Lithium hexafluorophosphate adsorption tower and regeneration process and parameter optimization method thereof

Abstract

The invention discloses a lithium hexafluorophosphate adsorption tower as well as a regeneration process and a parameter optimization method thereof. The adsorption tower comprises a first gas inlet, a second gas inlet, a gas distribution cavity, a pressure equalizing cavity, a grid plate, a porous pressing plate, an adsorbent and the like. Through the matching effect of all the components, regenerated gas sequentially passes through the gas distribution cavity, the adsorbent, the pressure equalizing cavity and the adsorbent in the process of passing through the adsorption tower from the second gas inlet, so that the radial diffusion is increased, the pressure drop of the regenerated gas is balanced, the short circuit of the gas is reduced, and the regeneration effect is improved. And after the size of the device is increased, the total regenerated nitrogen purging time is not greatly increased, so that the productivity and the operation stability of the device are improved. Meanwhile, through the methods of adsorption tower regeneration parameter acquisition, data integration and denoising, data division, machine learning model establishment and model training / verification / test, a machine learning model of adsorption tower regeneration time and related parameters can be obtained, and related process parameters are optimized through a genetic algorithm according to the capacity requirement of the device.

Description

Technical Field

[0001] The present invention relates to the technical field of adsorption separation, and in particular, to a lithium hexafluorophosphate adsorption tower, its regeneration process, and parameter optimization method. Background Art

[0002] Lithium hexafluorophosphate is the most widely used battery electrolyte at present. The industrial preparation methods of lithium hexafluorophosphate include gas-solid reaction method, hydrofluoric acid solvent method, ion exchange method, and organic solvent method. Among them, the organic solvent method has a simple process flow, mild operating conditions, less equipment investment, and no pollution generated during the production process. At the same time, the use of hydrogen fluoride is avoided, reducing the production risk and the material requirements for equipment. Moreover, the organic solvent method has a fast reaction rate, high product purity, and the obtained electrolyte can be directly used in lithium-ion batteries with a high yield. Therefore, the organic solvent method is widely used in current industrial installations for the production of lithium hexafluorophosphate.

[0003] However, during the production by the organic solvent method, the reaction raw materials will react with impurities in the organic solvent to generate acidic impurities such as HF. Since the battery industry has high requirements for the acidity of lithium hexafluorophosphate, an adsorption tower is needed to adsorb acidic impurities such as HF. At present, some existing technologies have studied adsorbents suitable for this process. In fact, the bottleneck of the lithium hexafluorophosphate industrial adsorption tower lies in the nitrogen purging process during the regeneration of the adsorption tower. The nitrogen purging process during the regeneration of the adsorption tower is the most time-consuming process, accounting for about two-thirds of the total regeneration time. If the size of the adsorption tower is enlarged, the nitrogen purging time during regeneration will increase disproportionately and even double, seriously affecting the production capacity of the lithium hexafluorophosphate industry. Constrained by the above factors, currently, industrial installations usually adopt the method of setting multiple small-sized adsorption towers to improve production capacity, but this greatly increases the equipment investment, and the adsorption capacity of a single adsorption tower is limited, so frequent switching of the adsorption tower is required, which greatly increases the labor cost and instability of production. Summary of the Invention

[0004] To solve the above defects in the prior art, the present invention provides a lithium hexafluorophosphate adsorption tower, its regeneration process, and parameter optimization method. By improving the structure of the adsorption tower, the nitrogen purging effect during regeneration is effectively improved. When the size of the adsorption tower is enlarged, the total nitrogen purging time during regeneration does not increase significantly, thereby reducing the equipment investment, improving the production capacity of the device and the stability of operation. At the same time, the present invention also proposes a parameter optimization method for the adsorption tower, which can optimize relevant parameters to obtain the adsorption tower size and operating parameters most suitable for the current production capacity.

