Coal-based n-alkane phase change material optimization extraction method based on data-driven model

Abstract

The invention discloses a coal-based n-alkane phase change material optimization extraction method based on a data-driven model, and relates to the technical field of data processing. The method comprises the following steps: step 1, collecting experimental data of coal-based n-alkanes; constructing a coal sample feature vector based on the experimental data; 2, constructing a dissolution selectivity prediction model based on the coal sample feature vector and a dissolution parameter matrix of the coal sample; 3, constructing a multi-field coupling transmission equation; establishing a separation efficiency prediction model and a phase change performance prediction function according to a multi-field coupling transmission equation; establishing a crystallization process control equation, and solving the crystallization process control equation to obtain the crystallinity; step 4, constructing a multi-objective optimization function; solving a multi-objective optimization function to obtain an optimal process parameter combination; according to the method, the energy consumption and the production cost are remarkably reduced while the material purity and the extraction efficiency are ensured.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a method for optimizing the extraction of coal-based normal alkane phase change materials based on a data-driven model. Background Art

[0002] In the research on the extraction and application of coal-based normal alkane phase change materials, the existing technology mainly focuses on the deep processing and separation methods of coal-based raw materials. Coal-based normal alkanes are an important type of phase change materials with unique thermophysical properties, such as high latent heat and controllable phase change temperature, which make them have broad application potential in energy storage, temperature regulation and thermal management. However, due to the complex composition, high carbon-hydrogen ratio and many impurities of the coal-based raw materials themselves, the extraction and purification process of coal-based normal alkanes faces great technical challenges. Most of the existing technologies rely on traditional chemical separation methods and thermal treatment technologies, but these methods have many shortcomings in terms of energy consumption, separation efficiency, material purity and so on. Therefore, how to efficiently extract high-purity normal alkanes from complex coal-based raw materials while reducing energy consumption and improving extraction quality has become a technical problem that needs to be solved urgently in the industry.

[0003] In the prior art, the most common methods for extracting coal-based normal alkanes include solvent extraction, distillation, fractionation and hydrodewaxing. The solvent extraction method uses the selective solubility of the solvent for different components to extract the normal alkanes in the coal-based raw materials from other components. This method has a certain degree of selectivity, but its solvent recovery cost is high, and impurities may be introduced during the dissolution process, affecting the purity of the extracted material. In addition, since the solvent is not selective for normal alkanes with different carbon chain lengths, it is difficult for this method to effectively extract normal alkanes with specific carbon chain lengths from complex coal-based raw materials. Distillation and fractionation technology separates components with different boiling points, but this method requires high temperature conditions, which can easily lead to thermal decomposition of the raw materials, reduce the thermal stability of the materials, and the energy consumption problem associated with high temperature operation has not been solved.

[0004] Hydrodewaxing is another common method for extracting normal alkanes, which uses hydrogenation reactions to remove non-normal components and improve the purity of normal alkanes. However, this method usually requires high pressure and catalysts, is relatively complex in operating conditions, has high equipment investment costs, and the selection and activity maintenance of catalysts is also a key issue. High pressure and hydrogenation atmosphere can also cause changes in the molecular structure of coal-based raw materials, causing instability during the conversion process. The by-products produced in this process may increase the difficulty of subsequent purification and further reduce the utilization efficiency of the materials. Therefore, although hydrodewaxing has advantages in extraction purity, it shows obvious deficiencies in energy consumption and equipment requirements. Summary of the invention

[0005] The purpose of the present invention is to provide a coal-based normal alkane phase change material optimization extraction method based on a data-driven model. Through multi-field coupling and multi-objective optimization, the present invention can efficiently extract normal alkanes of specific carbon chain lengths from complex coal-based raw materials, while achieving the best balance between separation efficiency and energy consumption. During the extraction process, the data-driven model is combined with the dynamic adjustment of variables such as temperature, pressure, and reaction time, so that the extraction process can be optimized under different conditions, thereby significantly reducing energy consumption and production costs while ensuring material purity and extraction efficiency.

[0006] In order to solve the above technical problems, the present invention provides a method for optimizing the extraction of coal-based normal alkane phase change materials based on a data-driven model, the method comprising:

[0007] Step 1: Collect experimental data of coal-based normal alkanes, including: coal sample component data, reaction temperature, reaction pressure and reaction time; construct coal sample feature vector based on experimental data;

[0008] Step 2: constructing a dissolution selectivity prediction model based on the coal sample characteristic vector and the dissolution parameter matrix of the coal sample; the output result of the dissolution selectivity prediction model is the solubility of coal-based normal alkanes in the coal sample at a specific reaction temperature;

[0009] Step 3: Based on the specific heat capacity and density of the coal sample, combined with the output results of the solubility selectivity prediction model, a multi-field coupling transmission equation is constructed, and the multi-field coupling transmission equation is solved to obtain the concentration of coal-based normal alkanes; then, according to the multi-field coupling transmission equation, a separation efficiency prediction model is established and a phase change performance prediction function is constructed; the output result of the separation efficiency prediction model is the separation efficiency; the phase change performance prediction function is solved to obtain the phase change temperature of coal-based normal alkanes; according to the separation efficiency prediction model and the phase change performance prediction function, a crystallization process control equation is established, and the crystallization process control equation is solved to obtain the crystallinity;

[0010] Step 4: Based on the crystallinity, separation efficiency, phase change temperature of coal-based normal alkanes and the concentration of coal-based normal alkanes, a multi-objective optimization function is constructed; by solving the multi-objective optimization function, an optimal process parameter combination is obtained; the optimal process parameter combination includes: an optimal reaction temperature, an optimal reaction pressure and an optimal reaction time; and the optimal process parameter combination is used to optimize the extraction of coal-based normal alkane phase change materials.

[0011] Furthermore, the coal sample component data include: reaction degree, volume change, number of components, carbon content, hydrogen content and oxygen content in the components.

[0012] Furthermore, the coal sample feature vector is expressed using the following formula:

[0013]

[0014] Where X(t) is the coal sample characteristic vector at reaction time t; N is the number of components in the coal sample; i is an integer subscript index ranging from 1 to N; C i is the carbon content of the i-th component; e is the natural base; T is the reaction temperature; E i is the activation energy of the i-th component; α j is the reaction degree of the jth component, j is an integer subscript index ranging from 1 to N, j≠i; H i is the hydrogen content of the i-th component; P is the reaction pressure; P0 is the standard pressure; O i is the oxygen content of the i-th component; ΔV i is the volume change of the i-th component; ΔG mix is the mixing Gibbs free energy; τ i is the characteristic relaxation time of the i-th component, and the calculation formula is:

[0015]

[0016] Furthermore, the dissolution selectivity prediction model is expressed using the following formula:

[0017]

[0018] Where S(n,T) represents the solubility of coal-based normal alkanes with a carbon chain length of n in the coal sample at the reaction temperature T; K d is the preset dissolution constant; X Z (t) is the transpose of X(t); μ is the average carbon chain length of coal-based normal alkanes; σ is the standard deviation of the carbon chain length of coal-based normal alkanes; D(T) is the solubility parameter matrix, expressed as:

[0019]

[0020] Where ΔH1 is the heat of solution of the first component, ΔH2 is the heat of solution of the second component, and ΔH N is the heat of solution of the Nth component; E1 is the activation energy of the first component, E2 is the activation energy of the second component, and E N is the activation energy of the Nth component; κ 12 (T) represents the dissolution influence coefficient of the first component on the second component at temperature T. Similarly, κ ij (T) represents the dissolution influence coefficient of the i-th component on the j-th component at temperature T; Z is the transposition operation.

