Prediction method for separation performance of selective permeable membrane

Abstract

The invention discloses a selective permeable membrane separation performance prediction method which comprises the following steps: constructing a membrane separation model which comprises a separation process module, a channel model module and a mass transfer module; acquiring data parameters and a model framework of a membrane separation experiment needing to be simulated, and constructing the model framework of the currently simulated membrane separation experiment in the membrane separation model; acquiring a model equation under the model architecture, introducing a trapezoid rule, correlating the relationship between the flow change of the permeation side and the flow change of the retentate side, and calculating the flow change of the retentate side on the basis of the component permeation flux; finally, all-around simulation results such as the membrane area, the separation coefficient, the terminal seepage flow, the terminal seepage pressure, the terminal retentate flow, the terminal retentate pressure, the concentration distribution, the speed distribution and the pressure distribution at the industrial scale level can be displayed, and all the aspects which cannot be concerned in an experiment are displayed. And the model result is more comprehensive and has reference.

Description

Technical Field

[0001] The present application relates to the field of membrane separation technology, and in particular to a method for predicting the separation performance of a selective permeation membrane. Background Art

[0002] The membrane separation process is the process of separating gas or liquid mixtures through membrane materials. The separation process generally involves three streams: a feed stream, a permeate stream, and a tail stream that is in material balance with the permeate and feed streams. The main components of the membrane separation process are tubular membrane filaments formed by selective membrane materials under certain preparation and processing methods. The structure and size of the membrane filaments will affect the membrane separation process. In the actual separation process, in order to achieve a certain separation effect, one or more membrane filaments can be arranged in parallel in the membrane device. The membrane separation process is widely used in the fields of environmental governance and industrial energy conservation, but there is currently a lack of computational models that can quickly and accurately describe the membrane separation process.

[0003] Existing models lack functionality, performance efficiency, usability, reliability, and portability, failing to meet the requirements of industrial applications. Functionally, existing models exhibit limited applicability and are often limited to calculations of single systems or processes. They fail to meet the accuracy requirements for multi-component systems or complex processes. Performance efficiency also leads to high computational memory usage, long calculation times, and slow response times. User-friendly operation prompts can be difficult for users to understand. Reliability is also unstable, prone to long periods of inaccessibility and system crashes. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a method for predicting the separation performance of selective permeable membranes. This method can accurately predict the separation performance of selective permeable membrane separation devices and obtain the apparent permeability and selectivity of membrane materials, thereby realizing the calibration of the membrane separation process and meeting the needs of membrane separation prediction in industrialization.

[0005] The present embodiment provides a method for predicting separation performance of a selective permeable membrane, comprising the following steps:

[0006] Step S1: constructing a membrane separation model based on the dynamic relationship between feed flow and permeate flow in membrane separation;

[0007] Step S2: obtaining data parameters and a model framework for the membrane separation to be simulated, and constructing the model framework for the membrane separation currently being simulated in the membrane separation model; wherein the input variables of the data parameters for the membrane separation to be simulated include at least feed flow rate, concentrations of each component, permeation rates of each component, structural parameters of the membrane assembly, operating pressure, and temperature;

[0008] Step S3: Simulation data is generated based on the constructed model framework, and a model equation under the model framework is obtained. The relationship between the flow change on the permeate side and the flow change on the retentate side is correlated during the solution of the model equation to quickly obtain the permeate flow rate, retentate flow rate, tail gas, changes in the concentration of each component, and changes in the pressure of each component.

[0009] Furthermore, in step S3), the trapezoidal rule is introduced into the process of solving the model equation to optimize the solution steps:

[0010] Based on the trapezoidal rule, the integral interval in the model equation is divided into several trapezoidal elements to obtain the area under the curve. An operator is constructed in each element in which the decrease in the permeate side flow is equal to the increase in the retentate side flow, and the operator is introduced into the model equation under the model framework to realize the deformation of the model equation. Under the constraints of boundary conditions, the permeate flow rate, the retentate flow rate, the changes in the concentration of each component, and the changes in the pressure of each component are quickly obtained.

[0011] Furthermore, the step of dividing the integral interval in the model equation into a plurality of trapezoidal elements based on the trapezoidal rule to obtain the area under the curve specifically includes:

[0012] Transform the model equation f(x) on the integration interval [a,b] based on the trapezoidal rule:

[0013]

[0014] Transformation of model equations based on n equidistant trapezoidal elements:

[0015] Where x0=a,x n =b.

[0016] Furthermore, taking the pressure change in the length direction of the membrane tube as an example, the equation is:

[0017]

[0018] The present invention discretizes it into n equally spaced subintervals on [0, L]: z0 = 0, z1 = Δz, z2 = 2Δz, ..., z n =L, where Δz = L / n, and the integral equation is transformed into a trapezoidal algebraic equation on n equally spaced subintervals, so that the equation can be solved:

[0019]

[0020] Furthermore, in step S1):

[0021] The membrane separation model includes a separation process module, a channel model module and a mass transfer module;

[0022] The separation process module is constructed based on the flow direction of the feed stream and the permeate stream;

[0023] The channel model module selects a corresponding flow channel based on the direction of the raw material flow;

[0024] The mass transfer module is constructed based on the transfer relationship of substances across the membrane, the change relationship along the length direction of the membrane component, and whether concentration polarization occurs in the membrane separation model.

[0025] Furthermore, the separation process module includes: a countercurrent unit, a parallel flow unit and a cross-flow unit;

[0026] The feed flow and the permeate flow on both sides of the membrane in the countercurrent unit flow in opposite directions;

[0027] The feed stream and the permeate stream on both sides of the membrane in the parallel flow unit have the same flow direction;

[0028] The flow directions of the feed flow and the permeate flow on both sides of the membrane in the cross-flow unit are staggered. In the approximate treatment, it is assumed that the flow directions of the two are the same and there are different concentration boundaries.

[0029] Furthermore, the channel model module includes: a tube-side unit and a shell-side unit;

[0030] In the tube-side unit, the permeable substances in the feed stream permeate from the inside of the membrane fibers to the shell side of the membrane separation device;

[0031] In the shell-side unit, the permeable substances in the feed stream permeate from the outside of the membrane fibers to the inside of the membrane fibers;

[0032] A total mass conservation relationship model, a material conservation relationship model and a pressure relationship model are constructed in the channel model module.