[0005] To achieve the above-mentioned invention object, a first aspect of the present invention provides a lithium hexafluorophosphate adsorption tower, which includes a tower body, at least one gas distribution chamber, at least one pressure equalizing chamber, a grid plate, a porous pressing plate and an adsorbent arranged inside the tower body, a first air inlet, a second air inlet and an exhaust port connected to the tower body; wherein,

[0006] The first air inlet and the second air inlet are used to transport the regeneration gas into the interior of the tower body from different directions, and the exhaust port is used to discharge the regeneration tail gas from the tower body;

[0007] The gas distribution chamber is a hollow annular cavity vertically arranged inside the tower body. The upper end wall and the lower end wall of the gas distribution chamber are both sealed, and the side wall of the gas distribution chamber is provided with mesh holes for gas to enter and exit;

[0008] The pressure equalizing chamber is a hollow annular cavity vertically arranged inside the tower body. The upper end wall and the lower end wall of the pressure equalizing chamber are both sealed, and the side wall of the pressure equalizing chamber is provided with mesh holes for gas to enter and exit. The gas distribution chamber is located in the peripheral area of the pressure equalizing chamber;

[0009] The porous pressing plate is radially arranged at the upper part of the tower body and is located above the gas distribution chamber and the pressure equalizing chamber. The lower end part of the gas distribution chamber and the lower end part of the pressure equalizing chamber are connected to the radially arranged grid plate, and the adsorbent is filled in the area between the grid plate and the porous pressing plate.

[0010] Further, the first air inlet is connected to the lower head of the tower body, the exhaust port is connected to the upper head of the tower body, and the second air inlet is connected to the side wall of the tower body.

[0011] Further, the height of the pressure equalizing chamber is not less than the height of the gas distribution chamber, and the volume of the pressure equalizing chamber is not less than the volume of the gas distribution chamber.

[0012] Further, the pressure equalizing chamber is located in the central area inside the tower body, and the gas distribution chamber is located in the peripheral area of the pressure equalizing chamber.

[0013] Further, the diameter of the gas distribution chamber is 0.02-0.2 times the diameter of the tower body.

[0014] Further, the diameter of the pressure equalizing chamber is 0.05-0.3 times the diameter of the tower body.

[0015] Further, the height of the pressure equalizing chamber is 0.6-1 times the height of the adsorbent filling.

[0016] Further, the distance between the centers of the cross-sections of the pressure equalizing chamber and the gas distribution chamber on the same plane is 0.1-0.3 times the diameter of the tower body.

[0017] Further, both the upper and lower ends of the gas distribution chamber and the pressure equalizing chamber are connected to the grid plate to achieve a sealed state.

[0018] In the second aspect of the present invention, a regeneration process for a lithium hexafluorophosphate adsorption tower is provided. Using the adsorption tower as described above, it includes: heating nitrogen to 80 - 100 °C, introducing it into the adsorption tower from the first inlet and the second inlet for purging, and discharging the tail gas from the exhaust port.

[0019] In the third aspect of the present invention, a method for optimizing the parameters of a lithium hexafluorophosphate adsorption tower is provided, including the following steps:

[0020] 1) Collect parameter data of the adsorption tower;

[0021] 2) Process the parameter data to remove deviation data;

[0022] 3) Divide the processed data into a training set, a validation set, and a test set according to the leave-one-out method;

[0023] 4) Create based on the neural network for the training set data to obtain the following training model:

[0024] Y = f(x1, x2, x3,..., x n )

[0025] where x1, x2, x3, x n are process operation parameters, f is the relationship formed by the trained neural network model with each process operation parameter, and Y is the final output result;

[0026] 5) Use the validation set data to validate the obtained training model, and use the test set data to test the obtained training model to obtain a machine learning model;

[0027] 6) Set the shortest nitrogen purging regeneration time as the objective function, and use the genetic algorithm in the machine learning model for optimization design to obtain the optimized parameters under the specified objective function.

[0028] Further, the parameters include but are not limited to production capacity, adsorbent bulk density, adsorbent saturated adsorption capacity, regeneration cycle, adsorption tower diameter, adsorption tower height, number of adsorption towers, gas distribution chamber diameter, pressure equalizing chamber diameter, pressure equalizing chamber height, total amount of regeneration gas, total regeneration nitrogen purging time, or a combination thereof.

[0029] Further, the processing in step 2) includes but is not limited to: integration and denoising, removal of duplicate data, handling of missing values, unification of data formats, correction of data errors, handling of outliers, to ensure the accuracy and consistency of the data.