[0021] Furthermore, the multi-field coupling transmission equation is expressed using the following formula:

[0022]

[0023] Among them, C n is the concentration of coal-based normal alkanes with carbon chain length n; is the gradient operator; k is the thermal conductivity of the coal sample; ρ is the density of the coal sample; C p is the specific heat capacity of the coal sample; is the reaction temperature change value; is the reaction pressure change value; D eff is the effective diffusion coefficient, which is calculated by the following formula:

[0024] D eff =φ·τ·D m S(n,T);

[0025] Where, φ is the porosity of the coal sample; D m is the molecular diffusion coefficient of the medium, that is, the free diffusion coefficient of coal-based normal alkanes in the coal sample; τ is the tortuosity, which is a set value greater than 1.

[0026] Furthermore, the separation efficiency prediction model is expressed using the following formula:

[0027]

[0028] Among them, θ(t) represents the separation efficiency at reaction time t; X(0) represents the coal sample characteristic vector at the initial reaction time.

[0029] Furthermore, the phase change performance prediction function is expressed using the following formula:

[0030]

[0031] Among them, T m (n) is the phase transition temperature of coal-based normal alkanes with a carbon chain length of n; ΔH f is the latent heat; T0 is the preset parameter reference temperature.

[0032] Furthermore, the crystallization process control equation is expressed using the following formula:

[0033]

[0034] Among them, X c is the crystallinity; k0 is the crystallization rate constant; is the saturated concentration of coal-based normal alkanes with a carbon chain length of n, calculated based on the Wilson equation:

[0035]

[0036] Where ΔH sol is the dissolution enthalpy change; C0 is the saturation concentration at the reference temperature T0.

[0037] Furthermore, the multi-objective optimization function is calculated using the following formula:

[0038]

[0039] Among them, T target is the target phase transition temperature; E total is the total energy consumption; P opt is the optimal process parameter combination.

[0040] The coal-based normal alkane phase change material optimization extraction method based on the data-driven model of the present invention has the following beneficial effects:

[0041] The present invention introduces a data-driven multi-objective optimization model in the extraction process. Most traditional extraction methods rely on a single process parameter, while the present invention achieves global optimization of reaction conditions by incorporating factors such as crystallinity, separation efficiency, solubility selectivity and total energy consumption into the multi-objective optimization model. This method ensures that the temperature, pressure and time in the extraction process can operate in the optimal combination, thereby achieving an optimal balance between extraction efficiency and energy consumption, avoiding cost issues caused by excessive energy consumption in traditional methods. At the same time, the multi-objective optimization model makes the extraction process flexible and adaptable, and can cope with changes in different raw material characteristics or production needs. Through the dynamic balance of energy consumption and separation efficiency, the present invention significantly improves the industrial extraction efficiency of coal-based normal alkanes.

[0042] Secondly, the multi-field coupled transmission equation and dissolution selectivity prediction model of the present invention significantly improve the accuracy and stability of the extraction process. In the extraction process of coal-based normal alkanes, the solubility and transmission path of the material have significant differences at different temperatures and pressures. Through the multi-field coupled transmission equation, the present invention can take into account the effects of factors such as temperature gradient, pressure gradient and concentration diffusion at the same time, and comprehensively simulate the dissolution and transmission behavior of coal-based normal alkanes, thereby achieving precise control of concentration and transmission path during the extraction process. The dissolution selectivity prediction model is based on the solubility characteristics of normal alkanes with specific carbon chain lengths, and accurately predicts the dissolution selectivity of different carbon chain structures at different temperatures. This process not only ensures the separation of specific target components during the extraction process, but also effectively improves the purity of the extracted product, so that the n-alkane material finally obtained has a uniform carbon chain distribution, ensuring the performance stability of the material.

[0043] The phase change performance prediction function and crystallization process control equation of the present invention play a key role in temperature control and crystallization behavior regulation during the extraction process. Phase change temperature is the core performance parameter of coal-based normal alkanes as phase change materials, which directly affects their energy storage and temperature regulation performance. The present invention predicts and adjusts the phase change temperature in real time during the extraction process through the phase change performance prediction function to ensure that the final product can meet the temperature requirements of specific application scenarios. The phase change performance prediction function can dynamically adjust the phase change temperature of the carbon chain length according to the heat accumulation during the extraction process, thereby controlling the phase change behavior of normal alkanes during the extraction process. Combined with the crystallization process control equation, the present invention accurately controls the crystallinity of coal-based normal alkanes during the extraction process, thereby ensuring the thermal stability of the material after the phase change. The crystallization process control equation ensures the adaptability of the crystallinity under different conditions by adjusting factors such as temperature and concentration, so that the phase change temperature and crystallization state of the material are both optimal. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] 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 or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0045] Figure 1 A schematic diagram of the system structure of a method for optimizing the extraction of coal-based normal alkane phase change materials based on a data-driven model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 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.

[0047] Example 1, reference Figure 1 :A method for optimizing the extraction of coal-based normal alkane phase change materials based on a data-driven model, the method comprising:

[0048] Step 1: Collect experimental data of coal-based normal alkanes, including: coal sample component data, reaction temperature, reaction pressure and reaction time; construct coal sample feature vector based on experimental data;

[0049] As a phase change material, the extraction process of coal-based normal alkanes is highly dependent on the composition of the coal sample and the changes in the reaction conditions. In traditional methods, the extraction process often requires repeated experiments to verify the reaction effects of different coal samples under various conditions, which is not only time-consuming and laborious, but also the accuracy of the data is limited by the slight errors of the experiment, and it is difficult to efficiently obtain accurate prediction results of the extraction conditions. The present invention introduces a data-driven method. First, by collecting the experimental data of the coal sample, including core parameters such as coal sample components, reaction temperature, reaction pressure and reaction time, basic data for a comprehensive description of each coal sample is formed. These data are no longer just experimental observations, but are converted into "feature vectors" of coal samples after feature extraction and vectorization processing, representing the reaction behavior and properties of coal samples under different conditions. The process of constructing the coal sample feature vector is not a simple data summary, but a weight analysis of each experimental parameter by mathematical methods to ensure that the feature vector can cover the multi-dimensional information of the coal sample. Specifically, the composition data of the coal sample includes the relative content of different components, such as the percentage of elements such as carbon, hydrogen, oxygen, sulfur, and other trace elements that may affect the solubility of normal alkanes. These composition information plays a decisive role in the extraction process of coal-based n-alkanes, while reaction temperature, pressure and time directly affect the dissolution behavior of coal-based n-alkanes under dynamic conditions. The construction of feature vectors effectively compresses this information into multidimensional space to form a unified numerical vector, so that the similarities or differences of different coal samples under the same or different conditions can be compared through the distance between vectors. The feature vectors generated in this way have the basis of data-driven models and can predict the dissolution behavior of coal samples under different conditions, especially the relationship between the solubility of n-alkanes in coal samples and specific experimental conditions. The role of feature vectors in data-driven models is similar to a "fingerprint", which is the unique response of coal samples under specific experimental conditions. Through repeated training and iteration, the data-driven model can learn the solubility differences of different coal samples under different conditions based on a large number of coal sample feature vectors and experimental conditions. In future predictions, even feature vectors of unknown coal samples can be used as input data, and the model will generate corresponding prediction results based on existing knowledge, greatly improving the accuracy and efficiency of the n-alkane extraction process.