[0033] Furthermore, the mass transfer module includes: an interface transfer module and a convection diffusion module;

[0034] The interface transfer module is based on a model of substance transfer across the membrane and constructs a permeation rate model for each component based on the relationship between feed flow, permeate flow, and retentate flow.

[0035] When the difference in permeation rate of each component does not exceed a preset value, and concentration polarization does not occur in the membrane separation model, the convection-diffusion module constructs a permeation model and boundary conditions for each component based on the relationship between the concentration gradient, pressure gradient, and velocity gradient along the length of the membrane assembly;

[0036] When the difference in permeation rates of the components exceeds a preset value, concentration polarization occurs in the membrane separation model. The convection-diffusion module then introduces a concentration polarization construction model to construct a permeation model of each component based on the concentration gradient, pressure gradient, velocity gradient and concentration polarization phenomenon along the length of the membrane assembly, as well as boundary conditions, the mass transfer driving force of the concentration polarization layer and the molar gas concentration of the components in the concentration polarization layer.

[0037] Furthermore, when concentration polarization occurs in the membrane separation model, the corresponding interface transfer equation is: d(ux)=k(x-xs)+xdu, where

[0038] Where u represents the fluid flow rate on the retentate side of the membrane module, v represents the fluid flow rate on the permeate side of the membrane module, x represents the molar concentration on the retentate side of the membrane module, y represents the molar concentration on the permeate side of the membrane module, and p i represents the pressure inside the wire, J1 is the permeability coefficient, J2 is the permeability coefficient of the difficult-to-permeate component, and k is the mass transfer coefficient;

[0039] The trapezoidal rule is used to optimize the solution of the interface transfer equation, including:

[0040] Step 1: Discretize the integration interval [0, L] into n subintervals: z0 = 0, z1 = Δz, z2 = 2Δz, ..., z n =L, where Δz = L / n.

[0041] Step 2: Create an alternative equation for the integral equation based on the trapezoidal rule:

[0042]

[0043] Step 3: Apply the substitution equation to solve the interface transfer equation.

[0044]

[0045] definition:

[0046] Establish an alternative equation for solving the interface transfer equation based on the trapezoidal rule:

[0047]

[0048] Step 4: Calculate f(z i ): Calculate each x(z i ) and x s (z i ), and then we get f(z i ).

[0049] Step 5: Iterative solution. According to each f(z i) value, iterate and update the distance z i The permeation flux ux(z i ).

[0050] Furthermore, it is assumed that the membrane tube calculation interval of length [0, L] is divided into 4 subintervals to solve the interface transfer equation.

[0051] Step 1: z0=0, z1=L / 4, z2=L / 2, z3=3L / 4, z4=L.

[0052] Step 2: Calculate x(z0) and x s (z0) value.

[0053] Step 3: According to Calculate the value of f(z0) and update the value of ux(z1).

[0054] Step 4: Based on the updated value of ux(z1), iteratively calculate f(z1). Similarly, we can obtain f(z2), f(z3), and f(z4).

[0055] Step 5: The final interface transfer flux is:

[0056] Actual interface saturation concentration x s The expression is very complicated:

[0057] The integral form of the interface transfer equation is:

[0058]

[0059] Using the solution method from step 1 to step 5 can increase the speed of solving the equation.

[0060] Furthermore, the selective permeability membrane separation performance prediction method further includes:

[0061] Step S4: Obtain the previous prediction results and the corresponding real data, and perform polynomial fitting on the errors between the two to obtain the error correction value. Train the output neural network model, where All are matrices, and the polynomial fitting is adjusted to obtain the final prediction correction model.

[0062] Furthermore, the change in the length direction of the membrane separation model is calculated by the following equation:

[0063] The calculation equation of the retentate flow rate u is: u(i+1)=u(i)-du, where u(i) is the flow rate at the i-th node on the retentate side;

[0064] The calculation equation of permeation flow v is: v(i+1)=v(i)+dv, where v(i) is the flow rate at the i-th node on the permeation side;

[0065] The calculation equation for the retentate concentration x is: x(i+1)=x(i)-dx, where x(i) is the concentration of the component at the i-th node on the retentate side;

[0066] The calculation equation of the permeate concentration y is: y(i+1)=y(i)+dy, where y(i) is the concentration of the component at the i-th node on the permeate side.

[0067] Furthermore, the membrane separation model has no less than 2 components;

[0068] The change of any component j is calculated by the following model:

[0069] Retentate flow u j Calculation equation: u j (i+1)=u j (i) -du;

[0070] Permeate flow v j Calculation equation: v j (i+1)=v j (i)+dv;

[0071] Retentate concentration x j Calculation equation: x j (i+1)=x j (i)-dx;

[0072] Osmotic concentration j Calculation equation: y j (i+1)=y j (i)+dy;

[0073] The total pressure is the sum of the pressures of the components: m is the number of components in the system;

[0074] The total retentate flow rate u is the sum of the retentate flow rates of each component:

[0075] The total retentate flow rate v is the sum of the retentate flow rates of each component:

[0076] The sum of the retentate concentrations calculated as percentages is 1:

[0077] The sum of the osmotic concentrations calculated as percentages is 1:

[0078] Beneficial effects of the present invention:

[0079] 1) The membrane separation model constructed in this application simulates industrial membrane separation processes (including pilot and industrial scales). Other existing models are mainly oriented towards chemical processes. The interfacial permeability performance of the two models is very different. The prediction scales of other models are usually at the centimeter, millimeter level or below, while the prediction scale of this model is at the meter level or above.

[0080] 2) The model in this application is based on the component permeation flux, and can ultimately form a comprehensive display of simulation results such as industrial-scale membrane area, separation coefficient, terminal permeation flow rate, terminal permeation pressure, terminal retentate flow rate, terminal retentate pressure, concentration distribution, velocity distribution, and pressure distribution, presenting all aspects that cannot be paid attention to simultaneously in the experiment. Compared with other models, the model results are more comprehensive and have reference value. At the same time, facing the membrane separation process in various application scenarios, the relative deviation of the data obtained by this scheme can be less than 20%, and the data reliability is high;

[0081] 3) The simulation time of each membrane separation process in this application can be maintained at about 20 milliseconds, achieving efficient and fast calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application, and other drawings obtained on this basis should also be attributed to this embodiment.