[0030] The beneficial effects of the present invention are as follows: In the present invention, a lithium hexafluorophosphate adsorption tower is provided, which includes a first air inlet, a second air inlet, a gas distribution chamber, a pressure equalizing chamber, a grid plate, a porous pressing plate, an adsorbent, etc. The first air inlet and the second air inlet are used to convey the regeneration gas. The gas distribution chamber is used to evenly distribute the regeneration gas conveyed by the second air inlet. The pressure equalizing chamber is used to balance the pressure drop of the regeneration gas. Through the cooperative action of the above-mentioned components, when the regeneration gas passes through the adsorption tower from the second air inlet, it sequentially passes through the gas distribution chamber, the adsorbent, the pressure equalizing chamber, and the adsorbent, increasing the radial diffusion, balancing the pressure drop of the regeneration gas, reducing the gas short circuit, and improving the regeneration effect. Moreover, after increasing the equipment size, the total regeneration nitrogen purging time does not increase significantly, thereby improving the device production capacity and operation stability and reducing the equipment investment. At the same time, through the methods of obtaining the adsorption tower regeneration parameters, data integration and denoising, data partitioning, establishing a machine learning model, and training / validating / testing the model. A machine learning model of the adsorption tower regeneration time and related parameters can be obtained, and the equipment size and related process parameters are optimized through a genetic algorithm according to the device production capacity requirements. For an already operating device, according to the device operation data, the direction of parameter optimization and improvement can also be proposed to improve the adsorption tower regeneration effect, reduce the operation cost, and improve the device production capacity.

[0031] Other features and advantages of the present invention will be described in detail through the following specific embodiments. Brief Description of the Drawings

[0032] Figure 1 It is a schematic structural diagram of the adsorption tower for lithium hexafluorophosphate provided by the embodiment of the present invention.

[0033] Figure 2 For illustration Figure 1 It is a cross-sectional view showing the positional relationship between the gas distribution chamber and the pressure equalizing chamber.

[0034] Wherein the reference numerals are: 1 tower body; 2 first air inlet; 3 second air inlet; 4 exhaust port; 5 adsorbent; 6 gas distribution chamber; 7 pressure equalizing chamber; 8 grid plate; 9 grille (including a vertically arranged annular grille 9.1 and a radially arranged planar grille 9.2); 10 porous pressing plate. Detailed Description of the Embodiments

[0035] Hereinafter, the concept of the present application and the technical effects generated will be clearly and completely described in combination with the embodiments to fully understand the purpose, features, and effects of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, other embodiments obtained by those skilled in the art without creative efforts shall fall within the scope of protection of the present application.

[0036] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0037] The first aspect of the present invention provides a lithium hexafluorophosphate adsorption tower, which includes a tower body, at least one gas distribution chamber, at least one pressure equalizing chamber, a grid plate, a porous pressing plate and an adsorbent provided in the tower body, a first air inlet, a second air inlet and an exhaust port connected to the tower body; wherein,

[0038] The first air inlet and the second air inlet are used to transport the regeneration gas into the interior of the tower body from different directions, and the exhaust port is used to discharge the regeneration tail gas from the tower body;

[0039] The gas distribution chamber is a hollow annular cavity vertically arranged inside the tower body. The upper end wall and the lower end wall of the gas distribution chamber are both sealed, and the side wall of the gas distribution chamber is provided with mesh holes for gas inlet and outlet;

[0040] The pressure equalizing chamber is a hollow annular cavity vertically arranged inside the tower body. The upper end wall and the lower end wall of the pressure equalizing chamber are both sealed, and the side wall of the pressure equalizing chamber is provided with mesh holes for gas inlet and outlet. The gas distribution chamber is located in the peripheral area of the pressure equalizing chamber;

[0041] The porous pressing plate is radially arranged at the upper part of the tower body and is located above the gas distribution chamber and the pressure equalizing chamber. The lower end part of the gas distribution chamber and the lower end part of the pressure equalizing chamber are connected to the radially arranged grid plate, and the adsorbent is filled in the area between the grid plate and the porous pressing plate.

[0042] In some embodiments, such as Figure 1 In the shown adsorption tower, the first air inlet 2 is connected to the lower head of the tower body, the exhaust port 4 is connected to the upper head of the tower body, and the second air inlet 3 is connected to the side wall of the tower body.

[0043] In the present invention, a part of the regeneration gas enters the adsorbent through the grid plate via the first air inlet, passes through the porous pressing plate, and is discharged from the tower body through the exhaust port. A part of the regeneration gas enters the gas distribution chamber through the second air inlet, enters the adsorbent and the pressure equalizing chamber through the mesh holes on the side wall of the gas distribution chamber, and then passes through the porous pressing plate and is discharged from the tower body through the exhaust port. The pressure equalizing chamber is located in the adsorbent filling area.