[0050] Step 2: constructing a dissolution selectivity prediction model based on the coal sample characteristic vector and the dissolution parameter matrix of the coal sample; the output result of the dissolution selectivity prediction model is the solubility of coal-based normal alkanes in the coal sample at a specific reaction temperature;

[0051] In the method of the present invention, the basis of the solubility selectivity prediction model is the solubility parameter matrix constructed by experimental data. The matrix contains the solubility characteristics of different coal samples at different reaction temperatures, integrates the multidimensional data of the solubility behavior, and forms a feature correlation relationship that can be learned by the model. This solubility parameter matrix not only describes the solubility of a single coal sample at a single temperature, but also integrates the various characteristics of the coal sample (such as composition, density, specific heat capacity, etc.) and reaction conditions (such as temperature, pressure, etc.) into a whole through the accumulation of a large amount of experimental data, so that the model can find the corresponding solubility relationship under the combination of different variables. During the training process, the model repeatedly learns these feature combinations, identifies the solubility selectivity law of coal-based normal alkanes in a data-driven manner, and ensures that efficient and accurate predictions can be achieved under specific reaction conditions. The solubility selectivity prediction model is not just a simple mapping of solubility data, but through machine learning methods, the model has the ability to "memorize" the correlation between different feature vectors. In actual operation, when a feature vector of a coal sample is input, the model will automatically find the sample data closest to the feature in the existing solubility parameter matrix, and obtain the solubility of the coal sample under the current conditions through interpolation, fitting and other methods. This method can significantly improve the accuracy of predictions, because the model does not rely on a single solubility data for judgment, but organically links various experimental data through the overall feature correlation, forming a way to expand dissolution behavior from known to unknown. The training of the model not only improves its ability to accurately predict known samples, but also can quickly give solubility prediction results when facing new coal samples or new conditions through the similarity between data. This flexibility makes the model have strong adaptability and generalization ability.

[0052] The model can predict the solubility of a specific coal sample at a specific reaction temperature by combining the coal sample feature vector and the dissolution parameter matrix, thereby providing guidance for the extraction process. This method has significant advantages over traditional experimental methods. Traditional methods usually require multiple experiments at different temperatures to observe the dissolution of normal alkanes in coal samples, which is not only labor-intensive and time-consuming, but may also lead to prediction errors due to the discreteness of the data. The solubility selectivity prediction model makes the effect of temperature on dissolution behavior a quantifiable functional relationship through systematic training and integration of data. After obtaining the coal sample feature vector, the model directly uses the vector and the data in the dissolution parameter matrix for calculation, thereby quickly obtaining the solubility results. This efficiency is particularly important in industrial applications, because the dissolution selectivity of different coal samples is significantly different, and through the prediction of the model, the dissolution of normal alkanes at a specific temperature can be directly obtained, avoiding the cumbersome steps of temperature control and sample testing in traditional experiments. Furthermore, the solubility selectivity prediction model is not only a data mapping tool, but also essentially utilizes the complex nonlinear relationship between the coal sample feature vector and the solubility parameter matrix, so that the model has a deep understanding of the solubility behavior of coal-based normal alkanes. Through training, this model can automatically identify the effect mode of different temperatures on solubility, making the prediction results closer to the actual situation. Whether it is high temperature or low temperature conditions, the model can quickly find the appropriate solubility prediction value based on the known data pattern. This flexible response capability greatly improves the extraction efficiency of normal alkanes. In the process of data input and output, the solubility selectivity prediction model is like a "virtual laboratory". It can give accurate solubility values ​​at any time through existing data, which not only shortens the experimental time, but also makes the prediction of solubility behavior more reliable through systematic model analysis. Finally, the solubility value output by the solubility selectivity prediction model becomes the key input of the subsequent steps, providing basic data for the establishment and optimization of multi-field coupled transmission equations. Through this prediction method, the present invention reduces the dependence on experimental steps in the extraction process, making the process of coal-based normal alkane extraction more automated and intelligent, and also makes the prediction of data-driven models an indispensable part of the extraction method. This model not only provides an accurate basis for the selective dissolution of coal samples, but also maintains efficient predictive capabilities under the influence of multiple variables such as temperature and coal sample components, thereby ensuring the stability and controllability of the extraction process.

[0053] Step 3: Based on the specific heat capacity and density of the coal sample, combined with the output results of the solubility selectivity prediction model, a multi-field coupling transmission equation is constructed, and the multi-field coupling transmission equation is solved to obtain the concentration of coal-based normal alkanes; then, according to the multi-field coupling transmission equation, a separation efficiency prediction model is established and a phase change performance prediction function is constructed; the output result of the separation efficiency prediction model is the separation efficiency; the phase change performance prediction function is solved to obtain the phase change temperature of coal-based normal alkanes; according to the separation efficiency prediction model and the phase change performance prediction function, a crystallization process control equation is established, and the crystallization process control equation is solved to obtain the crystallinity;

[0054] The main task of step 3 is to further construct and solve the multi-field coupled transport equation based on the output results of the solubility selectivity prediction model, combined with the specific heat capacity and density of the coal sample, to accurately simulate the transport process of normal alkanes in the coal sample and obtain its concentration distribution under different conditions. This step is not only a static prediction of solubility, but also requires the simulation of the dynamic changes of solubility under specific reaction conditions by coupling the transport behavior of multiple physical fields. By introducing parameters such as specific heat capacity and density, the multi-field coupled transport equation realizes a comprehensive description of various effects such as heat transfer, mass transfer, and flow, so that the dynamic characteristics of the extraction process can be truly reproduced. On this basis, step 3 further establishes a separation efficiency prediction model and a phase change performance prediction function, and constructs the control equation of the crystallization process based on these models to obtain the phase change temperature, separation efficiency and crystallinity of coal-based normal alkanes. This sophisticated coupling calculation provides a scientific basis for subsequent optimization. In the construction of the transport equation, the introduction of specific heat capacity and density plays an important role. The specific heat capacity reflects the amount of heat absorbed or released by the coal sample during heating, while the density is closely related to the transmission speed and distribution state of the substance. During the transmission process, the differences in specific heat capacity and density under different temperature conditions will cause local changes in solubility, thereby affecting the phase change process of n-alkanes. Through the combination of these parameters, the multi-field coupled transmission equation can capture the concentration distribution of n-alkanes in coal samples and predict the changing trend of solubility under different reaction conditions. This equation not only considers the distribution of n-alkanes in time and space, but also introduces variables such as temperature and concentration to form a dynamic correlation between the output results of the solubility selectivity prediction model and the transmission equation. Based on the solution of the transmission equation, the system can deduce the distribution concentration of coal-based n-alkanes, and then judge the efficiency of its transmission and dissolution process by the concentration change trend.