[0083] Figure 1 is a flow chart of the method for predicting separation performance of a selective permeable membrane of the present application;

[0084] Figure 2 It is a framework diagram of the membrane separation model in this application;

[0085] Figure 3 It is a schematic flow diagram of the countercurrent unit in this application;

[0086] Figure 4 It is a schematic flow diagram of the parallel flow unit in this application;

[0087] Figure 5 It is a schematic diagram of the cross-flow unit in this application. DETAILED DESCRIPTION

[0088] In order to make the purpose, features, and advantages of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described below are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0089] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0090] In the description of this application, it should be understood that the terms "upper", "lower", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing this application 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 understood as a limitation on this application.

[0091] The present application is described below with reference to specific embodiments:

[0092] The model of the present invention is aimed at industrial process objects, that is, industrial membrane separation processes (including pilot and industrial scales). Under this solution, the scale of membrane separation is usually at the meter level or above.

[0093] In order to cope with the multi-faceted calculation of experimental data in this model environment, this application proposes a method for predicting the separation performance of a large-scale membrane separation device, comprising the following steps:

[0094] Step S1: Construct a membrane separation model based on the dynamic relationship between the feed flow and the permeate flow in membrane separation, as a model constructed for the industrial process object, such as Figure 2 As shown in the figure, this model is divided into three levels, including separation process module, channel model module and mass transfer module.

[0095] The separation process module involves the selection of membrane separation process. Membrane separation process generally has three modes: countercurrent, parallel flow and cross flow. All three modes will appear in engineering practice. When building the model, the three flow modes of countercurrent, parallel flow and cross flow are provided for selection.

[0096] like Figure 2 As shown, as a specific embodiment, the separation process module includes: a countercurrent unit, a parallel flow unit and a cross-flow unit. In the countercurrent unit, the flow directions of the feed flow and the permeate flow on both sides of the membrane are opposite, in the parallel flow unit, the flow directions of the feed flow and the permeate flow on both sides of the membrane are the same, and in the cross-flow unit, the flow directions of the feed flow and the permeate flow on both sides of the membrane are staggered. In approximate processing, it is assumed that the flow directions of the two are the same and there are differences in concentration boundaries.

[0097] As a specific example, a hollow fiber membrane module is used as a membrane separation device in industry for description:

[0098] like Figure 3 The flow diagram of the countercurrent unit is shown, in which the feed flow and permeate flow on both sides of the hollow fiber membrane are in opposite directions.

[0099] like Figure 4 The flow diagram of the parallel flow unit is shown, in which the feed flow and permeate flow on both sides of the hollow fiber membrane are in the same direction.

[0100] like Figure 5 The flow diagram of the cross-flow unit is shown, in which the flow directions of the feed flow and the permeate flow on both sides of the hollow fiber membrane are assumed to be the same and there are different concentration boundaries.

[0101] In the above schematic diagram, u represents the fluid flow rate on the retentate side of the membrane module, v represents the fluid flow rate on the permeate side of the membrane module, x represents the molar concentration on the retentate side of the membrane module, and y represents the molar concentration on the permeate side of the membrane module.

[0102] The channel model module is designed to optimize the flow path selection. Feedstock can be routed through either tube or shell. Tube-pass allows permeable substances in the feedstock to permeate from the inside of the membrane fibers to the shell side of the membrane separation device. Shell-pass, on the other hand, allows permeable substances in the feedstock to permeate from the outside of the membrane fibers to the inside. From a material transport perspective, the two flow paths are essentially the same, differing only in the pressure drop between tube-pass and shell-pass.

[0103] like Figure 2 As shown, as a specific embodiment, the channel model module includes: a tube-side unit and a shell-side unit.

[0104] In the tube-side unit, the easily permeable substances in the feed flow permeate from the inside of the membrane fibers to the shell side of the membrane separation device.

[0105] In the shell-side unit, the permeable substances in the feed flow permeate from the outside of the membrane fibers to the inside of the membrane fibers.

[0106] In the channel model module, tube-side units or shell-side units are selected according to the actual membrane separation situation to construct the total mass conservation relationship model, material conservation relationship model and pressure relationship model.

[0107] As a specific example, taking the raw material flow as an example with two components, the conservation model equation constructed under the pipe-side unit of the channel model module is as follows:

[0108] Total mass conservation equation: du = dv;

[0109] Material conservation equation for component 1: udx+xdu=vdy+ydv;

[0110] Material conservation equation for component 2: -udx+(1-x)dv=-vdy+(1-y)dv;

[0111] Construct the pressure equation along the radial z direction:

[0112] Where μ represents the viscosity of the mixture, p i Indicates the pressure inside the wire, d i represents the inner diameter of the hollow fiber membrane, R is the gas constant, T is the temperature, and N is the number of membrane fibers in the membrane separation device.

[0113] As a specific example, the raw material flow takes two components as an example. Under the shell-side unit of the channel model module, the conservation model equation constructed is as follows:

[0114] Construct the conservation model equations:

[0115] Total mass conservation equation: du = dv;

[0116] Material conservation equation for component 1: udx+xdu=vdy+ydv;

[0117] Material conservation equation for component 2: -udx+(1-x)dv=-vdy+(1-y)dv;

[0118] Construct the pressure equation along the radial z direction:

[0119] Where S represents the average fluid flow rate on the shell side, p is the pressure, D represents the equivalent diameter of the membrane wire, D0 is the outer diameter of the membrane wire, and N is the number of membrane wires in the membrane separation device.

[0120] The mass transfer module is constructed based on the transfer relationship of substances on both sides of the membrane, the change relationship along the length of the membrane component, and whether concentration polarization occurs in the membrane separation model.

[0121] Specifically, the mass transfer module includes an interface transfer module and a convection diffusion module, wherein the interface transfer module is a model of material transfer across the two sides of the membrane. Under normal circumstances (the difference in component permeation rates can be limited to no more than a preset value, which is selected based on empirical values), the convection diffusion model includes models of concentration gradient, pressure gradient, and velocity gradient along the length of the membrane assembly. When the difference in component permeation rates exceeds the preset value, the effect of concentration polarization on the interface concentration distribution needs to be considered in the convection diffusion module. The concentration polarization effect here is a special transfer phenomenon caused by the surface properties of the membrane material, that is, when the concentration difference between the easily permeable component and the poorly permeable component on the membrane surface is extremely large (<10%, this value is used as the above empirical value), the poorly permeable component reacts on the easily permeable component, resulting in a decrease in the mass transfer capacity of the easily permeable component.

