A method and system for determining the content of rare earth extraction components
Through the improved standard deviation evolution algorithm and least squares method, the problem of large differences in component content calculation and actuality in the rare earth extraction process is solved, the prediction accuracy is improved, and an important reference is provided for the development and optimization of the rare earth extraction process.
Patent Information
- Application Number
- CN202111067887.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-09-13
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-09-13
AI Technical Summary
The existing rare earth extraction process does not consider the energy efficiency of the extraction tank when calculating the series, resulting in a large difference between the calculated component content and the actual production cannot be guided.
The optimized objective function is constructed using the improved standard deviation evolution algorithm, and the component content model in the water phase of the washing section and the organic phase of the extraction section is determined by the least squares method, and the actual energy efficiency separation coefficient is obtained, and a more accurate component content is calculated.
It improves the prediction accuracy of rare earth extraction components content, brings the calculation results closer to actual factory inspection data, and provides an important reference for new process development and existing process flow optimization.
Smart Images

Figure CN113777925B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of extraction process optimization, and in particular to a method and system for determining the content of rare earth extraction components. Background Art
[0002] my country is rich in rare earth resources, and its rare earth extraction technology is at the world's advanced level. When a factory develops a new process, it generally uses a separatory funnel method or a mechanism model to calculate the number of stages required for the process. The design is relatively ideal, and some stages are added on the basis of the calculated number of stages during the actual plant construction process. However, the energy efficiency of the extraction tank is not considered in the actual production process, so the calculated content of each level of components will lag behind by several levels or even dozens of levels, which is quite different from the actual situation. For example, the theoretical calculation of the component content of the 20th level is A, but in the actual production process, the component content of the 26th level extraction tank will be A. Obviously, the parameters of the optimal process designed cannot guide actual production. Summary of the invention
[0003] The purpose of the present invention is to provide a method and system for determining the content of rare earth extraction components, so as to improve the accuracy of predicting the content of rare earth extraction components.
[0004] To achieve the above object, the present invention provides a method for determining the content of rare earth extraction components, the method comprising:
[0005] Step S1: constructing a content model of each component in the water phase of the washing section and a content model of each component in the organic phase of the extraction section;
[0006] Step S2: constructing an optimization objective function based on a model for determining the content of each component in the water phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section using the least squares method;
[0007] Step S3: using an improved standard differential evolution algorithm to solve the optimization objective function, and obtaining the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section;
[0008] Step S4: Substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the model for determining the content of each component in the water phase of the washing section to obtain the content of each component in the organic phase of the washing section; substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the model for determining the content of each component in the organic phase of the extraction section to obtain the content of each component in the water phase of the extraction section.
[0009] Optionally, the improved standard differential evolution algorithm is used to solve the optimization objective function to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section, specifically including:
[0010] Step S31: Initialize parameters;
[0011] Step S32: Determine the initial population obtained by the G-generation evolution population; the initial population includes NP G Individuals of a population;
[0012] Step S33: using a double mutation strategy method to mutate the target vectors corresponding to the individuals of each population in the initial population, to obtain mutation vectors corresponding to the individuals of each population;
[0013] Step S34: performing a binomial crossover operation on the target vector and the mutation vector corresponding to each population individual in the initial population to obtain a test vector corresponding to each population individual in the initial population;
[0014] Step S35: adopting the greedy selection mode, determining the next generation population individuals based on the fitness value of the test vector and the fitness value of the target vector, and storing each population individual into the initial population;
[0015] Step S36: Determine whether G is greater than or equal to the maximum population evolution generation Maxgem; if G is greater than or equal to Maxgem, execute "Step S37"; if G is less than Maxgem, set G=G+1 and return to "Step S33";
[0016] Step S37: Substitute the target vectors corresponding to the individuals in each population in the initial population into the optimization objective function, and use the target vector of the individual in the population corresponding to the minimum fitness value as the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
[0017] Optionally, the specific formula of the model for determining the content of each component in the organic phase of the extraction stage is:
[0018]
[0019] The specific formula for determining the content of each component in the water phase of the washing stage is:
[0020]
[0021] Among them, Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X cirepresents the content of the i-th component at the water phase outlet, Y wi Represents the content of the i-th component at the organic phase outlet.
[0022] Optionally, the optimization objective function includes an objective function in the organic phase of the extraction section and an objective function in the water phase of the washing section;
[0023] The specific formula of the objective function in the organic phase of the extraction section is:
[0024]
[0025] The specific formula of the objective function in the water phase of the washing section is:
[0026]
[0027] Where J represents the sum of squares of the difference between the content of each component in the organic phase of the extraction section and the actual factory data, W represents the sum of squares of the difference between the content of each component in the water phase of the washing section and the actual factory data, minJ and minW are the minimum values of J and W respectively, and Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi represents the content of the i-th component at the outlet of the organic phase, and They respectively represent the content of the i-th component in the organic phase of the extraction section and the aqueous phase of the washing section obtained in the actual factory.