[0055] In addition to the construction of the transport equation, step 3 also introduces a separation efficiency prediction model and a phase change performance prediction function to further achieve precise control of the separation and phase change process of coal-based n-alkanes. The separation efficiency prediction model uses the results of the transport equation to determine the extraction efficiency of n-alkanes. Specifically, the separation efficiency reflects the proportion of n-alkanes in the coal sample extracted per unit time under specific transport conditions. With the solution of the multi-field coupled transport equation, the separation efficiency prediction model can make quantitative predictions on the separation effect in the extraction process based on data such as the gradient change of concentration and the time evolution of solubility. This prediction can not only help determine the extraction effect of coal-based n-alkanes under the current process conditions, but also serve as a basis for adjusting the extraction parameters to further optimize the efficiency of the entire process. The phase change performance prediction function models the phase change characteristics of coal-based n-alkanes in the extraction process by combining the results of the transport equation. Phase change is a key step in which the properties of coal-based n-alkanes change after being separated from coal samples. This process is affected by multiple factors such as temperature and pressure. The phase change performance prediction function is based on the solubility characteristics and thermodynamic parameters of n-alkanes to calculate the phase change temperature under specific conditions. Through the linkage of the transmission equation and the prediction function, the reaction conditions can be monitored and adjusted in real time to ensure that n-alkanes complete the phase change at an appropriate temperature. The determination of this phase change temperature not only directly affects the extraction purity and quality of the material, but also determines the stability and performance of the phase change material obtained in the end. On the basis of establishing the transmission equation, the separation efficiency prediction model and the phase change performance prediction function, step 3 further constructs the control equation of the crystallization process. The crystallization process is an indispensable part of the extraction process. By precisely controlling the crystallization conditions of n-alkanes, the crystallinity of the phase change material can be improved, thereby optimizing the performance of the material. Crystallinity is an important indicator to characterize the degree of ordered arrangement inside the material. Higher crystallinity usually means better thermodynamic stability and mechanical properties. The crystallization process control equation determines the key conditions such as temperature, time and pressure of n-alkanes in the crystallization process based on the output of the separation efficiency prediction model and the phase change performance prediction function, thereby maximizing its crystallinity under specific conditions. By solving this equation, the system can precisely control the reaction conditions to ensure that coal-based normal alkanes are precipitated under optimal crystallization conditions, ultimately forming a phase change material with higher crystallinity.

[0056] Step 4: Based on the crystallinity, separation efficiency, phase change temperature of coal-based normal alkanes and the concentration of coal-based normal alkanes, a multi-objective optimization function is constructed; by solving the multi-objective optimization function, an optimal process parameter combination is obtained; the optimal process parameter combination includes: an optimal reaction temperature, an optimal reaction pressure and an optimal reaction time; and the optimal process parameter combination is used to optimize the extraction of coal-based normal alkane phase change materials.

[0057] The core of step 4 is to find the optimal combination of process parameters in the extraction process, including the optimal reaction temperature, reaction pressure and reaction time, by constructing a multi-objective optimization function. This step is based on the data and prediction results obtained in the previous steps, with crystallinity, separation efficiency, phase change temperature and concentration of coal-based normal alkanes as key optimization targets, and the multi-objective optimization method is used to achieve fine adjustment of process parameters under complex conditions. The multi-objective optimization function plays a role in integrating all important influencing factors here, and measures and weighs different target parameters through a unified function to ensure that the final parameter combination can achieve the optimal balance in all indicators. The optimization process first introduces each target indicator into a unified evaluation framework. The construction of the evaluation framework involves the weight setting of multiple important factors and the comprehensive consideration of influencing factors. Specifically, crystallinity, as a key indicator of the performance of the final product, affects the structural stability and thermodynamic properties of coal-based normal alkane phase change materials, and therefore occupies an important proportion in the optimization. The separation efficiency determines the extraction rate and quality of coal-based normal alkanes, which directly affects the economy and efficiency of the extraction process. Phase change temperature is an important characteristic of materials in practical applications. It can determine the stability and performance of coal-based normal alkanes under different temperature conditions. Optimizing this temperature parameter can ensure the reliability and durability of the material in application. Concentration, as an immediate variable in the reaction process, reflects the content and distribution state of coal-based normal alkanes under reaction conditions, and can reflect the uniformity and effect of the dissolution process under different reaction conditions. By introducing these target factors, the optimization function integrates them in a multi-dimensional space to find a combination of process conditions that can meet various indicators at the same time.

[0058] The multi-objective optimization adopts a systematic solution method in this process, gradually adjusting the reaction temperature, pressure and time to find a balance point between all the target factors. This optimization solution method not only takes into account the individual extreme values ​​of each target, but also identifies and adjusts the mutual constraints between the targets through multiple iterations. Since there may be a competitive relationship between different targets, such as higher separation efficiency is often accompanied by higher energy consumption or higher reaction temperature, through the optimization process, the system can achieve an ideal balance between separation efficiency and other targets, so that all aspects of the extraction process can be optimized. In the actual solution, the optimization function will adjust the process condition combination as the weights and conditions of each target factor change, and finally find the best extraction condition under the comprehensive evaluation of multiple target parameters. The optimal process parameter combination obtained by solving the multi-objective optimization function enables the extraction of coal-based normal alkanes to achieve the highest extraction efficiency and quality. The entire optimization process is completed through multiple iterations of the model, enabling the system to dynamically respond to and adjust different condition combinations, thereby ensuring that the best extraction effect can be achieved in each extraction.

[0059] Example 2: Coal sample component data includes: reaction degree, volume change, component quantity, carbon content, hydrogen content and oxygen content in the components.

[0060] Specifically, the degree of reaction, as the reaction depth of the coal sample under specific conditions, reveals the process of dissolution and transformation of n-alkanes from the coal sample. In the data-driven model, the degree of reaction reflects the degree of chemical change inside the coal sample, which determines the extraction effect of n-alkanes and their performance in the subsequent separation steps. Therefore, the degree of reaction is not only an experimental variable, but also constitutes an important input in the solubility prediction model, affecting the judgment of the solubility selectivity prediction model on the extraction process. Volume change is another key parameter. Coal samples may undergo significant physical structure changes due to dissolution and separation during the extraction process, thereby changing the dissolution path and transport characteristics. The volume change captures this microstructural transformation, which affects the transport efficiency and path of n-alkanes during the dissolution process. By recording the volume change, the data-driven model can better simulate and predict the behavior of coal samples under different conditions. The data of volume change enables the model to dynamically adapt to the changes in the structure of the coal sample during the reaction process, thereby providing accurate parameters for the transport equation. The volume change also reflects the adjustment of the internal pore structure, which has an important impact on the solubility and transport efficiency of coal-based n-alkanes. The input of volume data ensures the precise control of the coal sample dissolution process at the microscopic level. The carbon, hydrogen and oxygen contents in the coal sample represent the basic chemical composition of the coal sample, which has a direct impact on the extraction process. The carbon content not only affects the basic properties of the coal sample, but also determines the potential for the generation of n-alkanes. The hydrogen content is directly related to the saturation and chemical stability of n-alkanes, and plays an important role in the dissolution and extraction process. The oxygen content may affect the redox characteristics of the coal sample, thereby changing the reaction activity and affecting the dissolution and extraction efficiency. The chemical composition of the coal sample determines the solubility and dissolution selectivity of different coal samples in the dissolution process through the ratio of carbon, hydrogen and oxygen. By analyzing these chemical data, the data-driven model can more accurately predict the extraction effect of n-alkanes under specific reaction conditions, and can dynamically adjust the process parameters such as temperature and pressure in the extraction process to achieve the best extraction effect. The number of components, as a structural feature of the coal sample, has an important application in the extraction process. The number of components not only determines the macroscopic characteristics of the coal sample, but also indirectly affects the degree of influence of the reaction conditions on the solubility. Different component quantities may lead to significant differences in the microstructure of coal samples, thereby changing the transport path of n-alkanes. The data-driven model can use the data on the component quantity to adjust the parameters of the solubility selectivity prediction model, making the extraction process more adaptable and flexible. With the help of information such as reaction degree, volume change, component quantity and carbon, hydrogen and oxygen content, the data-driven model enables the solubility selectivity prediction model to make corresponding adjustments according to the characteristics of different coal samples, thereby accurately controlling the extraction process.