[0044] In some embodiments, such as Figure 1For the adsorption tower shown, the flow rate distribution of the regeneration gas entering the first gas inlet 2 and the regeneration gas entering the second gas inlet 3 can be determined according to the height-diameter ratio of the adsorption tower. If the height-diameter ratio of the adsorption tower ≤ 1.5, then the preferred distribution ratio of the regeneration gas flow rates at the first gas inlet 2 and the second gas inlet 3 is 1 - 3; if the height-diameter ratio of the adsorption tower > 1.5, then the preferred distribution ratio of the regeneration gas flow rates at the first gas inlet 2 and the second gas inlet 3 is 0.4 - 1. Optimizing the flow rate according to the above ratios can increase the gas-phase mass transfer driving force, reduce the pressure drop across the entire tower, improve the uniformity of the gas flow field distribution, reduce local mass transfer dead zones, and enhance the regeneration effect.

[0045] In some embodiments, the height of the pressure equalization chamber is not less than the height of the gas distribution chamber, and the volume of the pressure equalization chamber is not less than the volume of the gas distribution chamber. The advantage of such a setting is that after the gas contacts the adsorbent through the distribution chamber, its axial pressure drop is relatively balanced with the radial pressure drop after passing through the pressure equalization chamber. The gas can achieve a better distribution effect and a uniform residence time within the adsorption tower, enhancing the overall mass transfer of the adsorption tower and improving the regeneration effect.

[0046] In some embodiments, the diameter of the gas distribution chamber is 0.02 - 0.2 times the diameter of the tower body.

[0047] In some embodiments, the diameter of the pressure equalization chamber is 0.05 - 0.3 times the diameter of the tower body.

[0048] In some embodiments, the height of the pressure equalization chamber is 0.6 - 1 times the height of the adsorbent filling.

[0049] In some embodiments, the distance between the centers of the cross-sections of the pressure equalization chamber and the gas distribution chamber on the same plane is 0.1 - 0.3 times the diameter of the tower body.

[0050] By setting the parameters of the above gas distribution chamber and pressure equalization chamber, it is possible to achieve good gas distribution, reduce short circuits and dead zones, and also make the practical adsorbent filling area of the adsorption tower as large as possible.

[0051] In some embodiments, the gas distribution chamber or the pressure equalization chamber can be fabricated by splicing grid plates and grating plates. The side wall of the gas distribution chamber or the pressure equalization chamber can be made of a vertically arranged annular grid plate A, whose mesh structure is used for gas to enter and exit from the side. The pore diameter is smaller than that of the adsorbent. Both the upper and lower ends of the grid plate A are connected to the annular grating plates to achieve a sealed state. Among them, the purpose of sealing the lower end is to prevent the gases at the first gas inlet and the second gas inlet from interfering with each other, and the purpose of sealing the upper end is to enable the gas to contact the adsorbent radially and avoid axial short circuits. In addition, a radially arranged planar grid plate B is connected below the annular grid plate A used to make the side walls of the gas distribution chamber and the pressure equalization chamber. The grid plate B has small holes, and the pore diameter is smaller than that of the adsorbent. It can both hold the adsorbent and allow the gas to pass through.

[0052] In some embodiments, the structure of the grid plate is an annular steel flat plate or the like, without mesh holes, and has a certain load-bearing capacity.

[0053] In some embodiments, the grid plate can adopt a multi-layer planar mesh framework such as a grille, which is composed of one or more groups of parallel metal grid bars. The mesh holes are 0.2-0.5 times the particle size of the adsorbent, such as 0.3 times, etc.

[0054] In some embodiments, the structure of the porous pressing plate is a planar mesh pressing plate or the like, which is composed of one or more groups of parallel metal grid bars. The mesh holes are 0.3-0.5 times the particle size of the adsorbent, such as 0.3 times.

[0055] In some embodiments, the adsorbent is a lithium-based fluoride adsorbent.

[0056] In some embodiments, the grid plate and the porous plate are fixed inside the tower body by welding, bolt connection, etc. For example, the lower grid plate of the gas distribution chamber and the grid plate B below the adsorbent are jointly welded to the tower body, and the grid plate A on the side wall of the gas distribution chamber is also welded to its lower grid plate. The upper grid plate of the gas distribution chamber is connected to the grid plate A on the side wall by bolts.