[0028] Optionally, the double mutation strategy method is used to mutate the target vector corresponding to each population individual in the initial population to obtain the mutation vector corresponding to each population individual. The specific calculation formula is:
[0029]
[0030] in, represents the mutation vector corresponding to the nth individual in the population, represents the target vector corresponding to the nth individual in the population, F n represents the proportional factor corresponding to the nth individual in the population, It means randomly selected from the first NP×p individuals in the population with the smallest fitness value in the Gth generation, p∈[0,1], Both represent individuals randomly selected from the Gth generation. represents a population individual randomly selected from the union of the current initial population and A, r1≠r2≠r3≠r4≠r5≠i means [1,NP G ] is a random number in , and γ represents the variation scale control factor.
[0031] The present invention also provides a system for determining the content of rare earth extraction components, the system comprising:
[0032] A model building module, used to build a model of the content of each component in the water phase of the washing section and a model of the content of each component in the organic phase of the extraction section;
[0033] An optimization objective function construction module, used to construct an optimization objective function based on a model for determining the content of each component in the water phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section by using a least square method;
[0034] A solution module, used to solve the optimization objective function by using an improved standard differential evolution algorithm to obtain an actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and an actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section;
[0035] The component content determination module is used to substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the component content determination model in the water phase of the washing section to obtain the content of each component in the organic phase of the washing section; substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the component content determination model in the organic phase of the extraction section to obtain the content of each component in the water phase of the extraction section.
[0036] Optionally, the solution module specifically includes:
[0037] Initialization unit, used to initialize parameters;
[0038] The initial population determination unit is used to determine the initial population obtained by the G-generation evolution population; the initial population includes NP G Individuals of a population;
[0039] A mutation unit, used for mutating the target vectors corresponding to the individuals of each population in the initial population by adopting a double mutation strategy method to obtain mutation vectors corresponding to the individuals of each population;
[0040] A binomial crossover operation unit is used to perform a binomial crossover operation on the target vector and the mutation vector corresponding to the individuals of each population in the initial population to obtain a test vector corresponding to the individuals of each population in the initial population;
[0041] A next generation population individual determination unit is used to adopt a greedy selection mode to determine the next generation population individuals based on the fitness value of the test vector and the fitness value of the target vector, and store each population individual into the initial population;
[0042] The judgment unit is used to judge whether G is greater than or equal to the maximum population evolution generation Maxgem; if G is greater than or equal to Maxgem, the "actual energy efficiency separation coefficient determination unit" is executed; if G is less than Maxgem, G=G+1 is set and the "mutation unit" is returned;
[0043] The actual energy efficiency separation coefficient determination unit is used to substitute the target vectors corresponding to the individuals in each population in the initial population into the optimization objective function, and use the target vector of the population individual corresponding to the minimum fitness value as the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
[0044] Optionally, the specific formula of the model for determining the content of each component in the organic phase of the extraction stage is:
[0045]
[0046] The specific formula for determining the content of each component in the water phase of the washing stage is:
[0047]
[0048] Among them, Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi Represents the content of the i-th component at the organic phase outlet.
[0049] Optionally, the optimization objective function includes an objective function in the organic phase of the extraction section and an objective function in the water phase of the washing section;
[0050] The specific formula of the objective function in the organic phase of the extraction section is:
[0051]
[0052] The specific formula of the objective function in the water phase of the washing section is:
[0053]
[0054] Where J represents the sum of squares of the difference between the content of each component in the organic phase of the extraction section and the actual factory data, W represents the sum of squares of the difference between the content of each component in the water phase of the washing section and the actual factory data, minJ and minW are the minimum values of J and W respectively, and Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi represents the content of the i-th component at the outlet of the organic phase, and They respectively represent the content of the i-th component in the organic phase of the extraction section and the aqueous phase of the washing section obtained in the actual factory.
[0055] Optionally, the double mutation strategy method is used to mutate the target vector corresponding to each population individual in the initial population to obtain the mutation vector corresponding to each population individual. The specific calculation formula is:
[0056]
[0057] in, represents the mutation vector corresponding to the nth individual in the population, represents the target vector corresponding to the nth individual in the population, F n represents the proportional factor corresponding to the nth individual in the population, It means randomly selected from the first NP×p individuals in the population with the smallest fitness value in the Gth generation, p∈[0,1], Both represent individuals randomly selected from the Gth generation. represents a population individual randomly selected from the union of the current initial population and A, r1≠r2≠r3≠r4≠r5≠i means [1,NP G ] is a random number in , and γ represents the variation scale control factor.
[0058] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0059] The invention discloses a method and system for determining the content of rare earth extraction components. The method solves the optimization objective function by using an improved standard differential evolution algorithm, and substitutes the solved actual energy efficiency separation coefficient into a model for determining the content of each component in a washing section water phase and a model for determining the content of each component in an extraction section organic phase that meet actual working conditions, respectively, to obtain the content of each component in the washing section organic phase and the content of each component in the extraction section water phase, so that the calculated content of each component is closer to the content of each level of components obtained by actual factory detection, and provides an important reference for new process development, reorganization of the process flow of an existing rare earth extraction process and re-optimization of process parameters, and has important practical significance. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0061] Figure 1 This is a flow chart of the method for determining the content of rare earth extraction components of the present invention;
[0062] Figure 2 This is a rare earth cascade extraction production flow chart of the present invention;
[0063] Figure 3 This is a flow chart of the improved differential evolution algorithm LCTADE of the present invention;
[0064] Figure 4 The structure diagram of the system for determining the content of rare earth extraction components of the present invention;
[0065] Figure 5 It is a convergence curve diagram of the fitness value error of LCTADE of the present invention and other DE algorithms on functions 5, 8, 12, 26, and 30 in CEC2017 as the evolutionary generations change;
[0066] Figure 6 A schematic diagram of the actual energy efficiency separation coefficient obtained by optimizing the improved differential evolution algorithm of the present invention;
[0067] Figure 7 This is a schematic diagram comparing the content of each component calculated based on the actual energy efficiency separation coefficient of the present invention with the actual value;
[0068] Figure 8 This is a schematic diagram of the error between the calculated value and the actual value of the content of each component calculated by the present invention. DETAILED DESCRIPTION
[0069] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of 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.