[0061] Example 3: The coal sample feature vector is expressed using the following formula:

[0062]

[0063] Where X(t) is the coal sample characteristic vector at reaction time t; N is the number of components in the coal sample; i is an integer subscript index ranging from 1 to N; C i is the carbon content of the i-th component; e is the natural base; T is the reaction temperature; E i is the activation energy of the i-th component; α j is the reaction degree of the jth component, j is an integer subscript index ranging from 1 to N, j≠i; H i is the hydrogen content of the i-th component; P is the reaction pressure; P0 is the standard pressure; O i is the oxygen content of the i-th component; ΔV i is the volume change of the i-th component; ΔG mix is the mixing Gibbs free energy; τ i is the characteristic relaxation time of the i-th component, and the calculation formula is:

[0064]

[0065] Specifically, each item in the eigenvector corresponds to the behavior of the three key elements of carbon, hydrogen, and oxygen in the coal sample during the reaction process. The first item involves the carbon content C i Description of the activation energy E i , reaction temperature T, and reaction degree α j , etc., represent the reaction trend of carbon under different reaction conditions. The activation energy and reaction temperature determine the reaction rate of each component, among which, represents the dependence of the reaction rate on activation energy and temperature. The higher the temperature or the lower the activation energy, the larger the value of this term, which means that the carbon component reacts faster under these conditions. This part of the formula effectively links parameters such as temperature and activation energy, allowing the data-driven model to accurately describe the reaction behavior of carbon in the extraction process of coal-based normal alkanes based on these variables. The second term involves the hydrogen content H i, which combines the ratio of pressure P to standard pressure P0 and the exponential term of the reaction degree. This formula describes the diffusion behavior of hydrogen components during dissolution and reaction by combining pressure and reaction conditions. Pressure directly affects the transmission rate of hydrogen components, while the degree of reaction in the exponential term reflects the reaction state of hydrogen at different stages of the process. Through the ratio of pressure to standard pressure, this formula not only captures the effect of pressure on the reaction, but also enables the model to predict the dissolution and diffusion characteristics of hydrogen under different pressure conditions. The dynamic change of hydrogen content provides key support for the prediction of the solubility selectivity of normal alkanes. Therefore, the model can predict the behavior of hydrogen components under different conditions based on the actual reaction pressure, thereby providing effective guidance for the extraction process. The third term is about the oxygen content O i The description of the volume change ΔV i and temperature, reflecting the dynamic transport behavior of oxygen due to volume change during the reaction. Oxygen content is crucial to the extraction process of coal-based normal alkanes, and volume change significantly affects the transfer efficiency of oxygen components during dissolution and transport. The exponential term of this formula It reflects the solubility characteristics of oxygen under pressure, volume and temperature conditions. Increased pressure or volume contraction will limit the transmission of oxygen, while increased temperature will promote transmission. This design allows the data-driven model to accurately capture the behavior of oxygen components under different volume changes and temperature conditions, thereby achieving more precise control during the extraction process. In addition, the eigenvector is multiplied by an exponential correction term The effects of the energy state and reaction time of the coal sample are further considered. The mixing Gibbs free energy ΔG mix It describes the energy interaction between different components of the coal sample. Its negative value indicates that the reaction is more stable under lower free energy conditions. The combination of reaction temperature T and mixing Gibbs free energy can capture the energy change characteristics of the system during temperature changes. Therefore, the model can evaluate the stability of each component in the coal sample by calculating temperature and free energy. In addition, It indicates the influence of the characteristic relaxation time of each component on the overall reaction time. Characteristic relaxation time τ i The formula is defined as The reaction rate and delay of each component under different temperature and pressure conditions are described. This time scale allows the model to capture the temporal changes of different components during the reaction process, especially when the reaction conditions change. The model can respond quickly based on the adjustment of these parameters to ensure the stability and efficiency of the extraction and dissolution process of coal-based n-alkanes.

[0066] Example 4: The dissolution selectivity prediction model is expressed using the following formula:

[0067]

[0068] Where S(n,T) represents the solubility of coal-based normal alkanes with a carbon chain length of n in the coal sample at the reaction temperature T; K d is the preset dissolution constant; X Z (t) is the transpose of X(t); μ is the average carbon chain length of coal-based normal alkanes; σ is the standard deviation of the carbon chain length of coal-based normal alkanes; D(T) is the solubility parameter matrix, expressed as:

[0069]

[0070] Where ΔH1 is the heat of solution of the first component, ΔH2 is the heat of solution of the second component, and ΔH N is the heat of solution of the Nth component; E1 is the activation energy of the first component, E2 is the activation energy of the second component, and E N is the activation energy of the Nth component; κ 12 (T) represents the dissolution influence coefficient of the first component on the second component at temperature T. Similarly, κ ij (T) represents the dissolution influence coefficient of the i-th component on the j-th component at temperature T; Z is the transposition operation.

[0071] Specifically, the model first introduces the characteristic vector transposition X of the coal sample Z (t), the role of this step in the model is to directly integrate the chemical composition, physical properties and reactivity characteristics of the coal sample into the solubility calculation. Through the transposition operation, the coal sample feature vector is multiplied by the solubility parameter matrix D(T) to ensure that the dissolution behavior is directly related to the proportion and reactivity of different components, so that the calculation results have a higher material matching. This approach of directly introducing coal sample characteristics into solubility calculations enables the model to quickly reflect the internal structure changes of coal samples under different temperatures and reaction conditions, reflecting the importance of data-driven feature analysis in the prediction of dissolution behavior. The solubility parameter matrix D(T) is one of the key components of the model and contains the heat of solution ΔH of each component. i and activation energy E i , and the solubility influence coefficient κ between different components ij (T). The heat of solution and activation energy represent the absorption of energy and the "difficulty" of the reaction in the model, respectively. They determine the driving force and resistance of the dissolution process at a specific temperature. Specifically, the heat of solution describes the energy requirement and the activation energy reflects the threshold of the reaction. Together, they affect the temperature dependence of the dissolution behavior. For example, when the temperature rises, the contribution of the heat of solution decreases, which accelerates the dissolution process; while lower activation energy makes dissolution behavior more likely to occur. Through this relationship between temperature and energy, the dissolution parameter matrix enables the solubility to be dynamically adjusted in actual temperature changes, thereby realistically simulating the temperature response of the solubility.

[0072] In addition, the dissolution influence coefficient κ ij (T) expresses the nature of the interaction between different components during the dissolution process. These coefficients indicate how one component affects the behavior of another component when it is dissolved, that is, there is a synergistic or competitive effect between the dissolved molecules. For example, the presence of the i-th component may accelerate or inhibit the dissolution of the j-th component, resulting in the dissolution efficiency of a certain n-alkane being affected by other components. This interaction is represented by the dissolution influence coefficient in the matrix, so that the dissolution behavior is not just the result of a single component, but the synergistic effect of the overall components. This design ensures that the solubility prediction model can reasonably express its actual dissolution process when facing complex components. In order to reflect the selectivity of dissolution behavior, the model adds a Gaussian distribution function based on carbon chain length. The distribution of carbon chain length n is described with the mean value μ and standard deviation σ as the center. This design makes the solubility show a certain distribution trend with the change of carbon chain length. For example, when the carbon chain length is close to the average value, the solubility reaches a peak; when the carbon chain length deviates from the average value, the solubility gradually decreases. Through this distribution function, the model realizes the simulation of the dissolution selectivity of n-alkane. Specifically, the model can identify the difference in the effect of carbon chains of different lengths on solubility. This difference reflects the structural preference of n-alkane molecules, which causes long-chain or short-chain molecules to be selectively dissolved or retained during the extraction process. The whole formula also contains a normalization constant K d , which is a preset solubility constant that ensures that the calculated solubility results can be standardized and compared under different conditions. This constant provides a benchmark that enables the model's solubility prediction results under different temperature, composition and structure conditions to be analyzed and compared consistently, ensuring that the results of solubility selectivity have a basis for quantification and comparison. Under different operating conditions, the model is adjusted through a combination of variables to reflect the solubility behavior during the extraction process, making the solubility prediction of the entire system more accurate and effective.