[0057] In some embodiments, the numbers of the gas distribution chamber and the pressure equalizing chamber are 1-3 respectively, and the distribution mode between them is arranged at intervals. Preferably, the pressure equalizing chamber is located in the central area inside the tower body, and the gas distribution chamber is located in the peripheral area of the pressure equalizing chamber. The sizes of each gas distribution chamber and the pressure equalizing chamber meet the above description.

[0058] The second aspect of the present invention provides a regeneration process for a lithium hexafluorophosphate adsorption tower. Using the adsorption tower as described above, it includes: heating nitrogen to 80-100 °C, and introducing it into the adsorption tower from the first air inlet and the second air inlet for purging, and the tail gas is discharged from the exhaust port.

[0059] The third aspect of the present invention provides a method for optimizing the parameters of a lithium hexafluorophosphate adsorption tower, including the following steps:

[0060] 1) Collect the parameter data of the adsorption tower;

[0061] 2) Process the parameter data to remove the deviation data;

[0062] 3) Divide the processed data into a data set according to the leave-one-out method, and divide it into a training set, a validation set and a test set;

[0063] 4) Create based on the neural network for the data of the training set to obtain the following training model:

[0064] Y = f(x1, x2, x3,..., x n )

[0065] where x1, x2, x3, x n are process operation parameters, f is the relational expression formed by the trained neural network model with each process operation parameter, and Y is the final output result;

[0066] 5) Use the validation set data to verify the obtained training model, and use the test set data to test the obtained training model to obtain a machine learning model;

[0067] 6) Set the shortest nitrogen purge regeneration time as the objective function, and use the genetic algorithm in the machine learning model for optimization design to obtain the optimized parameters under the specified objective function.

[0068] In the present invention, after obtaining the training model, use the validation set data to verify the obtained model. If the verification result of the validation set data deviates from the preset output accuracy, add the training set model and retrain the model; if the verification result of the validation set data meets the preset output accuracy, use the test set data to test the obtained model. After obtaining the data-based machine learning model as above, the parameters can be optimized according to the model using the objective function. The optimization refers to using the shortest total regeneration nitrogen purge time as the objective function and using the genetic algorithm for optimization design, and various optimal parameters under the specified objective function can be obtained.

[0069] In some embodiments, the parameters include but are not limited to production capacity, adsorbent bulk density, adsorbent saturated adsorption capacity, regeneration period, adsorption tower diameter, adsorption tower height, number of adsorption towers, gas distribution chamber diameter, equalization chamber diameter, equalization chamber height, total amount of regeneration gas, total regeneration nitrogen purge time, or a combination thereof.

[0070] In some embodiments, the processing in step 2) includes but is not limited to: integrating and denoising, removing duplicate data, handling missing values, unifying data formats, correcting data errors, handling outliers to ensure the accuracy and consistency of the data.

[0071] In some embodiments, the creation in step 4) includes establishing an input layer, establishing a hidden layer, establishing an output layer, and model training.

[0072] The present invention will be further described below with examples, but the present invention is not limited thereto.

[0073] Please refer to Figure 1As shown in the figure, an embodiment of the present invention provides a lithium hexafluorophosphate adsorption tower, which includes a tower body 1, a first air inlet 2, a second air inlet 3, an exhaust port 4, an adsorbent 5, a gas distribution chamber 6, a pressure equalizing chamber 7, a grid plate 8, a grid plate 9 (including a vertically arranged annular grid 9.1 and a radially arranged planar grid 9.2), and a porous pressing plate 10. The lower head of the tower body is connected to the first air inlet 2 through an opening, and the side wall of the tower body is connected to the second air inlet 3 through an opening. The first air inlet 2 and the second air inlet 3 are used to transport the regeneration gas into the tower body. The upper head of the tower body is connected to the exhaust port 4 through an opening, which is used to discharge the regeneration tail gas from the tower body. The adsorbent 5 is filled between the grid 9.2 and the porous pressing plate 10; the gas distribution chamber 6 and the pressure equalizing chamber 7 are located inside the tower body and are connected to the adsorbent 5. A part of the regeneration gas enters the adsorbent 5 through the radially arranged grid 9.2 via the first air inlet 2, passes through the porous pressing plate 10, and is discharged from the tower body through the exhaust port 4. A part of the regeneration gas enters the gas distribution chamber 6 through the second air inlet 3, enters the adsorbent 5 and the pressure equalizing chamber 7 through the vertically arranged grid 9.1, passes through the porous pressing plate 10, and is discharged from the tower body through the exhaust port 4. The adsorbent 5 is a lithium-based fluoride adsorbent. The gas distribution chamber 6 is a ring-shaped cavity structure. The gas distribution chamber 6 is connected to the tower body, and the radial direction of the gas distribution chamber 6 is connected to the adsorbent 5 through the grid 9.1. The lower end of the gas distribution chamber is connected to the grid 9.2, and a grid plate 8 is provided at this end for sealing. The upper end of the gas distribution chamber is connected to the grid plate 8, and this end is sealed. The pressure equalizing chamber 7 is a ring-shaped cavity structure. The pressure equalizing chamber 7 is located in the adsorbent filling area. The radial direction of the pressure equalizing chamber 7 is connected to the adsorbent 5 through the grid 9.1. The lower end of the pressure equalizing chamber is connected to the grid 9.2, and a grid plate is provided at this end for sealing. The upper end of the pressure equalizing chamber 7 is provided with a grid plate for sealing. The height of the pressure equalizing chamber 7 is not less than the height of the gas distribution chamber 6, and the volume of the pressure equalizing chamber 7 is not less than the volume of the gas distribution chamber 6.