[0070] The purpose of the present invention is to provide a method and system for determining the content of rare earth extraction components, so as to improve the accuracy of predicting the content of rare earth extraction components.
[0071] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0072] Example 1
[0073] like Figure 1 As shown, the present invention provides a method for determining the content of rare earth extraction components, the method comprising:
[0074] Step S1: constructing a model for determining the content of each component in the aqueous phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section.
[0075] Step S2: constructing an optimization objective function based on a model for determining the content of each component in the aqueous phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section using the least squares method.
[0076] Step S3: using an improved standard differential evolution algorithm to solve the optimization objective function, and obtaining the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
[0077] Step S4: Substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the model for determining the content of each component in the water phase of the washing section to obtain the content of each component in the organic phase of the washing section; substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the model for determining the content of each component in the organic phase of the extraction section to obtain the content of each component in the water phase of the extraction section.
[0078] The following is a detailed discussion of each step:
[0079] like Figure 2As shown, the cascade extraction separation process is to connect several extraction tanks in series so that the rare earth elements are in continuous contact with the organic phase and the aqueous phase to achieve the separation of different elements. The rare earth cascade extraction production process with a-level washing and b-level extraction, the extractant is added from the first level and flows from left to right under the action of the motor and the stirrer in the extraction tank; the detergent is added from the a+b level and flows from right to left; the feed liquid is added to the organic phase from the b-level and moves from right to left. Through continuous fusion with the extractant, different rare earth elements are extracted into different solvents to achieve separation, and are divided into N components in total, among which the first component is the difficult-to-extract component B in the organic phase with a content of Y B , the Nth component is the easily extractable component A, and the organic phase content is Y A .
[0080] According to the cascade extraction theory, when the rare earth extraction system reaches a steady state, A is the easily extractable component, B is the difficultly extractable component, Y is the organic phase composition, X is the aqueous phase composition, and N is the number of components. The adjacent separation coefficient is defined as:
[0081]
[0082] Among them, β A / B Y represents the adjacent separation coefficient of the easily extractable component A relative to the difficultly extractable component B. B Indicates the content of difficult-to-extract component B in the organic phase, Y A Indicates the content of easily extractable component A in the organic phase, X B represents the content of difficult-to-extract component B in the aqueous phase, X A It indicates the content of easily extractable component A in the water phase.
[0083] Based on the definition of adjacent separation coefficient, the general expression of the relative separation coefficient of the i-th component relative to the difficult-to-extract component B (the first component) can be deduced as follows:
[0084]
[0085] Among them, β 1 / i represents the relative separation coefficient of the difficult-to-extract component B relative to the i-th component, β i / 1 represents the relative separation coefficient of the i-th component relative to the difficult-to-extract component B, Y i represents the content of the i-th component in the organic phase, X i represents the content of the i-th component in the aqueous phase, i = 1, 2, ..., N, N is the number of components, Y B Indicates the content of difficult-to-extract component B in the organic phase, X B Indicates the content of difficult-to-extract component B in the water phase.
[0086] Similarly, based on the definition of adjacent separation coefficient, the general expression of the relative separation coefficient of the i-th component relative to the easily extractable component A (the last component) can be derived as follows:
[0087]
[0088] Among them, β N / i represents the relative separation coefficient of the easily extractable component A relative to the i-th component, β i / N represents the relative separation coefficient of the i-th component relative to the easily extractable component A, Y i represents the content of the i-th component in the organic phase, X i represents the content of the i-th component in the aqueous phase, i = 1, 2, ..., N, N is the number of components, Y A represents the organic phase content in the easily extractable component A, X A It indicates the water content in the easily extractable component A.
[0089] When the rare earth extraction system is in equilibrium, the calculation formula for the relative separation coefficient is shown in Table 1, in which L is the element number of the cutting position for dividing the difficult-to-extract components and the easily-to-extract components, the component number is N, and k is the current element number. For example, when it is a three-component extraction and separation process of Ce, Pr, and Nd, Pr is the element for dividing the cutting position of the easily-to-extract component Ce and the difficult-to-extract component Nd, the component number N is 3, and L is 2.