[0122] As a specific embodiment, the state equation is constructed in the interface transfer module:

[0123] Among them, u in Indicates the feed flow rate, v out represents the permeate flow rate, u out represents the retentate flow rate, x out represents the retentate composition, y out represents the permeate composition, J A represents the permeability coefficient of component 1, J B represents the permeability coefficient of component 2, Q A is the permeation rate of component 1, Q B is the permeation rate of component 2, x in represents the feed concentration, p1 represents the retentate side pressure, and p2 represents the permeate side pressure.

[0124] As a specific embodiment, when the difference in permeation rate of each component does not exceed a preset value, concentration polarization does not occur in the membrane separation model. Under this condition, the convection-diffusion module constructs a permeation model and boundary conditions for each component based on the relationship between the concentration gradient, pressure gradient and velocity gradient in the length direction of the membrane assembly.

[0125] As a specific embodiment, the state equation is constructed in the convection-diffusion model:

[0126] Permeation equation for component 1: J A (p i xp o y)dz / L=-udx-xdu;

[0127] Permeation equation for component 2: J A [p i (1-x)-p o (1-y)]dz / (Lα)=udx-(1-x)du;

[0128] Among them, the permeability coefficient J changes with temperature and conforms to the Arrhenius relationship;

[0129] Permeability coefficient:

[0130] Separation coefficient α = J1 / J2;

[0131] Viscosity of the mixture:

[0132] Boundary conditions:

[0133] At z=0, x=x0, u=u0, p i =p i,0 ,

[0134] z=L,v=0,p o =p o ,

[0135] p o Indicates the external pressure of the wire, p i represents the pressure inside the wire, L represents the wire length, E1 represents the activation energy, C1 represents the constant, J2 is the permeability coefficient of the difficult-to-permeate component, x0 is the feed concentration, u0 is the feed flow rate, p i,0 is the feed pressure, x L is the concentration of the retentate side fluid at the position where the membrane tube length is L, y L is the fluid concentration on the permeate side at the position where the membrane tube length is L;

[0136] When the difference in permeation rates of each component exceeds a preset value, concentration polarization occurs in the membrane separation model. Under this condition, the convection-diffusion module introduces a concentration polarization construction model to construct a permeation model of each component based on the concentration gradient, pressure gradient, velocity gradient and concentration polarization phenomenon along the length of the membrane assembly, as well as boundary conditions, mass transfer driving force of the concentration polarization layer and molar gas concentration of the components in the concentration polarization layer.

[0137] As a specific embodiment, when the difference in component permeation rate exceeds a preset value, a concentration polarization model is introduced into the convection-diffusion model to construct the equation of state:

[0138] Permeation equation for component 1: J A (p i x s -p o y)dz / L=-udx-xdu;

[0139] Permeation equation for component 2: J A [p i (1-x s )-p o(1-y)]dz / (Lα)=udx-(1-x)du;

[0140] Concentration polarization layer mass transfer driving force: d(ux) = k(x-xs) + xdu;

[0141] Mass transfer coefficient is the kinematic viscosity, d h is the equivalent diameter of the hollow fiber membrane; the molar gas concentration of component 1 in the concentration polarization layer

[0142] Boundary conditions when considering concentration polarization:

[0143] When z=0, x=x0,u=u0,p i =p i,0 , p o =p o

[0144] When z=L, v=0, p o =p o ,

[0145]

[0146] As a specific embodiment, under the condition of membrane separation, this model is selected according to the condition: the separation process module, the channel model module and the mass transfer module are selected in sequence, so that the model corresponding to the final simulated experiment can be established and constructed.

[0147] Among them, the separation process module is selected as the highest priority, that is, the separation process needs to be determined first, then the logistics channel is determined in the channel model module, and finally the mass transfer model is determined.

[0148] Step S2: Obtain the data parameters and model architecture of the membrane separation to be simulated, and construct the model architecture of the membrane separation currently being simulated in the membrane separation model; wherein the input variables of the data parameters of the membrane separation experiment to be simulated include at least the feed flow rate, the concentration of each component, the permeation rate of each component, the structural parameters of the membrane assembly, the operating pressure and the temperature.

[0149] Step S3: Generate simulation data based on the constructed model framework, obtain the model equation under the model framework and output the permeate flow and tail flow rate, the mole percentage of each component, and the pressure loss.

[0150] In order to improve the calculation speed under this model, the trapezoidal rule is introduced in this application, and numerical discretization is performed in the convection-diffusion module and the convection-diffusion module with concentration polarization introduced.

[0151] As a specific embodiment, the model architecture constructed in the concentration-free polarization process includes the following eight differential equations:

[0152] (a1) Total mass conservation equation: du = dv;

[0153] (a2) Material conservation equation for component 1: udx + xdu = vdy + ydv;

[0154] (a3) Material conservation equation for component 2: -udx+(1-x)dv=-vdy+(1-y)dv;

[0155] (a4) Permeation equation for component 1: J A (p i xp o y)dz / L=-udx-xdu;

[0156] (a5) Permeation equation for component 2: J A [p i (1-x)-p o (1-y)]dz / (Lα)=udx-(1-x)du;

[0157] (a6) Pressure equation:

[0158] (a7) Viscosity of the mixture: μ i represents the viscosity of component i;

[0159] (a8) The boundary conditions are set as:

[0160] At Z=0, x=x0, u=u0, p i =p i,0 , p o =p o

[0161] At Z=L, v=0, p o =p o ,

[0162] In the above series of equations, according to the above method, the trapezoidal rule is introduced here, and the specific optimization steps are:

[0163] Step A1: The flow rate change on the permeate side is related to the flow rate change on the retentate side by introducing the equation du=dv, that is, the decrease in the permeate side flow rate is equal to the increase in the retentate side flow rate.

[0164] Step A2: Substitute equation (a1) into equations (a2) and (a3) ​​to obtain the updated mass balance equation:

[0165] udx+xdv=vdy+ydv……(a2)

[0166] -udx+(1-x)dv=-vdy+(1-y)dv……(a3)

[0167] Step A3: Solve by updating the mass balance equation

[0168] Step A4: Substitute the dv expression into the permeation equation (a4) to obtain the updated permeation rate equation:

[0169] Further simplification yields:

[0170] Step A5: Rearrange equation (a5) to obtain the expression for dx:

[0171] Further simplification yields:

[0172]

[0173] Step A6: Solve the pressure drop equation by combining equations (a6) and (a7).