[0073] Embodiment 5: The multi-field coupling transmission equation is expressed using the following formula:

[0074]

[0075] Among them, C n is the concentration of coal-based normal alkanes with carbon chain length n; is the gradient operator; k is the thermal conductivity of the coal sample; ρ is the density of the coal sample; C p is the specific heat capacity of the coal sample; is the reaction temperature change value; is the reaction pressure change value; D eff is the effective diffusion coefficient, which is calculated by the following formula:

[0076] D eff =φ·τ·Dm S(n,T);

[0077] Where, φ is the porosity of the coal sample; D m is the molecular diffusion coefficient of the medium, that is, the free diffusion coefficient of coal-based normal alkanes in the coal sample; τ is the tortuosity, which is a set value greater than 1.

[0078] Specifically, the first term of the equation is expressed by the effective diffusion coefficient D eff and concentration gradient The effective diffusion coefficient plays a regulatory role in the transmission process. It not only depends on the internal pore structure of the coal sample, but also is limited by the molecular characteristics and solubility selectivity of coal-based normal alkanes. The calculation formula of the effective diffusion coefficient is D eff =φ·τ·D m S(n,T), where the porosity φ represents the size of the space in which molecules can flow freely in the coal sample, and the tortuosity τ is used to characterize the complexity of the transmission path. These factors work together to reflect the permeability and difficulty of molecules in the coal sample. Molecular diffusion coefficient D m It describes the free diffusion rate of coal-based n-alkanes in the coal sample medium, while the solubility selectivity S(n,T) combines the effects of carbon chain length n and reaction temperature T on solubility. As an important factor, solubility selectivity controls the transport rate of n-alkanes under specific conditions, allowing the model to accurately simulate the transport behavior under different carbon chain structures and temperatures. Through the combined influence of these parameters, the effective diffusion coefficient has become the core of expressing the diffusion behavior of coal-based n-alkanes in coal samples. The second term of the transport equation involves the thermal diffusion effect, which is expressed by the thermal conductivity k, the coal sample density ρ and the specific heat capacity C p The temperature gradient is described by a combination of The thermal conductivity k describes the ability of the coal sample to transfer heat from the high temperature area to the low temperature area, while the density ρ and specific heat capacity C p It reflects the characteristics of coal sample absorbing and releasing heat during heat flow conduction. p It represents the ability of coal sample to absorb heat per unit mass, while density ρ represents the mass of coal sample per unit volume. These characteristics are reflected in the transport model as the regulatory mechanism of thermal diffusion on the transport behavior of n-alkanes. By introducing the temperature gradient, the equation can reflect the distribution change of temperature at different spatial positions and the direct impact of this change on dissolution and diffusion. The temperature gradient will cause coal-based n-alkanes to diffuse from high-temperature areas to low-temperature areas. This is because the molecules have higher kinetic energy in high-temperature environments, making it easier to enter the pores of coal samples for transport.

[0079] The third term is the expression of the pressure-driven effect, through the pressure gradient The introduction of , the equation can describe the driving effect of pressure changes on the transport of n-alkanes. During the extraction process, pressure changes will significantly affect the movement trend and transport speed of coal-based n-alkanes in the coal sample. The solubility selectivity S(n,T) in this term appears again, combining the solubility under specific temperature and carbon chain length conditions, combined with the gas constant R and temperature T to form The coefficient of . This coefficient indicates that the dissolution and transport process of n-alkanes will be significantly enhanced under high temperature or high pressure difference conditions. In actual operation, by adjusting the reaction pressure, the equation can reasonably predict the transport behavior of dissolved substances under different pressure conditions, so that the model can flexibly adapt to different reaction conditions. The introduction of pressure gradient enables the equation to simulate the transport trend of coal-based n-alkanes from high pressure areas to low pressure areas, ensuring that the transport model fully reflects the influence of multiple fields. The design of multi-field coupled transport equation enables the temperature, pressure and diffusion effects to be organically combined in the transport model, and the mutual influence between the three is fully expressed in the dynamic changes of the dissolution and extraction process. In this model, the coupling of each field reflects the complex physical relationship in the actual extraction process, so that the transport path, velocity and distribution characteristics of n-alkanes are accurately described. The change of concentration over time is manifested through the joint action of these multiple physical fields. Different temperature, pressure and pore conditions will affect the dissolution, diffusion and transport efficiency of n-alkanes, making the equation more adaptable and predictive.

[0080] Example 6: The separation efficiency prediction model is expressed using the following formula:

[0081]

[0082] Among them, η(t) represents the separation efficiency at reaction time t; X(0) represents the coal sample characteristic vector at the initial reaction time.

[0083] Specifically, in the separation efficiency prediction model, the numerator of the formula contains the integral term It is used to describe the cumulative effect of the dissolution rate of coal-based normal alkanes over time. The solubility selectivity s(n,T) represents the selective preference for dissolution under carbon chain length n and temperature T, which means that the dissolution behavior of normal alkanes with a specific carbon chain length will be different under different temperature conditions. It is a kinetic index in the dissolution process, which indicates the concentration change of n-alkanes with a carbon chain length of n per unit time. The integral result of this item provides the cumulative amount of the dissolution rate from the initial moment to the reaction time t, which reflects the total amount of n-alkanes dissolved in a specific time. Therefore, this part describes the process of gradual dissolution and separation of n-alkanes under reaction conditions. The denominator X Z(0)·X(0) is the product of the initial coal sample feature vector X(0) and its transpose, which serves as a normalization factor to quantify the overall characteristics of the initial coal sample. This normalization operation ensures that the calculation results of the separation efficiency are comparable under different coal samples and conditions. The initial coal sample feature vector X(0) contains key component information of the coal sample, such as the content of carbon, hydrogen, oxygen, etc., as well as other solubility characteristics and physical parameters. These characteristics determine the basic dissolution behavior of the coal sample at the initial moment of the extraction process. By dividing by the result of the time integration, the model is able to normalize the separation efficiency so that it has a relatively consistent separation metric under different coal samples and extraction conditions. In the last part of the formula, the exponential term The mixing Gibbs free energy ΔG is introduced mix The term reflects the regulatory effect of the thermodynamic stability of the system on the separation efficiency. The mixing Gibbs free energy describes the energy interaction between different components in the mixed state, and a negative value indicates that the system tends to have a higher thermodynamic stability at the current temperature T. A lower free energy means that the system is more likely to remain dissolved in this state, thereby improving the separation efficiency. On the contrary, if the Gibbs free energy is high, the system is unstable, and the solubility and separation efficiency will be reduced. By combining the mixing Gibbs free energy term with the gas constant R and temperature T, the exponential term affects the prediction of the separation efficiency under different temperature conditions, ensuring that the model can adaptively adjust the separation efficiency results under various thermodynamic states. The separation efficiency prediction model therefore accurately describes the separation effect of n-alkanes under specific reaction conditions from both kinetic and thermodynamic aspects, and the structure of the formula ensures a reasonable trade-off between various factors. In terms of kinetics, the accumulation of the dissolution rate reflects that the separation efficiency gradually increases with time, while in terms of thermodynamics, the upper limit and stability of the separation efficiency are adjusted by the mixing Gibbs free energy term. This dual-factor coupling ensures that the model can provide accurate predictions of separation efficiency under different reaction times and temperatures, so that every moment of the extraction process is carried out under optimal conditions, providing a scientific basis for the optimized extraction of coal-based n-alkanes.