[0090] Table 1 Actual separation coefficients in various stages of rare earth extraction and separation
[0091]
[0092]
[0093] When simulating and calculating step by step from the first stage to the intermediate feed stage, it is known that the content of the i-th component at the water phase outlet is X ci , we need to solve the i-th component Y in the organic phase of the extraction section ci . From formula (2), we get:
[0094]
[0095] The sum of the components in the organic phase of formula (4) is equal to 1, that is, (Y1+Y2+…+Y N =1), so we can conclude that:
[0096]
[0097] Substituting equation (5) back into equation (4) yields the general formula for calculating the content of each component in the organic phase of the extraction section:
[0098]
[0099] Similarly, it is known that the content of the i-th component at the organic phase outlet is Y wi , the general formula for calculating the content of each component in the water phase of the washing section is:
[0100]
[0101] Based on the above two equations, the theoretical aging separation coefficient β ck or β wk Calculate the content of each element component between different phases in each level.
[0102] Considering that the actual extraction industrial site has a certain amount of energy consumption due to the agitator in the extraction tank, resulting in different extraction efficiencies, the compositions X and Y of the aqueous phase and the organic phase are not accurate when the rare earth extraction system reaches stability, resulting in incorrect relative separation coefficient values from formulas (2) and (3), making the output values of the rare earth element component content calculation model based on the relative separation coefficient inconsistent with the actual working conditions. Therefore, it is proposed that the energy efficiency separation coefficient reduces the influence of the energy efficiency of the extraction tank on the model calculation of the component content, and the specific formula for establishing the model for determining the content of each component in the organic phase of the extraction section based on the actual energy efficiency separation coefficient in step S1 is:
[0103]
[0104] The specific formula for establishing the model for determining the content of each component in the water phase of the washing section based on the actual energy efficiency separation coefficient is:
[0105]
[0106] Among them, Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi Represents the content of the i-th component at the organic phase outlet.
[0107] Step S2: constructing an optimization objective function based on the content determination model of each component in the water phase of the washing section and the content determination model of each component in the organic phase of the extraction section by using the least squares method; the optimization objective function includes the objective function in the organic phase of the extraction section and the objective function in the water phase of the washing section.
[0108] The specific formula of the objective function in the organic phase of the extraction section is:
[0109]
[0110] The specific formula of the objective function in the water phase of the washing section is:
[0111]
[0112] Where J represents the sum of squares of the difference between the content of each component in the organic phase of the extraction section and the actual factory data, W represents the sum of squares of the difference between the content of each component in the water phase of the washing section and the actual factory data, minJ and minW are the minimum values of J and W respectively, and Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β′ ck represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi represents the content of the i-th component at the outlet of the organic phase, and They respectively represent the content of the i-th component in the organic phase of the extraction section and the aqueous phase of the washing section obtained in the actual factory.
[0113] This embodiment uses the least squares method to minimize the sum of the squares of the differences between the contents of each component at each level calculated by the two content determination models and the actual factory data. When using the optimization algorithm to solve, they need to be solved separately, and the two optimization objective functions are replaced by the fitness function f() instead of J and W when solving.
[0114] This embodiment uses the improved DE algorithm to solve the two optimization objectives respectively. First, the objective function J is minimized to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section. Similarly, the objective function W is minimized to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section.
[0115] like Figure 3 As shown, the improved standard differential evolution algorithm is used to solve the optimization objective function to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section, specifically including:
[0116] Step S31: Initialize parameters; the parameters include: population evolution generation G, space dimension D, maximum population number NP max , the minimum population size NP min , maximum population evolution number Maxgem, mean μCR, external archive A. Where, D = N-1.
[0117] Step S32: Determine the initial population obtained by the G-generation evolution population; the initial population includes NPG Individuals of a population, n=1,2,…,NP G ,m=1,2,...,D; where, represents the vector set corresponding to the initial population, Indicates the NP in the G-th generation initial population G The target vector corresponding to each individual in the population, each target vector consists of D-dimensional parameters, β' m,n,G Represents the parameter corresponding to the mth dimension of the nth population individual in the Gth generation initial population.
[0118] The method of mapping the chaotic space to the solution space of the objective function:
[0119] 1) Randomly generate a range of -1≤h(1) m ≤1; 1≤m≤d, the D-dimensional chaotic variable H=(h(1)1,h(1)2,...,h(1) d ), where h(1) m Represents the m-th dimension parameter of the generated chaotic vector.
[0120] 2) Each dimension of the generated H is subjected to NP G -1 iteration to get NP G -1 Chaos variable.
[0121] Using the cubic mapping in the chaotic model to initialize the population, the resulting chaotic sequence distribution is more evenly distributed than the commonly used Logistic mapping. Therefore, the iterative expression for chaotic sequence generation is:
[0122] h(n+1) m =4h(n) m 3 -3h(n) m ; -1≤h(n) m ≤1; n=0,1,2,...,NP G ; m = 1, ..., D (12);
[0123] Among them, h(n) m ,h(n+1) m are the values obtained from the n-1th and nth iterations of the mth dimension of the chaotic vector.
[0124] 3) According to Will generate NP G The D-dimensional chaotic variables are mapped to the solution space of the optimization objective function, and the population size is NP G , where β' min ,β' maxare the minimum and maximum values of each dimension of each population individual (i.e., target vector), Represents the m-th dimension parameter of the n-th population individual in the mapped solution space, h(n) m Represents the m-th dimension parameter of the n-th population individual of the iteratively generated chaotic vector.