[0174] Step A7: Substitute the boundary conditions into equation (a8) to obtain the solution for each output variable.

[0175] As a specific embodiment, for the concentration polarization model, some equations and boundary conditions are different from those of the non-concentration polarization model.

[0176] Equation (b1): du = dv

[0177] Equation (b2): udx + xdu = vdy + ydv

[0178] Equation (b3): ​​-udx + (1-x) dv = -vdy + (1-y) dv

[0179] Equation (b4): J A (p i x s -p o y)dz / L=-udx-xdu

[0180] Equation (b5): J A [p i (1-x s )-p o (1-y)]dz / (Lα)=udx-(1-x)du

[0181] Equation (b6): d(ux) = k(xxs )+xdu

[0182] Equation (b7):

[0183] Equation (b8):

[0184] Equation (b9):

[0185] Boundary conditions:

[0186] At Z=0, x=x0, u=u0, p i =p i,0 , p o =p o

[0187] At Z=L, v=0, p o =p o ,

[0188] In the above series of equations, according to the above method, the trapezoidal rule is introduced here, and the specific optimization steps are:

[0189] Step B1: Solve equation (b8):

[0190] Integrate with z as the integration interval:

[0191] Consider μ, u, N, d i is a constant:

[0192] Substituting the pressure boundary condition, p i (0) = p i，0 , p i (L) = p i，L :

[0193]

[0194] therefore, Here i,0 is the feed pressure at the position where the membrane tube length is 0, p i，L is the feed pressure at the position where the membrane tube length is L;

[0195] Step B2: Solve equation (b6): d(ux) = k(xx s )+xdu

[0196] Substituting equation (b8)x s Substituting the expression into the equation and integrating equation (b6) with z as the integration interval, we obtain:

[0197] Considering the complexity of the equation, it is difficult to solve it by direct integration. A similar solution is performed according to steps A1 to A5 in the summary of the invention.

[0198] Step B3: Substitute the boundary conditions and solve equations (b1) to (b5) one by one:

[0199] For the differential equation du=dv, we have u=v+C1

[0200] For equation (b2), we have

[0201] Solving equation (b3) yields:

[0202] For equation (b4), we have

[0203] For equation (b5), we have

[0204] Step B4: Substitute the boundary conditions into the above equations (b1) to (b5).

[0205] Substituting the boundary conditions at z = 0 and z = L into the above equations (b1) to (b5), we can obtain x(z), y(z), p i (z), p o The analytical solutions of (z), u(z), and v(z), namely, the concentration distribution, pressure distribution, and velocity distribution in the membrane channel direction.

[0206] Since the above equations (b1) to (b5) are relatively complex, their rates of change can be set as constants when solving them, or they can be solved using numerical methods.

[0207] As a specific embodiment, the present application may introduce a numerical solution method for the above equation:

[0208] The present invention uses the trapezoidal algorithm to solve the integral equation:

[0209]

[0210] By dividing the integral interval into n subintervals, the original integral function is infinitely approximated:

[0211]

[0212] Where x0=a,x n =b.

[0213] Taking equation (b8) as an example, the present invention illustrates how to implement numerical integration using the trapezoidal algorithm:

[0214] Discretize it into n subintervals on [0, L]: z0 = 0, z1 = Δz, z2 = 2Δz, ..., z n =L, where Δz = L / n. Then equation (b8) can be transformed into the following form, and then solved:

[0215]

[0216] For equation (11): d(ux) = k(xx s )+xdu,

[0217] in

[0218] Numerical integration of equation (b6) using the trapezoidal algorithm involves the following steps:

[0219] Step C1: Discretize the integral interval [0, L] into n subintervals: z0 = 0, z1 = Δz, z2 = 2Δz, ..., z n =L, where Δz = L / n.

[0220] Step C2: Establish the discrete algebraic equation to estimate the integral function:

[0221]

[0222] Step C3: Apply the discrete algebraic equation to equation (b6).

[0223]

[0224] definition:

[0225] Apply the trapezoidal algorithm to the function f(z):

[0226]

[0227] Step C4: Discretize x(z) and x s (z): At each z i Calculate x(z i ) and x s (z i ).

[0228] Step C5: Iterative solution. According to each x(z i ) and x s (z i ) value, iterate and update ux(z i ).

[0229] As a specific example:

[0230] Assume that the interval [0, L] is divided into 4 subintervals and solve equation (b6).

[0231] Step C1: z0=0, z1=L / 4, z2=L / 2, z3=3L / 4, z4=L.

[0232] Step C2: Calculate x(z i ) and x s (z i )value.

[0233] Step C3: According to Calculate f(z i ) value.

[0234] Step C4: Perform equivalent substitution on the integral function f(x) in the interval [0, L].

[0235]

[0236] Step C5: At each z i Iteratively update ux(z i ).

[0237] In addition, in view of the fact that nonlinear iteration may cause accumulation of calculation errors in the process of adopting discrete calculation scheme of this model, an error module is constructed to fit the calculation results of this model with the actual true values ​​through a polynomial equation. On the basis of existing experimental data, a polynomial correlation is obtained by mapping and regressing the calculation results of the model with the actual true values, and the output results of each time are further corrected by this polynomial correlation to achieve higher precision training.

[0238] As a specific embodiment, the previous prediction results and the corresponding real data can be obtained, and the errors between the two can be fitted with a polynomial to obtain the error correction value. Train the output neural network model, where All are matrices, and the polynomial fit is adjusted to obtain the final prediction correction model.