[0084] Example 7: The phase change performance prediction function is expressed using the following formula:

[0085]

[0086] Among them, T m (n) is the phase transition temperature of coal-based normal alkanes with a carbon chain length of n; ΔH f is the latent heat; T0 is the preset parameter reference temperature.

[0087] Specifically, in the first part of the formula, the integral term It is used to calculate the total heat released during the separation of normal alkanes within the reaction time t. The separation efficiency η(t) in this term represents the dissolution and extraction effect at the reaction time t. Its introduction in the formula ensures the time dependence of the heat calculation and reflects the dynamic changes of the reaction process over time. The separation efficiency directly affects the rate of the dissolution process, which in turn determines the rate of heat release, and the concentration change rate This describes the dissolution and extraction rate of normal alkanes, further incorporating the kinetics of the extraction process into the model. This rate is multiplied by the separation efficiency and the latent heat of phase change ΔH f The product of ΔH and the integration over time can give a dynamic heat accumulation value, which describes the total heat of dissolution from the initial moment to the current moment. This heat directly reflects the reaction of the extraction process of normal alkanes near the phase transition temperature, thus providing a basis for the prediction of the phase transition temperature. f is another key parameter that describes the energy conversion required for the phase change of n-alkanes. In general, latent heat represents the amount of heat required for a substance to undergo a phase change, whether from solid to liquid or from liquid to gas. The introduction of this term ensures that the energy requirements of n-alkanes in a specific state are taken into account in the prediction of the phase transition temperature, so as to accurately calculate how the energy accumulated during its dissolution drives the phase change process. The combination of latent heat of phase change and separation efficiency enables the model to accurately predict the trend of phase transition temperature based on the energy accumulated during the current reaction time. The denominator is the heat capacity of the coal sample, where C p is the specific heat capacity of the coal sample. p It indicates the amount of heat absorbed by a unit mass of coal sample under temperature change. This item directly affects the ability of coal sample to absorb heat during dissolution. n Integrate and multiply by the specific heat C p , which is designed to express the heat absorption capacity of the coal sample during the entire reaction process. The larger the heat capacity of the coal sample, that is, the specific heat capacity C pThe higher the value, the smaller the rate of temperature change of the system, which means that higher heat capacity will make it more difficult for the system to reach the phase change temperature. Therefore, by using heat capacity as the denominator, the model achieves a reasonable balance between heat accumulation and temperature change, ensuring that the calculation results of the phase change temperature are accurate and adaptable under different coal samples and reaction conditions. The preset reference temperature T0 is the benchmark temperature of the model. By adding it to the heat accumulation value, the actual phase change temperature prediction under different reaction times is obtained. The setting of this reference temperature provides a basic value for the model, so that the temperature can return to the initial benchmark level in the absence of reaction or heat change. This temperature benchmark makes the prediction of phase change temperature more flexible and adaptable. By accumulating heat effects on the basis of the benchmark temperature, the model can accurately adjust the temperature prediction value under different environments and reaction conditions.

[0088] Example 8: The crystallization process control equation is expressed using the following formula:

[0089]

[0090] Among them, X c is the crystallinity; k0 is the crystallization rate constant; is the saturated concentration of coal-based normal alkanes with a carbon chain length of n, calculated based on the Wilson equation:

[0091]

[0092] Where ΔH sol is the dissolution enthalpy change; C0 is the saturation concentration at the reference temperature T0.

[0093] Specifically, the main structure of the formula is based on the crystallinity X c The rate of change Where k0 is the crystallization rate constant, which determines the crystallization speed under given conditions. This rate constant is related to factors such as the properties of the material, solubility, and solution environment. Through model settings, it is ensured that the crystallization process has high adaptability under different conditions. The separation efficiency η(t) is introduced into the formula to describe the extraction efficiency of normal alkanes at a specific time point, which directly affects the crystallization rate. The higher the separation efficiency, the more normal alkanes are separated from the solution per unit time, thereby providing more nucleation materials for the crystallization process. In the formula, The term reflects the effect of temperature on the crystallization process. T is the actual reaction temperature, T m(n) is the phase transition temperature of normal alkane with a carbon chain length of n. The absolute value of the difference between the two indicates the degree of deviation between the actual temperature and the phase transition temperature. By introducing the combination of the gas constant R and the temperature T, the exponential term indicates that when the actual temperature is close to the phase transition temperature, the crystallization rate will be significantly accelerated; conversely, when the temperature deviation is large, the crystallization rate will slow down. This term ensures that the crystallization process is optimized when the temperature conditions are close to the phase transition temperature by adjusting the driving effect of the temperature on the crystallization process, thereby achieving the best crystallization effect. The closer the actual temperature is to the phase transition temperature, the closer the exponential term value is to 1, the greater the rate of change of crystallinity, and the more efficient the crystallization process of normal alkanes. The last part of the formula is the dynamic feedback control term for the concentration of normal alkanes. n is the concentration of normal alkanes with a carbon chain length of n at the current moment, is the saturation concentration of the carbon chain length at a specific temperature. It indicates the maximum amount of normal alkanes that can be dissolved in the solution at the current temperature T. By controlling C n and The formula ensures that when the current concentration approaches the saturation concentration, the rate of increase of crystallinity will gradually slow down to avoid instability of the crystallization process caused by oversaturation. The design of the saturation concentration control item allows the crystallization process to proceed rapidly when the concentration of normal alkanes is far below the saturation concentration; and when the concentration approaches the saturation state, the crystallization rate slows down, thus forming a stable crystal structure. In addition, the saturation concentration It is calculated based on the Wilson equation, and its formula is This equation describes the relationship between the saturation concentration of n-alkanes and the heat of solution ΔH. sol , gas constant R, current temperature T and reference temperature T0. Dissolution heat change ΔH sol represents the energy conversion required by n-alkanes during the dissolution process; when the temperature changes, the saturation concentration will be adjusted accordingly, so that the equation can dynamically calculate the saturation concentration of n-alkanes with different carbon chain lengths according to the actual temperature conditions, thereby further affecting the results of the crystallization process control equation. This saturation concentration calculation formula based on the heat of solution ensures the temperature dependence of the crystallization process, so that the crystallization behavior of n-alkanes can be precisely controlled at different reaction temperatures.

[0094] Example 9: The multi-objective optimization function is calculated using the following formula:

[0095]

[0096] Among them, T target is the target phase transition temperature; E total is the total energy consumption; P opt is the optimal process parameter combination.