[0125] The initial population generated by chaos initialization is composed of NP G The target vector corresponding to each individual in the population is composed of parameters with dimension D. The population evolution generation G is initialized to 0, so the vector set corresponding to the initial population is obtained as follows:
[0126]
[0127] The performance requirements of mutation strategies are different at different stages of evolution. "DE / current-to-pbest / 1" and "DE / rand / 1" are selected as the multi-scale mutation strategy set. In order to balance the convergence speed and accuracy of the algorithm, a double mutation strategy method based on DE / current-to-pbest / 1 and DE / rand / 1 is proposed. At the Gth generation, the target vectors corresponding to the individuals of each population in the initial population are Perform mutation to obtain mutation vector The specific steps are as follows:
[0128] Step S33: using a double mutation strategy method to mutate the target vector corresponding to each population individual in the initial population, to obtain the mutation vector corresponding to each population individual, the specific calculation formula is:
[0129]
[0130] in, represents the mutation vector corresponding to the nth individual in the population, represents the target vector corresponding to the nth individual in the population, F n represents the proportional factor corresponding to the nth individual in the population, It means randomly selecting from the first NP×p(p∈[0,1]) individuals in the population with the smallest (better) fitness value in the G-th generation. Both represent individuals randomly selected from the Gth generation. represents a population individual randomly selected from the union of the current initial population and A, r1≠r2≠r3≠r4≠r5≠i means [1,NP G ] is a random number in , and γ represents the variation scale control factor. and They are obtained by mutation strategies DE / current-to-pbest / 1 and DE / rand / 1, respectively. When the fitness value of the individual after evolution is better than that before evolution, the individual before evolution is stored in the external archive A. When obtaining the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section, the target vector corresponding to the individual of each population before evolution is substituted into formula (10) to obtain J, and J is used to represent the fitness f().
[0131] γ∈[0,1] is the mutation scale control factor, which decreases with the evolutionary generation in the optimization process. The early stage of the search process of most evolutionary algorithms is more important than the later stage. It is necessary to increase the proportion of mutation vectors generated by DE / current-to-pbest / 1 to make the search range close to the optimal solution and speed up the convergence. γ is larger. In the later stage, due to the poor diversity of the population, the algorithm is prone to fall into the local optimal solution. Therefore, it is necessary to increase the proportion of mutation vectors generated by DE / rand / 1. The γ value should be smaller in the later stage. According to the evolutionary characteristics, γ is defined as:
[0132]
[0133] Among them, γ represents the mutation scale control factor, γ∈[0,1], G represents the current population evolution generation, and Maxgem represents the maximum iteration generation, which is generally set to 3000.
[0134] F n According to the fitness value of each generation of individuals in the evolutionary process, it is updated, so it is concluded that F n The specific calculation formula is:
[0135]
[0136] Among them, F n represents the proportional factor corresponding to the nth individual in the initial population, They represent the target vectors corresponding to the individuals with the largest and smallest fitness values in the initial population of the Gth generation, randc i (0.5,0.1) is a random number generated according to the Cauchy distribution. represents the maximum fitness value in the initial population of the Gth generation, represents the minimum fitness value in the initial population of the Gth generation, Represents the target vector corresponding to the nth individual in the population.
[0137] Step S34: The target vectors corresponding to the individuals in each population in the initial population are and mutation vector Perform a binomial crossover operation to obtain the test vectors corresponding to the individuals in each population in the initial population. Among them, um,n,G Represents the test vector corresponding to the nth individual in the Gth generation initial population The m-th dimension parameter in .
[0138] The specific formula for calculating the binomial crossover operation is:
[0139]
[0140] Among them, CR n represents the crossover probability of the nth individual in the initial population, CR n ∈(0,1), if rand n,m [0,1] The generated random number is less than or equal to the crossover probability CR n Or the dimension m of the current vector is equal to a randomly generated dimension m rand , then the test vector corresponding to the nth individual in the Gth generation initial population is The m-th dimension parameter u in m,n,G The mutation vector corresponding to the nth individual in the initial population of the Gth generation The m-th dimension parameter v m,n,G Instead; on the contrary, the test vector corresponding to the nth individual in the initial population of the Gth generation The m-th dimension parameter u in m,n,G The target vector corresponding to the nth individual in the Gth generation initial population The m-th dimension parameter β′ m,n,G replace.
[0141] Calculate the crossover probability CR corresponding to the nth population individual in the Gth generation initial population n,G The specific formula is:
[0142] CR n,G =randi(μCR G ,0.1) (18);
[0143] Among them, CR n,G Indicates the crossover probability CR corresponding to the nth population individual in the Gth generation initial population n,G , randi(μCR G ,0.1) indicates the mean value is μCR G The standard deviation is 0.1. The initial μCR G Set to 0.5. After each iteration, update the mean μCR G The specific formula is:
[0144] μCR G+1 =(1-c CR )×μCR G +c CR ×meanWL (S CR,G ) (19);
[0145] Among them, S CR,G represents the set of crossover probabilities corresponding to all successful population individuals, μCR G The range is [0,1], maen WL (S CR,G ) represents S CR,G The crossover probability CR corresponding to all successful individuals in the population i,G The weighted average of CR is a constant value, set to 0.5.