[0074] Nitrogen is heated to 90 °C as the regeneration gas, the particle size of the adsorbent is 0.6 - 1.2 mm, the mesh size of the grid and the porous pressing plate is 0.3 mm, the grid plate is a ring-shaped steel plate, the grid is a planar mesh frame, and the porous pressing plate is a planar mesh pressing plate.

[0075] Data collection is carried out to obtain data on production capacity, adsorbent bulk density, saturated adsorbent capacity, regeneration cycle, adsorption tower diameter, adsorption tower height, number of adsorption towers, gas distribution chamber diameter, equalizing chamber diameter, equalizing chamber height, total amount of regeneration gas, and total regeneration nitrogen purge time through experiments. Then, data preprocessing is performed to integrate and denoise. After that, the processed data is divided into datasets, and a machine learning model is trained based on the divided data. Then, the parameters are optimized using the objective function. The optimization refers to using the shortest total regeneration nitrogen purge time as the objective function and using the genetic algorithm for optimal design to obtain various optimal parameters under the specified objective function.

[0076] Example 1:

[0077] The production capacity of lithium hexafluorophosphate is 500 kg / h, the acid value of the feed is 500 ppm, an adsorbent with a saturated adsorption capacity of 60 mg / g of adsorbent is selected, and the adsorbent bulk density is 850 kg / m 3 , the adsorbent particle size is 0.8 mm, and the regeneration cycle is 3 months. In the genetic algorithm optimization model, the diameter of the gas distribution chamber is restricted to be 0.02 - 0.2 times the tower diameter, the distance between the equalizing chamber and the gas distribution chamber is 0.1 - 0.3 times the tower diameter, the diameter of the equalizing chamber is 0.05 - 0.3 times the tower diameter, and the height of the equalizing chamber is 0.6 - 1 times the adsorbent filling height. Machine learning is performed to obtain a surrogate model, and parameter optimization is carried out based on the shortest total regeneration nitrogen purge time as the objective function. The specific method is as follows:

[0078] Select a group of adsorption towers, and collect the production capacity Q of the adsorption tower, the adsorbent bulk density ρ, and the saturated adsorbent capacity C sat , regeneration cycle T reg , adsorption tower diameter D col , adsorption tower height H col , number of adsorption towers N col , number of gas distribution chambers N dist , gas distribution chamber diameter D dist , gas distribution chamber height H dist , number of equalizing chambers N eq , equalizing chamber diameter D eq , equalizing chamber height H eq , distance L between the equalizing chamber and the gas distribution chamber, continuously increase the total amount of regeneration gas V reg , obtain the total regeneration nitrogen purge time T total data, and the above completes a set of data records. Select another group of adsorption towers, repeat the above operations, and obtain data for machine learning. Perform preprocessing on the original data, and use the wavelet threshold denoising algorithm to eliminate measurement noise. Its mathematical expression is:

[0079]

[0080] Among them, WT is the wavelet transform, Thresh is the threshold function, and λ is the denoising threshold.

[0081] The preprocessed dataset is divided into a training set Strain and a test set Stest, with a division ratio of 8:2. A machine learning model is established to express the non-linear relationship between parameters:

[0082] Ttotal = f(Q, ρ, Csat, Treg, Dcol, Hcol, Ncol, Ndist, Ddist, Hdist, Neq, Deq, Heq, L, Vreg)

[0083] The random forest regression algorithm is used for model training, and key parameters are screened through feature importance evaluation. The objective function is to minimize the predicted mean squared error:

[0084]

[0085] With the objective of minimizing the total regeneration nitrogen purge time, a constrained optimization problem is constructed and a genetic algorithm is implemented for optimization.