[0239] The accuracy of the prediction method of this application is explained below in combination with experimental data:

[0240] Scenario 1: At an operating temperature of 35°C, a 100mm long, 12mm outer diameter, and 8mm inner diameter tubular membrane was used to separate a mixture of carbon dioxide (CO2) and nitrogen (N2). The tubular membrane assembly had an inner diameter of 20mm, and a single ceramic membrane was installed within the membrane assembly. A single tubular membrane was installed within the membrane assembly. The mixed gas passed through the shell side (inner diameter 20mm), and the separated CO2 was collected from the membrane tube. The gas separation process used a countercurrent flow. The main focus was on the effects of different feed pressures on separation performance. The specific experimental data and calculation results are shown in Tables 1-1 to 1-5 below:

[0241]

[0242] Table 1-1 Selectivity data comparison

[0243]

[0244] Table 1-2 Comparison of retentate flow rate data

[0245]

[0246] Table 1-3 Comparison of retentate concentration data

[0247]

[0248] Table 1-4 Comparison of permeation flow data

[0249]

[0250] Table 1-5 Comparison of osmotic concentration data

[0251]

[0252] Scenario 2: Under operating temperature of 35°C and feed pressure of 300kPa, a tubular membrane with a length of 100mm, an outer diameter of 12mm, and an inner diameter of 8mm was used to separate a mixture of carbon dioxide (CO2) and nitrogen (N2). The inner diameter of the tubular membrane module was 20mm, and a single tubular membrane was installed inside the membrane module. The mixed gas flowed through the shell side, and the separated CO2 was collected from the membrane tube. The gas separation process adopted a countercurrent flow process. The main purpose was to investigate the effect of different feed concentrations (compositions) on separation performance. The specific experimental data and the corresponding calculation results of the tested software are shown in Tables 2-1 to 2-5 below:

[0253]

[0254] Table 2-1 Selectivity data comparison

[0255]

[0256] Table 2-2 Comparison of retentate flow rate data

[0257]

[0258] Table 2-3 Comparison of retentate concentration data

[0259]

[0260] Table 2-4 Comparison of permeation flow data

[0261]

[0262] Table 2-5 Comparison of osmotic concentration data

[0263]

[0264] Scenario 3: Under a feed pressure of 300kPa, a 100mm long, 12mm outer diameter, and 8mm inner diameter tubular membrane was used to separate a mixture of carbon dioxide (CO2) and nitrogen (N2). The tubular membrane module had an inner diameter of 20mm and contained a single tubular membrane. The mixed gas flowed through the shell side, and the separated CO2 was collected from the membrane tube. The gas separation process employed a countercurrent flow. The primary objective was to examine the effects of varying operating temperatures on separation performance. The specific experimental data and the calculated results of the tested software are shown in Tables 3-1 to 3-5 below:

[0265]

[0266] Table 3-1 Selectivity data comparison

[0267]

[0268] Table 3-2 Comparison of retentate flow rate data

[0269]

[0270] Table 3-3 Comparison of retentate concentration data

[0271]

[0272] Table 3-4 Comparison of permeation flow data

[0273]

[0274] Table 3-5 Comparison of osmotic concentration data

[0275]

[0276] Scenario 4: At an operating temperature of 25°C, a zinc chloride (ZnCl2) solution filtration experiment was conducted using a hollow fiber membrane with a length of 100 mm, an outer diameter of 1 mm, and an inner diameter of 0.8 mm. The tubular membrane module had an inner diameter of 20 mm and contained 10 hollow fiber membranes. The feed liquid passed through the shell side, and the permeate was collected from the fiber membrane tubes. The separation process used a countercurrent flow. The main purpose was to investigate the effects of different feed flow rates and feed concentrations (compositions) on membrane separation performance. The specific experimental data and the calculation results of the tested software are shown in Tables 4-1 to 4-6 below:

[0277]

[0278] Table 4-1 Comparison of interception rate data

[0279]

[0280] Table 4-2 Comparison of retentate flow rate data

[0281]

[0282]

[0283] Table 4-3 Comparison of retentate concentration data

[0284]

[0285] Table 4-4 Comparison of permeation flow data

[0286]

[0287] Table 4-5 Comparison of osmotic concentration data

[0288]

[0289] Table 4-6 Shell side pressure drop data comparison

[0290]

[0291] Scenario 5: At an operating temperature of 25°C, a sodium sulfate (Na2SO4) solution filtration experiment was conducted using a hollow fiber membrane with a length of 100 mm, an outer diameter of 1 mm, and an inner diameter of 0.8 mm. The tubular membrane module had an inner diameter of 20 mm and contained 10 hollow fiber membranes. The feed liquid passed through the shell side, and the permeate was collected from the fiber membrane tubes. The separation process used a countercurrent flow. The main purpose was to investigate the effects of different feed pressures and feed concentrations (compositions) on membrane separation performance. The specific experimental data and the calculation results of the tested software are shown in Tables 5-1 to 5-6 below:

[0292]

[0293] Table 5-1 Comparison of interception rate data

[0294]

[0295] Table 5-2 Comparison of retentate flow rate data

[0296]

[0297] Table 5-3 Comparison of retentate concentration data

[0298]

[0299] Table 5-4 Comparison of permeation flow data

[0300]

[0301] Table 5-5 Comparison of osmotic concentration data

[0302]

[0303]

[0304] Table 5-6 Shell side pressure drop data comparison

[0305]

[0306] Scenario 6: A sodium sulfate (Na2SO4) solution filtration experiment was conducted using a 100mm long, 1mm outer diameter, and 0.8mm inner diameter hollow fiber membrane at a transmembrane pressure of 500kPa. The tubular membrane module had an inner diameter of 20mm and contained 10 hollow fiber membranes. The feed solution flowed through the shell, and the permeate was collected from the fiber membrane tubes. The separation process employed a countercurrent flow. The primary objective was to examine the effects of varying operating temperatures on membrane separation performance. The specific experimental data and the calculated results of the tested software are shown in Tables 6-1 to 6-6 below:

[0307]

[0308] Table 6-1 Comparison of interception rate data

[0309]

[0310] Table 6-2 Comparison of retentate flow rate data

[0311]

[0312]

[0313] Table 6-3 Comparison of retentate concentration data

[0314]

[0315] Table 6-4 Comparison of permeation flow data

[0316]

[0317] Table 6-5 Comparison of osmotic concentration data

[0318]

[0319] Table 6-6 Shell side pressure drop data comparison

[0320]

[0321] Scenario 7: Under operating conditions of 35°C and a feed pressure of 400 kPa, a 100 mm long, 12 mm outer diameter, and 8 mm inner diameter tubular membrane was used to separate a helium-methane mixture. The tubular membrane module had an inner diameter of 20 mm and contained a single ceramic membrane. The mixed gas passed through the shell side, and the separated helium was collected from the membrane tube. The gas separation process employed a countercurrent flow.