[0146] Calculate maen WL (S CR,G ) is as follows:
[0147]
[0148] Among them, ω k is the weight coefficient, Δfk is the fitness value of the nth individual in the initial population after evolution Fitness value before evolution The absolute value of the difference, S k,CR,G Indicates S CR,G The value of the k-th crossover probability in the set, represents the test vector corresponding to the nth individual in the initial population of the Gth generation, the evolved vector, represents the target vector corresponding to the nth individual in the initial population of the Gth generation, the vector before evolution, S CR,G Represents the set consisting of the crossover probabilities corresponding to all successful population individuals.
[0149] Step S35: adopt the greedy selection mode, based on the test vector The fitness value and target vector The fitness value of the next generation population is determined, and each population individual is stored in the initial population. At this time, the initial population includes NP G+1 There are population individuals, each population individual corresponds to a target vector, and each target vector includes D-dimensional parameters.
[0150] The number of individuals in the population decreases linearly. In the early stage of evolution, the population is large, which increases the diversity of the population. In the later stage of evolution, the number of individuals in the population is small, which speeds up the convergence speed. Therefore, the number of individuals in the G+1 generation population NP is calculated at this time. G+1 The specific formula is:
[0151]
[0152] Among them, NPmax is the maximum population size, generally set to 18*D; NP min is the minimum population size, which is set to 4; NFE is the number of fitness value evaluations, which increases with the number of iterations; MAX_NFE is the maximum number of fitness value evaluations; round() represents the rounding function.
[0153] Use the greedy selection mode to make the offspring test vector Always better than or equal to the parent's target vector To minimize the optimization objective function value, that is, the fitness value, the greedy selection mode is adopted based on the test vector The fitness value and target vector The specific formula for determining the fitness value of individuals in the next generation population is:
[0154]
[0155] in, Represents the test vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the initial population of the G+1th generation.
[0156] The above formula shows that when the evolutionary test vector The fitness value of Better than the target vector The fitness value of When the test vector Replace the target vector Evolve into an individual in the next generation population, otherwise retain the original target vector
[0157] Step S36: Determine whether G is greater than or equal to Maxgem; if G is greater than or equal to Maxgem, execute "Step S37"; if G is less than Maxgem, set G=G+1 and return to "Step S33".
[0158] Step S37: Substitute the target vectors corresponding to the individuals of each population in the initial population into the optimization objective function, and use the target vector of the individual of the population corresponding to the minimum fitness value as the actual energy efficiency separation coefficient. Specifically, the target vectors corresponding to the individuals of each population in the initial population are Substitute the multiple parameters in into β′ in formula 10 ck , the minimum fitness value minJ corresponds to the target vector of the population individual As the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section. Similarly, the target vector corresponding to each population individual in the initial population Substitute the multiple parameters in into β′ in formula 11 wk , the minimum fitness value minW corresponds to the target vector of the population individual As the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section.
[0159] Example 2
[0160] like Figure 4 As shown, the present invention also discloses a system for determining the content of rare earth extraction components, the system comprising:
[0161] The model building module 401 is used to build a model of the content of each component in the water phase of the washing section and a model of the content of each component in the organic phase of the extraction section.
[0162] The optimization objective function construction module 402 is used to construct the optimization objective function based on the content determination model of each component in the water phase of the washing section and the content determination model of each component in the organic phase of the extraction section by using the least square method.
[0163] The solution module 403 is used to solve the optimization objective function using an improved standard differential evolution algorithm to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
[0164] The component content determination module 404 is used to substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the component content determination model in the water phase of the washing section to obtain the content of each component in the organic phase of the washing section; substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the component content determination model in the organic phase of the extraction section to obtain the content of each component in the water phase of the extraction section.
[0165] As an optional implementation, the solution module 403 of the present invention specifically includes:
[0166] Initialization unit, used to initialize parameters.
[0167] The initial population determination unit is used to determine the initial population obtained by the G-generation evolution population; the initial population includes NP G Individuals of a population.
[0168] The mutation unit is used to mutate the target vectors corresponding to the individuals of each population in the initial population by adopting a double mutation strategy method to obtain mutation vectors corresponding to the individuals of each population.
[0169] The binomial crossover operation unit is used to perform a binomial crossover operation on the target vector and the mutation vector corresponding to the individuals of each population in the initial population to obtain the test vector corresponding to the individuals of each population in the initial population.
[0170] The next generation population individual determination unit is used to adopt a greedy selection mode, determine the next generation population individuals based on the fitness value of the test vector and the fitness value of the target vector, and store each population individual into the initial population.
[0171] The judgment unit is used to judge whether G is greater than or equal to the maximum population evolution generation Maxgem; if G is greater than or equal to Maxgem, the "actual energy efficiency separation coefficient determination unit" is executed; if G is less than Maxgem, G=G+1 is set and the "mutation unit" is returned;
[0172] The actual energy efficiency separation coefficient determination unit is used to substitute the target vectors corresponding to the individuals in each population in the initial population into the optimization objective function, and use the target vector of the population individual corresponding to the minimum fitness value as the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
[0173] The same contents as those in Example 1 are not repeated here one by one.