[0086]

[0087] Comparative Example 1:

[0088] The production capacity of lithium hexafluorophosphate is 500 kg / h, the acid value of the feed is 500 ppm, the saturated adsorption capacity of the selected adsorbent is 60 mg / g of adsorbent, and the bulk density of the adsorbent is 850 kg / m 3 , the particle size of the adsorbent is 0.8 mm, and the regeneration period is 3 months. The improved adsorption tower structure of this application is not adopted, and the sizes of the pressure equalization chamber and the gas distribution chamber in the optimization model are limited to 0, and other conditions are the same as those in Example 1.

[0089] The comparison results between Example 1 and Comparative Example 1 are as follows:

[0090] In Example 1, due to the increased radial diffusion, the pressure drop of the regeneration gas is balanced, the gas short-circuit is reduced, and the regeneration nitrogen purge effect is improved. The total regeneration nitrogen purge time does not increase significantly and exponentially with the increase of the adsorption tower size. The optimized parameter results are as follows: the diameter of the adsorption tower is 2.4 m, the height of the adsorption tower is 3.8 m, the number of adsorption towers is 1, the diameter of the gas distribution chamber is 0.15 m, the diameter of the pressure equalization chamber is 0.12 m, the number of the gas distribution chamber and the pressure equalization chamber is 1, the distance between the gas distribution chamber and the pressure equalization chamber is 0.4 m, the pressure equalization chamber is located in the central area inside the tower body, the gas distribution chamber is located in the peripheral area of the pressure equalization chamber, the height of the gas distribution chamber is 2.8 m, the height of the pressure equalization chamber is 3 m, and the total volume of the regeneration gas is 1650 m 3 / h, the total regeneration nitrogen purging time is 25 days, and the flow ratio of the regeneration gas entering the first inlet 2 to the regeneration gas entering the second inlet 3 is 0.8; the results of Comparative Example 1 are as follows: the adsorption tower diameter is 1 m, the adsorption tower height is 1.6 m, the number of adsorption towers is 10, and the total regeneration gas volume is 2100 m 3 / h, and the total regeneration nitrogen purging time is 70 days. The equipment investment, total regeneration gas consumption, and total regeneration nitrogen purging time of Comparative Example 1 are all higher than those of Example 1.

[0091] Comparative Example 2

[0092] It is different from Example 1 in that: one gas distribution chamber is provided, the diameter of the gas distribution chamber is 0.04 m, and the height of the gas distribution chamber is 3.6 m. The number of equalizing chambers is set to 0.

[0093] The comparison results between Example 1 and Comparative Example 2 are as follows:

[0094] Since no equalizing chamber is provided in Comparative Example 2, the total regeneration nitrogen purging time of Comparative Example 2 is extended to 31 days, and the regeneration effect of the adsorption tower is worse than that of Example 1.

[0095] Comparative Example 3

[0096] It is different from Example 1 in that: the height of the equalizing chamber is less than the height of the gas distribution chamber, and the volume of the equalizing chamber is less than the volume of the gas distribution chamber. Specifically, the number of the equalizing chamber and the gas distribution chamber is 1 each, the heights are 0.5 m and 2 m respectively, and the diameters are 0.1 m and 0.2 m respectively.

[0097] The comparison results between Example 1 and Comparative Example 3 are as follows:

[0098] Since the parameter settings of the equalizing chamber in Comparative Example 3 are unreasonable, the gas phase distribution effect of Comparative Example 3 is poor, the pressure drop of the whole tower is large, there is a mass transfer dead zone in the tower, and the regeneration effect is poor.

[0099] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description should not be construed as a limitation of the present invention. Those skilled in the art can understand that under the teaching of this specification, some modifications or adjustments can be made to the present invention. These modifications or adjustments should also be within the scope defined by the claims of the present invention.