[0322] The main focus is on the effects of different feed flow rates and feed concentrations (compositions) on separation performance. Specific experimental data and the calculation results of this application are shown in Tables 7-1 to 7-5 below:

[0323]

[0324]

[0325] Table 7-1 Selectivity data comparison

[0326]

[0327] Table 7-2 Comparison of retentate flow rate data

[0328]

[0329] Table 7-3 Comparison of retentate concentration data

[0330]

[0331] Table 7-4 Comparison of permeation flow data

[0332]

[0333] Table 7-5 Comparison of osmotic concentration data

[0334]

[0335] Scenario 8: Under a feed pressure of 300kPa, a 100mm long, 12mm outer diameter, and 8mm inner diameter tubular membrane was used to separate a mixture of helium (He) and methane (CH4). The tubular membrane assembly had an inner diameter of 20mm and contained a single ceramic membrane. The mixed gas passed through the shell side, and the separated helium (He) was collected from the membrane tube. The gas separation process employed a countercurrent flow. The primary objective was to investigate the effects of varying operating temperatures on separation performance. Specific experimental data and the calculation results for this application are shown in Tables 8-1 to 8-5 below:

[0336]

[0337] Table 8-1 Selectivity data comparison

[0338]

[0339] Table 8-2 Comparison of retentate flow rate data

[0340]

[0341]

[0342] Table 8-3 Comparison of retentate concentration data

[0343]

[0344] Table 8-4 Comparison of permeation flow data

[0345]

[0346] Table 8-5 Comparison of osmotic concentration data

[0347]

[0348] Scenario 9: At an operating temperature of 25°C, a tubular membrane with a length of 100 mm, an outer diameter of 12 mm, and an inner diameter of 8 mm was used to conduct a retention experiment for dextran solutions of varying molecular weights. The tubular membrane assembly had an inner diameter of 20 mm and contained a single ceramic membrane. The filtered liquid was collected from the membrane tube, and cross-flow filtration was used for liquid separation. The experiment primarily examined the effects of varying feed pressures and glucose molecular weight on retention performance. The specific experimental data and the calculation results of this application are shown in Tables 9-1 to 9-6 below:

[0349]

[0350]

[0351] Table 9-1 Comparison of interception rate data

[0352]

[0353] Table 9-2 Comparison of retentate flow rate data

[0354]

[0355] Table 9-3 Comparison of retentate concentration data

[0356]

[0357] Table 9-4 Comparison of permeation flow data

[0358]

[0359]

[0360] Table 9-5 Comparison of osmotic concentration data

[0361]

[0362] Table 9-6 Shell side pressure drop data comparison

[0363]

[0364] Based on the above nine experiments we can see:

[0365] Under the simulation of the membrane separation experimental environmental parameters of the first six different scenarios, the membrane separation process data of carbon dioxide and nitrogen mixtures and sodium sulfate and aqueous solution were calculated. The six scenarios are membrane separation experiments of carbon dioxide and nitrogen mixtures at different feed pressures, membrane separation experiments of carbon dioxide and nitrogen mixtures at different feed concentrations (compositions), membrane separation experiments of carbon dioxide and nitrogen mixtures at different operating temperatures, membrane separation experiments of zinc chloride and aqueous solution at different feed flow rates and feed concentrations (compositions), membrane separation experiments of sodium sulfate and aqueous solution at different feed pressures and feed concentrations (compositions), and membrane separation experiments of sodium sulfate and aqueous solution at different operating temperatures. The calculated data were compared with the experimental data under the same conditions. The relative deviations of the selectivity, permeate flow rate, permeate concentration, retentate flow rate, retentate concentration, rejection rate, and shell-side pressure drop data in the six scenarios were all less than 20%, and the two data were consistent.

[0366] Under the simulation of the environmental parameters of the membrane separation experiments in the latter three different scenarios, the membrane separation process data for a helium-methane mixture and a dextran-water solution were calculated. The three membrane separation experiments included a helium-methane mixture at different feed flow rates and feed concentrations (compositions), a helium-methane mixture at different operating temperatures, and a dextran-water solution at different feed pressures and glucose molecular weights. The calculated data were compared with experimental data under the same conditions. The relative deviations of the selectivity, permeate flow rate, permeate concentration, retentate flow rate, retentate concentration, rejection rate, and shell-side pressure drop data for the three experimental scenarios were all less than 20%, indicating a good agreement between the two data.

[0367] After compiling the solution of the present invention into software, according to actual measurements, when calculating the membrane separation process data of the oxygen and nitrogen mixture, the calculation times for three executions were 19.56 milliseconds, 18.72 milliseconds, and 17.83 milliseconds, respectively, with an average calculation time of 18.70 milliseconds.

[0368] When calculating the membrane separation process data of nitrogen and hydrogen mixture, the calculation time for three executions was 21.36 milliseconds, 20.52 milliseconds, and 19.61 milliseconds, respectively, with an average calculation time of 20.50 milliseconds.

[0369] Basically, it is possible to compress each simulation calculation process to about 20 milliseconds, and the error in the final data parameters compared with the experimental data does not exceed 20%, which makes the reliability very high.

[0370] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0371] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations (such as quantity, shape, position, etc.) can be made to the technical solution of the present invention, and these equivalent transformations are all protected by the present invention.

Claims

1. A method for predicting separation performance of a selective permeable membrane, characterized in that: The steps include: Step S1: Constructing a membrane separation model based on the dynamic relationship between the feed flow and the permeate flow in membrane separation; wherein the membrane separation model includes a separation process module, a channel model module, and a mass transfer module; the separation process module is constructed based on the flow directions of the feed flow and the permeate flow; the channel model module is constructed by selecting a corresponding flow channel based on the direction of the feed flow; and the mass transfer module is constructed based on the transfer relationship of gases across the membrane, the change relationship along the length direction of the membrane module, and whether concentration polarization occurs in the membrane separation model; Step S2: obtaining data parameters and a model framework for the membrane separation to be simulated, and constructing the model framework for the membrane separation currently being simulated in the membrane separation model; wherein the input variables of the data parameters for the membrane separation to be simulated include at least feed flow rate, concentrations of each component, permeation rates of each component, structural parameters of the membrane assembly, operating pressure, and temperature; Step S3: Simulation data is generated based on the constructed model framework, and a model equation under the model framework is obtained. The model equation is associated with the relationship between the flow change on the permeate side and the flow change on the retentate side to quickly obtain the permeate flow rate, the retentate flow rate, the change in the concentration of each component, and the change in the pressure of each component.

2. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: In step S3), the relationship between the flow change on the permeate side and the flow change on the retentate side is associated with the model equation to quickly obtain the permeate flow rate, the retentate flow rate, the change in the concentration of each component and the change in the pressure of each component, specifically including: Based on the trapezoidal rule, the integral interval in the model equation is divided into several trapezoidal elements to obtain the area under the curve. An operator is constructed in each element in which the decrease in the permeate side flow is equal to the increase in the retentate side flow, and the operator is introduced into the model equation under the model framework to realize the deformation of the model equation. Under the constraints of boundary conditions, the permeate flow rate, the retentate flow rate, the changes in the concentration of each component, and the changes in the pressure of each component are quickly obtained.

3. The method for predicting separation performance of a selective permeable membrane according to claim 2, wherein: The step of dividing the integral interval in the model equation into a plurality of trapezoidal elements based on the trapezoidal rule to obtain the area under the curve specifically includes: Transform the model equation f(x) on the integration interval [a,b] based on the trapezoidal rule: Transformation of model equations based on n equidistant trapezoidal elements: Where x0=a,x n =b.

4. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The separation process module includes: a countercurrent unit, a parallel flow unit and a cross-flow unit; The feed flow and the permeate flow on both sides of the membrane in the countercurrent unit flow in opposite directions; The feed stream and the permeate stream on both sides of the membrane in the parallel flow unit have the same flow direction; The flow directions of the feed flow and the permeate flow on both sides of the membrane in the cross-flow unit are staggered. In the approximate treatment, it is assumed that the flow directions of the two are the same and there are different concentration boundaries.

5. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The channel model module includes: a tube-side unit and a shell-side unit; In the tube-side unit, the permeable substances in the feed stream permeate from the inside of the membrane fibers to the shell side of the membrane separation device; In the shell-side unit, the permeable substances in the feed stream permeate from the outside of the membrane fibers to the inside of the membrane fibers; A total mass conservation relationship model, a material conservation relationship model and a pressure relationship model are constructed in the channel model module.

6. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The mass transfer module includes: an interface transfer module and a convection diffusion module; The interface transfer module is based on a model of substance transfer across the membrane and constructs a permeation rate model for each component based on the relationship between feed flow, permeate flow, and retentate flow. When the difference in permeation rate of each component does not exceed a preset value, and concentration polarization does not occur in the membrane separation model, the convection-diffusion module constructs a permeation model and boundary conditions for each component based on the relationship between the concentration gradient, pressure gradient, and velocity gradient along the length of the membrane assembly; When the difference in permeation rates of the components exceeds a preset value, concentration polarization occurs in the membrane separation model. The convection-diffusion module then introduces a concentration polarization construction model to construct a permeation model of each component based on the concentration gradient, pressure gradient, velocity gradient and concentration polarization phenomenon along the length of the membrane assembly, as well as boundary conditions, the mass transfer driving force of the concentration polarization layer and the molar gas concentration of each component in the concentration polarization layer.

7. The method for predicting separation performance of a selective permeable membrane according to claim 6, wherein: When concentration polarization occurs in the membrane separation model, the corresponding interface transfer equation is: d(ux)=k(xx s )+xdu, where Where u represents the fluid flow rate on the retentate side of the membrane module, v represents the fluid flow rate on the permeate side of the membrane module, x represents the molar concentration on the retentate side of the membrane module, y represents the molar concentration on the permeate side of the membrane module, and p o Indicates the external pressure of the wire, p i represents the pressure inside the wire, J1 is the permeability coefficient, J2 is the permeability coefficient of the difficult-to-permeate component, and k is the mass transfer coefficient; Optimize and solve the interface transfer equations, including: Step 1: Discretize the integral interval [0, L] into n subintervals: z0 = 0, z1 = Δz, z2 = 2Δz, …, z n =L, where Δz = L / n; Step 2: Create an alternative equation for the integral equation based on the trapezoidal rule: Step 3: Apply the substitution equation to solve the interface transfer equation: definition: Establish an alternative equation for solving the interface transfer equation based on the trapezoidal rule: Step 4: Calculate f(z i ): Calculate each x(z i ) and x s (z i ), and then we get f(z i ); Step 5: Iterative solution, according to each f(z i ) value, iterate and update the distance z i The permeation flux ux(z i ).

8. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The selective permeability membrane separation performance prediction method further comprises: Step S4: Obtain the previous prediction results and the corresponding real data, and perform polynomial fitting on the errors between the two to obtain the error correction value. Train the output neural network model, where and All are matrices, and the polynomial fitting is adjusted to obtain the final prediction correction model.

9. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The change in the length direction of the membrane separation model is calculated by the following equation: The calculation equation of the retentate flow rate u is: u(i+1)=u(i)-du, where u(i) is the flow rate at the i-th node on the retentate side; The calculation equation of permeation flow v is: v(i+1)=v(i)+dv, where v(i) is the flow rate at the i-th node on the permeation side; The calculation equation for the retentate concentration x is: x(i+1)=x(i)-dx, where x(i) is the concentration of the component at the i-th node on the retentate side; The calculation equation of the permeate concentration y is: y(i+1)=y(i)+dy, where y(i) is the concentration of the component at the i-th node on the permeate side.

10. The method for predicting separation performance of a selective permeable membrane according to claim 1, wherein: The membrane separation model has no less than 2 components; The change of any component j is calculated by the following model: Retentate flow u j Calculation equation: u j (i+1)=u j (i) -du; Permeate flow v j Calculation equation: v j (i+1)=v j (i)+dv; Retentate concentration x j Calculation equation: x j (i+1)=x j (i)-dx; Osmotic concentration j Calculation equation: y j (i+1)=y j (i)+dy; The total pressure is the sum of the pressures of the components: m is the number of components in the system; The total retentate flow rate u is the sum of the retentate flow rates of each component: The total retentate flow rate v is the sum of the retentate flow rates of each component: The sum of the retentate concentrations calculated as percentages is 1: The sum of the osmotic concentrations calculated as percentages is 1:

Images