[0174] Example 3
[0175] The initialization parameter values of the present invention are: D=2, NP max =18*D,NP min =4,μCR G =0.5, MAX_NFE=D*1000, Maxgem=3000; the same steps as those in Example 1 are not described again.
[0176] Figure 3 FIG. 1 is a specific flow chart of the improved differential evolution algorithm LCTADE of the present invention. The convergence curves of the fitness value error when LCTADE and other improved DE algorithms optimize different functions with the change of evolutionary generations are shown in FIG. Figure 5 As shown, (a) is the convergence curve of the fitness value error of function 5 as the evolutionary algebra changes, (b) is the convergence curve of the fitness value error of function 8 as the evolutionary algebra changes, (c) is the convergence curve of the fitness value error of function 12 as the evolutionary algebra changes, (d) is the convergence curve of the fitness value error of function 26 as the evolutionary algebra changes, and (e) is the convergence curve of the fitness value error of function 30 as the evolutionary algebra changes; the present invention uses the improved algorithm LCTADE to optimize the objective functions (10) and (11) to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the extraction section and the washing section. The results are shown in FIG. Figure 6As shown, (a) is the actual energy efficiency separation coefficient of the extraction section, and (b) is the actual energy efficiency separation coefficient of the washing section; the actual energy efficiency separation coefficient corresponding to each component content obtained by optimization is substituted into the rare earth element component content calculation model (8) and (9) established based on the energy efficiency separation coefficient to obtain the actual component content values at each level, which are compared with the actual component content as shown in the figure Figure 7 As shown in the error diagram Figure 8 shown.
[0177] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0178] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A method for determining the content of rare earth extraction components, It is characterized in that The method comprises: Step S1: constructing a content model of each component in the water phase of the washing section and a content model of each component in the organic phase of the extraction section; Step S2: constructing an optimization objective function based on a model for determining the content of each component in the water phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section using the least squares method; Step S3: using an improved standard differential evolution algorithm to solve the optimization objective function, and obtaining the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section; Step S4: Substituting the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the model for determining the content of each component in the water phase of the washing section, to obtain the content of each component in the organic phase of the washing section; substituting the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the model for determining the content of each component in the organic phase of the extraction section, to obtain the content of each component in the water phase of the extraction section; The improved standard differential evolution algorithm is used to solve the optimization objective function to obtain the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section, specifically including: Step S31: Initialize parameters; Step S32: Determine the initial population obtained by the G-generation evolution population; the initial population includes NP G Individuals of a population; Step S33: using a double mutation strategy method to mutate the target vector corresponding to each population individual in the initial population to obtain a mutation vector corresponding to each population individual; using a double mutation strategy method to mutate the target vector corresponding to each population individual in the initial population to obtain a mutation vector corresponding to each population individual, the specific calculation formula is: in, represents the mutation vector corresponding to the nth individual in the population, represents the target vector corresponding to the nth individual in the population, F n represents the proportional factor corresponding to the nth individual in the population, It means randomly selected from the first NP×p individuals in the population with the smallest fitness value in the Gth generation, p∈[0,1], Both represent individuals randomly selected from the Gth generation. represents a population individual randomly selected from the union of the current initial population and A, r 1 ≠r 2 ≠r 3 ≠r 4 ≠r 5 ≠i means [1,NP G ], γ represents the variation scale control factor, and the expression of γ is: Among them, γ represents the mutation scale control factor, γ∈[0,1], G represents the current population evolution generation, and Maxgem represents the maximum iteration generation; Step S34: performing a binomial crossover operation on the target vector and the mutation vector corresponding to each population individual in the initial population to obtain a test vector corresponding to each population individual in the initial population; Step S35: Adopting the greedy selection mode, determine the next generation population individuals based on the fitness value of the test vector and the fitness value of the target vector, and store each population individual into the initial population; adopting the greedy selection mode, based on the test vector The fitness value and target vector The specific formula for determining the fitness value of individuals in the next generation population is: in, Represents the test vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the initial population of the G+1th generation; Step S36: Determine whether G is greater than or equal to the maximum population evolution generation Maxgem; if G is greater than or equal to Maxgem, execute "Step S37"; if G is less than Maxgem, set G=G+1 and return to "Step S33"; Step S37: Substitute the target vectors corresponding to the individuals in each population in the initial population into the optimization objective function, and use the target vector of the individual in the population corresponding to the minimum fitness value as the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
2. The method for determining the content of rare earth extraction components according to claim 1, It is characterized in that The specific formula for determining the content of each component in the organic phase of the extraction stage is: The specific formula for determining the content of each component in the water phase of the washing stage is: Among them, Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β c ' k represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi Represents the content of the i-th component at the organic phase outlet.
3. The method for determining the content of rare earth extraction components according to claim 1, It is characterized in that The optimization objective function includes an objective function in the organic phase of the extraction section and an objective function in the water phase of the washing section; The specific formula of the objective function in the organic phase of the extraction section is: The specific formula of the objective function in the water phase of the washing section is: Where J represents the sum of squares of the difference between the content of each component in the organic phase of the extraction section and the actual factory data, W represents the sum of squares of the difference between the content of each component in the water phase of the washing section and the actual factory data, minJ and minW are the minimum values of J and W respectively, and Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β c ' k represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi represents the content of the i-th component at the outlet of the organic phase, and They respectively represent the content of the i-th component in the organic phase of the extraction section and the aqueous phase of the washing section obtained in the actual factory.
