density of the uniform fluid. The total volume V calculated by NVT flash evaporation is total Total molar N total and temperature T are constant; G and L represent the gas phase and liquid phase respectively; For a two-phase system, the volumes of the gas phase and the liquid phase are denoted by V and G and V L , the moles of gas phase and liquid phase are expressed as N G and N L , the molar density of the gas phase and the liquid phase are denoted by n L and n G ; Two constraints define the mathematical model of NVT flash calculation, where: Constraint 1: The following constraints on moles and volumes must be met in the NVT flash calculation: VG+VL=Vtotal NG+NL=Ntotal VG≥0, VL≥0, NG≥0, NL≥0 In gas phase molar N G and volume V G is a decision variable, then the decision variable must take values within the feasible domain, that is, and V G ∈[0,V total ]; Based on the correct definition of Helmholtz free energy, chemical potential and pressure, constraint 2 is defined, and the exterior point method, i.e., penalty function method, is used to process constraint 2, where constraint 2 is: Among them, parameter b k It can be obtained from the universal gas constant R, critical temperature T ck , and the critical pressure P ck The calculation results are: The following relative errors of chemical potential and pressure are required to meet the convergence criteria: The mathematical model for NVT flash calculation is: The mathematical model of NVT flash calculation is iteratively solved based on the constrained grey prediction evolutionary algorithm to obtain a phase diagram when phase equilibrium is reached, and as the number of iterations increases, a total Helmholtz free energy convergence diagram is obtained, including:
51. In the feasible region of mole and volume, randomly generate the first three generations of populations, process each individual in the first three generations of populations with constraint condition 2, and calculate the objective function value H for the individuals that satisfy constraint condition 2; 52. Randomly select three individuals from three consecutive generations of population as a group of selected individuals, wherein an individual is randomly selected from each generation of population, a test individual is generated based on the randomly selected group of individuals, and a test population is generated based on the randomly selected multiple groups of individuals; 53. Constraint condition 2 is processed for each individual in the test population, and the objective function value H is calculated for the test individuals that meet constraint condition 2; 54. By comparing the objective function value H of the target individual and the test individual, the better individual is selected to enter the next generation; 55. Determine whether the current number of iterations t is less than the maximum number of iterations t max If yes, then t=t+1, then execute 52, if no, then calculate the relative error of pressure and chemical potential energy, and judge whether the relative error of pressure and chemical potential energy converges to the standard, if yes, then output the global optimal value, if no, then t max =t max +λ, execute 51.
2. The phase equilibrium calculation method according to claim 1, It is characterized in that The data set collected for NVT flash calculation includes: Six substances, namely, pure n-butane, pure carbon dioxide, 35% methane and 65% pentane, 54.7413% methane and 45.2587% pentane, 55% methane and 45% decane, 50% methane, 30% pentane and 20% decane, were collected as six sets of data sets.
3. The phase equilibrium calculation method according to claim 2, It is characterized in that The parameters of the constraint grey prediction evolution algorithm are set, including: Set the population size and maximum number of iterations t generated by the constrained grey prediction evolutionary algorithm max and the step size λ.
4. The phase equilibrium calculation method according to claim 1, It is characterized in that The above 51 randomly generates the first three generations of populations in the feasible region of moles and volumes, and calculates the objective function value H for each individual in the first three generations of populations, including: Three generations of populations are randomly generated in the feasible region, where the jth dimension of the i-th individual in the t-th generation population is denoted as Generated by the following formula: Where D is the dimension of the problem, up j and low j are the upper and lower bounds of the j-th dimension respectively; rand(0,1) represents a random number generated by a uniform distribution in the range [0,1]; Determine whether each individual in the first three generations of the generated population satisfies constraint 2, calculate the objective function value H for the individuals that satisfy constraint 2, and regenerate the individuals that do not satisfy constraint 2 within the feasible region so that the regenerated individuals satisfy constraint 2, and calculate the objective function value H for the regenerated individuals that satisfy constraint 2.
5. The phase equilibrium calculation method according to claim 1, It is characterized in that The step 52 randomly selects three individuals from three consecutive generations of populations as a group of selected individuals, wherein an individual is randomly selected from each generation of population, a test individual is generated based on the randomly selected group of individuals, and a test species is generated based on the randomly selected multiple groups of individuals, including: Consider three consecutive populations Xt-2, Xt-1, and Xt as a population sequence. t Randomly select three individuals and make Maxdr,j=max{d12,j,d13,j,d23,j},Mindr,j=min{d12,j,d13,j,d23,j},test individual Generated by the following formula: Among them, t≥2, δ is a parameter for controlling the differential threshold of prediction, α, β and θ are the grey development coefficient, grey control parameter and disturbance coefficient respectively, and the formulas for calculating α, β and θ are as follows: θ=rand(-1,1)·δ; Constraint 2 is processed for each individual in the generated experimental population, and the objective function value H of the experimental individual that satisfies constraint 2 is calculated.
6. The phase equilibrium calculation method according to claim 5, It is characterized in that The step 54, by comparing the objective function value H of the target individual and the test individual, selects the better individual to enter the next generation, including: The selection operation is performed based on the greedy selection mechanism by comparing the target individuals and experimental individuals The objective function value is used to select the better individuals to enter the next generation. The selection operation is performed through the following formula: in, is the objective function value of the test individual, is the objective function value of the target individual.
Citation Information
Patent Citations
Multi-modal multi-objective optimization method based on grey prediction evolution algorithm
CN111582428A
Electric power unit combination method based on grey prediction evolution algorithm
CN111582487A
Method for solving multi-demand and multi-dimensional knapsack problem based on constrained differential evolution algorithm
CN117057247A
Method to estimate the influence of pore-size distribution on phase equilibrium of multi-component hydrocarbon systems in unconventional shale gas and oil reservoirs
US20170292373A1