Claims

1. A lithium hexafluorophosphate adsorption tower, characterized in that: It comprises a tower body, at least one gas distribution chamber, at least one pressure equalization chamber, a grid plate, a porous pressure plate and an adsorbent arranged in the tower body, and a first air inlet, a second air inlet and an air outlet connected to the tower body; wherein, The first air inlet and the second air inlet are used to transport the regeneration gas into the tower body from different directions, and the exhaust port is used to discharge the regeneration tail gas out of the tower body; The gas distribution cavity is a hollow annular cavity vertically arranged inside the tower body, the upper end wall and the lower end wall of the gas distribution cavity are both sealed, and the side wall of the gas distribution cavity is provided with mesh holes for gas to enter and exit; The pressure equalizing chamber is a hollow annular chamber vertically arranged inside the tower body, the upper end wall and the lower end wall of the pressure equalizing chamber are sealed, the side wall of the pressure equalizing chamber is provided with mesh holes for gas inlet and outlet, and the gas distribution chamber is located in the peripheral area of ​​the pressure equalizing chamber; The porous pressure plate is radially arranged at the upper part of the tower body and is located above the gas distribution chamber and the pressure equalization chamber. The lower end of the gas distribution chamber and the lower end of the pressure equalization chamber are connected to the radially arranged grid plate, and the adsorbent is filled in the area between the grid plate and the porous pressure plate.

2. The lithium hexafluorophosphate adsorption tower according to claim 1, characterized in that: The first air inlet is connected to the lower end cap of the tower body, the exhaust port is connected to the upper end cap of the tower body, and the second air inlet is connected to the side wall of the tower body.

3. The lithium hexafluorophosphate adsorption tower according to claim 1, characterized in that: The height of the pressure equalizing chamber is not less than the height of the gas distribution chamber, and the volume of the pressure equalizing chamber is not less than the volume of the gas distribution chamber; and / or, the pressure equalizing chamber is located in the central area inside the tower body, and the gas distribution chamber is located in the peripheral area of ​​the pressure equalizing chamber.

4. The lithium hexafluorophosphate adsorption tower according to any one of claims 1 to 3, characterized in that: The diameter of the gas distribution cavity is 0.02-0.2 times the diameter of the adsorption tower body.

5. The lithium hexafluorophosphate adsorption tower according to claim 1, characterized in that: The diameter of the pressure equalization chamber is 0.05-0.3 times the diameter of the adsorption tower body; and / or, The height of the pressure equalization chamber is 0.6-1 times the filling height of the adsorbent; and / or, The distance between the centers of the cross sections of the pressure equalizing chamber and the gas distribution chamber on the same plane is 0.1-0.3 times the diameter of the tower body.

6. The lithium hexafluorophosphate adsorption tower according to any one of claims 1 to 5, characterized in that: The upper end wall and the lower end wall of the gas distribution chamber and the pressure equalization chamber are both connected to the annular grid plate to achieve a sealed state.

7. A regeneration process for a lithium hexafluorophosphate adsorption tower, using the adsorption tower according to any one of claims 1 to 6, characterized in that: include: The nitrogen is heated to 80-100°C and enters the adsorption tower from the first air inlet and the second air inlet for purging, and the tail gas is discharged from the exhaust port.

8. A method for optimizing parameters of a lithium hexafluorophosphate adsorption tower according to any one of claims 1 to 6, characterized in that: The steps include: 1) Collecting adsorption tower parameter data; 2) Process the parameter data and remove the deviation data; 3) The processed data is divided into training set, validation set and test set according to the leave-one-out method; 4) The training set data is created based on a neural network to obtain the following training model: Y=f(x1,x2,x3,,,x n ) where x1, x2, x3, x n is the process operation parameter, f is the relationship between the trained neural network model and each process operation parameter, and Y is the final output result; 5) Using the validation set data to verify the obtained training model, and using the test set data to test the obtained training model to obtain a machine learning model; 6) Setting the shortest nitrogen purge regeneration time as the objective function, using genetic algorithm to optimize the design in the machine learning model to obtain the optimized parameters under the specified objective function.

9. The parameter optimization method according to claim 8, characterized in that: The parameters include but are not limited to production capacity, adsorbent bulk density, adsorbent saturated adsorption capacity, regeneration cycle, adsorption tower diameter, adsorption tower height, number of adsorption towers, gas distribution chamber diameter, pressure equalization chamber diameter, pressure equalization chamber height, total regeneration gas volume, total regeneration nitrogen purge time, or a combination thereof.

10. The parameter optimization method according to claim 8, characterized in that: The processing in step 2) includes but is not limited to: integration and denoising, removing duplicate data, processing missing values, unifying data formats, correcting data errors, and processing outliers to ensure the accuracy and consistency of the data.

Images