4. A system for determining the content of rare earth extraction components, It is characterized in that The system comprises: A model building module, used to build a model of the content of each component in the water phase of the washing section and a model of the content of each component in the organic phase of the extraction section; An optimization objective function construction module, used to construct an optimization objective function based on a model for determining the content of each component in the water phase of the washing section and a model for determining the content of each component in the organic phase of the extraction section by using a least square method; A solution module, used to solve the optimization objective function by using an improved standard differential evolution algorithm to obtain an actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and an actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section; The component content determination module is used to substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section into the component content determination model in the water phase of the washing section to obtain the content of each component in the organic phase of the washing section; substitute the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section into the component content determination model in the organic phase of the extraction section to obtain the content of each component in the water phase of the extraction section; The solution module specifically includes: Initialization unit, used to initialize parameters; The initial population determination unit is used to determine the initial population obtained by the G-generation evolution population; the initial population includes NP G Individuals of a population; The mutation unit is used to mutate the target vector corresponding to each population individual in the initial population by adopting a double mutation strategy method to obtain the mutation vector corresponding to each population individual; the double mutation strategy method is used to mutate the target vector corresponding to each population individual in the initial population to obtain the mutation vector corresponding to each population individual, and the specific calculation formula is: in, represents the mutation vector corresponding to the nth individual in the population, represents the target vector corresponding to the nth individual in the population, F n represents the proportional factor corresponding to the nth individual in the population, It means randomly selected from the first NP×p individuals in the population with the smallest fitness value in the Gth generation, p∈[0,1], Both represent individuals randomly selected from the Gth generation. represents a population individual randomly selected from the union of the current initial population and A, r 1 ≠r 2 ≠r 3 ≠r 4 ≠r 5 ≠i means [1,NP G ], γ represents the variation scale control factor, and the expression of γ is: Among them, γ represents the mutation scale control factor, γ∈[0,1], G represents the current population evolution generation, and Maxgem represents the maximum iteration generation; A binomial crossover operation unit is used to perform a binomial crossover operation on the target vector and the mutation vector corresponding to the individuals of each population in the initial population to obtain a test vector corresponding to the individuals of each population in the initial population; The next generation population individual determination unit is used to adopt a greedy selection mode to determine the next generation population individuals based on the fitness value of the test vector and the fitness value of the target vector, and store each population individual into the initial population; adopt a greedy selection mode based on the test vector The fitness value and target vector The specific formula for determining the fitness value of individuals in the next generation population is: in, Represents the test vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the Gth generation initial population The fitness value of Represents the target vector corresponding to the nth individual in the initial population of the G+1th generation; The judgment unit is used to judge whether G is greater than or equal to the maximum population evolution generation Maxgem; if G is greater than or equal to Maxgem, the "actual energy efficiency separation coefficient determination unit" is executed; if G is less than Maxgem, G=G+1 is set and the "mutation unit" is returned; The actual energy efficiency separation coefficient determination unit is used to substitute the target vectors corresponding to the individuals in each population in the initial population into the optimization objective function, and use the target vector of the population individual corresponding to the minimum fitness value as the actual energy efficiency separation coefficient corresponding to the content of each component in the water phase of the washing section and the actual energy efficiency separation coefficient corresponding to the content of each component in the organic phase of the extraction section.
5. The system for determining the content of rare earth extraction components according to claim 4, It is characterized in that The specific formula for determining the content of each component in the organic phase of the extraction stage is: The specific formula for determining the content of each component in the water phase of the washing stage is: Among them, Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β c ' k represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi Represents the content of the i-th component at the organic phase outlet.
6. The system for determining the content of rare earth extraction components according to claim 4, It is characterized in that The optimization objective function includes an objective function in the organic phase of the extraction section and an objective function in the water phase of the washing section; The specific formula of the objective function in the organic phase of the extraction section is: The specific formula of the objective function in the water phase of the washing section is: Where J represents the sum of squares of the difference between the content of each component in the organic phase of the extraction section and the actual factory data, W represents the sum of squares of the difference between the content of each component in the water phase of the washing section and the actual factory data, minJ and minW are the minimum values of J and W respectively, and Y ci represents the content of the i-th component in the organic phase of the extraction section, X wi Indicates the content of the i-th component in the water phase of the washing section, i = 1, 2, ..., N, N is the number of components, β' wk represents the actual energy efficiency separation coefficient corresponding to the kth component in the water phase of the washing section, β c ' k represents the actual energy efficiency separation coefficient corresponding to the kth component in the organic phase of the extraction section, k and i are both independent variables, k = 1, 2, ..., i, X ci represents the content of the i-th component at the water phase outlet, Y wi represents the content of the i-th component at the outlet of the organic phase, and They respectively represent the content of the i-th component in the organic phase of the extraction section and the aqueous phase of the washing section obtained in the actual factory.