[0097] Specifically, the first part of the multi-objective optimization function Combining crystallinity, separation efficiency, dissolution selectivity and total energy consumption constitutes the core measure of extraction efficiency. C It is a key parameter to describe the order of the internal structure of the material. Higher crystallinity often means better stability and performance of the phase change material in the application. Therefore, combinations with higher crystallinity are given priority in the optimization process. The separation efficiency η(t) represents the extraction efficiency of coal-based normal alkanes within the reaction time t, which reflects the speed and proportion of the material separated from the solution. Since the separation efficiency is affected by the dissolution rate and the solution concentration, the separation efficiency gradually increases over time. This change ensures that the optimization function can dynamically feedback the extraction process. The solubility selectivity S(n,T) is a solubility characteristic index based on the carbon chain length n and the temperature T, which is used to represent the solubility behavior under specific conditions. Normal alkanes of different carbon chain lengths have different dissolution tendencies at different temperatures. By incorporating the solubility selectivity into the optimization model, the best dissolution effect can be ensured at the target temperature. The product of these three represents the overall performance of the extraction efficiency under specific conditions. By integrating them, the optimization function can accurately evaluate the extraction effect under different process parameter combinations. Energy consumption is an important constraint of this optimization function. The formula is expressed by the total energy consumption E total The denominator of standardizes the extraction efficiency. Total energy consumption refers to the total amount of energy consumed in the extraction process. In actual operation, the lower the energy consumption, the more conducive it is to saving costs and improving the economy of the extraction process. Therefore, by putting the total energy consumption into the denominator, the optimization function tends to choose a combination of process conditions with high efficiency and low consumption. Standardization ensures that the extraction efficiency under different parameter combinations is comparable, so that the model can find the best solution under different energy consumption levels. The optimization function uses the ratio of extraction effect to energy consumption as the evaluation criterion, which improves the extraction quality while ensuring the economic feasibility of the process. Another key part of the formula is This index term measures the actual temperature T and the target phase transition temperature T by the temperature matching degree. target The degree of closeness. m (n) represents the actual phase transition temperature under the current carbon chain length, and T target is the pre-set target temperature. In this item, the absolute difference between the actual temperature and the target temperature is normalized by the gas constant R and the temperature T, and characterized in an exponential manner. This design ensures that when the actual temperature is close to the target temperature, the exponential term is close to 1, and the overall value of the optimization function increases; when the deviation between the two is large, the exponential term approaches 0, and the overall optimization effect decreases. This item not only ensures that the extraction process is as close to the ideal temperature conditions as possible, but also incorporates tolerance for temperature changes in the process parameter selection, ensuring that the model can still find the best process parameter combination under a wide range of temperature adjustments.

[0098] The present invention is described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the extraction of coal-based normal alkane phase change materials based on a data-driven model, characterized in that: The method comprises: Step 1: Collect experimental data of coal-based normal alkanes, including: coal sample component data, reaction temperature, reaction pressure and reaction time; construct coal sample feature vector based on experimental data; Step 2: constructing a dissolution selectivity prediction model based on the coal sample characteristic vector and the dissolution parameter matrix of the coal sample; the output result of the dissolution selectivity prediction model is the solubility of coal-based normal alkanes in the coal sample at a specific reaction temperature; Step 3: Based on the specific heat capacity and density of the coal sample, combined with the output results of the solubility selectivity prediction model, a multi-field coupling transmission equation is constructed, and the multi-field coupling transmission equation is solved to obtain the concentration of coal-based normal alkanes; then, according to the multi-field coupling transmission equation, a separation efficiency prediction model is established and a phase change performance prediction function is constructed; the output result of the separation efficiency prediction model is the separation efficiency; the phase change performance prediction function is solved to obtain the phase change temperature of coal-based normal alkanes; according to the separation efficiency prediction model and the phase change performance prediction function, a crystallization process control equation is established, and the crystallization process control equation is solved to obtain the crystallinity; Step 4: Based on the crystallinity, separation efficiency, phase change temperature of coal-based normal alkanes and the concentration of coal-based normal alkanes, a multi-objective optimization function is constructed; by solving the multi-objective optimization function, an optimal process parameter combination is obtained; the optimal process parameter combination includes: an optimal reaction temperature, an optimal reaction pressure and an optimal reaction time; and the optimal process parameter combination is used to optimize the extraction of coal-based normal alkane phase change materials.

2. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 1, characterized in that: The coal sample component data include: reaction degree, volume change, component quantity, carbon content, hydrogen content and oxygen content in the components.

3. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 2, characterized in that: The coal sample feature vector is expressed using the following formula: Where X(t) is the coal sample characteristic vector at reaction time t; N is the number of components in the coal sample; i is an integer subscript index ranging from 1 to N; C i is the carbon content of the i-th component; e is the natural base; T is the reaction temperature; E i is the activation energy of the i-th component; α j is the reaction degree of the jth component, j is an integer subscript index ranging from 1 to N, j≠i; H i is the hydrogen content of the i-th component; P is the reaction pressure; P0 is the standard pressure; O i is the oxygen content of the i-th component; ΔV i is the volume change of the i-th component; ΔG mix is the mixing Gibbs free energy; τ i is the characteristic relaxation time of the i-th component, and the calculation formula is:

4. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 3, characterized in that: The dissolution selectivity prediction model is expressed using the following formula: Where S(n,T) represents the solubility of coal-based normal alkanes with a carbon chain length of n in the coal sample at the reaction temperature T; K d is the preset dissolution constant; X Z (t) is the transpose of X(t); μ is the average carbon chain length of coal-based normal alkanes; σ is the standard deviation of the carbon chain length of coal-based normal alkanes; D(T) is the solubility parameter matrix, expressed as: Where ΔH1 is the heat of solution of the first component, ΔH2 is the heat of solution of the second component, and ΔH N is the heat of solution of the Nth component; E1 is the activation energy of the first component, E2 is the activation energy of the second component, and E N is the activation energy of the Nth component; κ 12 (T) represents the dissolution influence coefficient of the first component on the second component at temperature T. Similarly, κ ij (T) represents the dissolution influence coefficient of the i-th component on the j-th component at temperature T; Z is the transposition operation.

5. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 4, characterized in that: The multi-field coupled transmission equation is expressed using the following formula: Among them, C n is the concentration of coal-based normal alkanes with carbon chain length n; is the gradient operator; k is the thermal conductivity of the coal sample; ρ is the density of the coal sample; C p is the specific heat capacity of the coal sample; is the reaction temperature change value; is the reaction pressure change value; D eff is the effective diffusion coefficient, which is calculated by the following formula: D eff =φ·τ·D m ·S(n,T); Where, φ is the porosity of the coal sample; D m is the molecular diffusion coefficient of the medium, that is, the free diffusion coefficient of coal-based normal alkanes in the coal sample; τ is the tortuosity, which is a set value greater than 1.

6. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 5, characterized in that: The separation efficiency prediction model is expressed using the following formula: Among them, η(t) represents the separation efficiency at reaction time t; X(0) represents the coal sample characteristic vector at the initial reaction time.

7. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 6, characterized in that: The phase change performance prediction function is expressed using the following formula: Among them, T m (n) is the phase transition temperature of coal-based normal alkanes with a carbon chain length of n; ΔH f is the latent heat; T0 is the preset parameter reference temperature.

8. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 7, characterized in that: The crystallization process control equation is expressed as follows: Among them, X c is the crystallinity; k0 is the crystallization rate constant; is the saturated concentration of coal-based normal alkanes with a carbon chain length of n, calculated based on the Wilson equation: Where ΔH sol is the dissolution enthalpy change; C0 is the saturation concentration at the reference temperature T0.

9. The method for optimizing extraction of coal-based normal alkane phase change materials based on a data-driven model according to claim 8, characterized in that: The multi-objective optimization function is calculated using the following formula: Among them, T target is the target phase transition temperature; E total is the total energy consumption; P opt is the optimal process parameter combination.

Images