Method and system for evaluating axial mixing in extraction tower based on CFD simulation

Abstract

The invention discloses an evaluation method and system for axial mixing in an extraction tower based on CFD simulation, and relates to the technical field of liquid extraction. The method comprises the following steps: constructing geometric and physical models of an extraction tower, carrying out CFD flow field simulation in the extraction tower, obtaining a component concentration sequential sequence {smt} at a simulated monitoring point and a component concentration sequential sequence {szt} at a real monitoring point in the tower, carrying out difference processing on a simulated value and a real value of the component concentration to obtain a residual sequence {delta st} of the component concentration, and carrying out wavelet transform processing to obtain a component concentration residual sequence {delta st}; the method comprises the following steps: defining a data state of a component concentration sampling value at a t moment, constructing a data state transfer matrix A = (aij) 2 * 2, calculating a probability value pt1 that the state of the component concentration sampling value is 0 or 1, judging an abnormal condition of the obtained component concentration at a real monitoring point in a tower at the t moment according to a calculation result, identifying and eliminating abnormal data in a data sequence {szt}, and determining the abnormal condition of the component concentration at the real monitoring point in the tower at the t moment. And obtaining a component concentration sequential sequence {szt '} at a real monitoring point in the reconstructed tower, and calculating an axial mixing coefficient in the extraction tower according to the sequence {szt'}.

Description

Technical Field

[0001] The invention relates to the technical field of liquid extraction, and in particular to an evaluation method and system for axial mixing in an extraction tower based on CFD simulation. Background Art

[0002] At present, the research on axial mixing is mainly based on the axial mixing model, that is, a single axial mixing coefficient is used to describe the non-ideality of the flow. Generally, after obtaining the residence time distribution, the axial mixing coefficient can be calculated using the approximate solution of the axial mixing model to evaluate the degree of axial backmixing in the tower. This method has been successfully applied to the calculation of the axial mixing coefficient in the rotating disk tower.

[0003] At present, computational fluid dynamics (CFD) simulation is often used to obtain residence time distribution curves. During the simulation process, it is necessary to construct the geometric and physical model of the turntable tower, set the sampling points, obtain the functional relationship between the component concentration and the residence time at the sampling points, and thus obtain the residence time distribution curve of the components in the tower. The entire simulation process is similar to being in an idealized environment, but in actual situations, the component concentration is often affected by many factors. Any change in any influencing factor may cause the component concentration at the sampling point to change, resulting in a deviation between the actual situation and the simulation results, and failing to reflect the actual distribution of axial mixing in the tower. To this end, we propose an evaluation method and system for axial mixing in an extraction tower based on CFD simulation. Summary of the invention

[0004] The main purpose of the present invention is to provide an evaluation method and system for axial mixing in an extraction tower based on CFD simulation, which can effectively solve the problems in the background technology.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] The evaluation method of axial mixing in the extraction column based on CFD simulation includes:

[0007] The geometric parameters and processing parameters of the extraction tower to be evaluated are obtained, and the geometric and physical models of the extraction tower are constructed. The CFD flow field in the extraction tower is simulated according to the constructed model. After the CFD flow field in the tower is stable, the pulse signal is simulated by injecting pure components within a fixed time interval Δt to obtain the component concentration time series {s mt}, t = 1, 2, ..., N;

[0008] The extraction tower to be evaluated is operated with the same processing parameters to obtain the time series of component concentrations at the real monitoring points in the tower. zt}, the simulated value of the component concentration is subtracted from the true value to obtain the residual sequence of the component concentration {Δs t}, where Δst =s mt -s zt , and perform wavelet transform on the residual sequence;

[0009] Define the data state of the component concentration sampling value at time t as S, where S = {0, 1}, S = 1 means the data state is normal; S = 0 means the data state is abnormal, and construct the data state transfer matrix A = (a ij ) 2×2 , where a ij represents the probability value of data transitioning from state i to state j in one step, a ij =P(s t =js t-1 =i); i, j∈S;

[0010] Calculate the probability value p that the component concentration sampling value state is 0 or 1 at time t t1 , the calculation formula is:

[0011]

[0012] Where N[·|·] is the Gaussian distribution function; W(t,f) is the wavelet coefficient at time t; W a is the average value of normal wavelet coefficients; W v is the variance of normal wavelet coefficients; T is the Gaussian distribution function relationship; according to the calculation results, the abnormal situation of the component concentration at the real monitoring point in the tower at time t is determined, and the data sequence {s zt}, and obtain the reconstructed component concentration time series at the real monitoring point in the tower {s zt '}, the abnormal situation judgment principle is:

[0013] Define the function θ t (1) θ t (0), where θ t (1) = a i1 ×p t1 θ t (0) = a i0 ×p t0 ;

[0014] When θ t (1)≥θ t (0), there is s t =1, indicating that the component concentration at the actual monitoring point in the tower is a normal data value;

[0015] When θ t (1)<θ t (0), there is s t =0, indicating that the component concentration at the actual monitoring point in the tower is an abnormal data value;

[0016] sequence{s zt The acquisition process of '} includes the following steps:

[0017] Get data sequence {s zt Abnormal data in} zq , where q∈N;

[0018] Extract data sequence {s zt Abnormal data in zq The first m data of are s z(q-1) 、s z(q-2) ,...,s z(q-m) , where qm≥1;

[0019] Substitute the extracted data into the calculation formula: In the formula, s zq ' is abnormal data s zq The reconstructed data value; λ is a constant coefficient;

[0020] The data sequence {s zt Abnormal data in} zq Replaced with the reconstructed data value s zq ', that is, to obtain the sequence {s zt '};

[0021] According to the obtained sequence {s zt '}Calculate the axial mixing coefficient in the extraction tower to be evaluated. The calculation formula for the axial mixing coefficient is:

[0022]

[0023] Where Pe is the axial mixing coefficient; u is the fluid velocity in the tower; L is the characteristic length; t a Expressed as the average residence time; σ 2 It is expressed as the variance of residence time distribution; D is the diffusion coefficient of the fluid in the tower; the calculation formula for the average residence time is: In the formula, s k Component concentration sampling value; t k The sampling value of the component concentration is s k The calculation formula of the residence time distribution variance is:

[0024] An evaluation system for axial mixing in an extraction tower based on CFD simulation, comprising a tower type data acquisition module, a CFD flow field simulation module, a component data acquisition module, a data fitting processing module, an abnormal data determination module, and an evaluation module for axial mixing in a tower;

[0025] The tower data acquisition module is used to obtain the geometric parameters and processing parameters of the extraction tower to be evaluated, and to construct the geometric and physical model of the extraction tower;

[0026] The CFD flow field simulation module is used to perform CFD flow field simulation in the extraction tower according to the constructed model. After the CFD flow field in the tower is stable, a pulse signal is simulated by injecting pure components within a fixed time interval Δt to obtain a time series of component concentrations at the simulated monitoring point.

[0027] The component data acquisition module is used to obtain the component concentration time series at the real monitoring points in the tower when the extraction tower to be evaluated is operated with the same processing parameters;

[0028] The data fitting processing module is used to perform difference processing on the simulated value of the component concentration and the real value to obtain the residual sequence of the component concentration, and perform wavelet transformation processing on the residual sequence;

[0029] The abnormal data determination module is used to determine the abnormal situation of the component concentration at the real monitoring point in the tower at time t, and identify and eliminate the abnormal data in the data sequence;

[0030] The in-tower axial mixing evaluation module is used to obtain the reconstructed component concentration time series at the real monitoring points in the tower, and calculate the axial mixing coefficient in the extraction tower to be evaluated based on the obtained sequence;

[0031] The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor.

[0032] The present invention has the following beneficial effects:

[0033] Compared with the existing technology, by constructing the geometric and physical model of the extraction tower, CFD flow field simulation is performed in the extraction tower to obtain the component concentration time series at the simulated monitoring point {s mt}, the extraction tower to be evaluated is operated with the same processing parameters to obtain the time series of component concentrations at the real monitoring points in the tower {s zt}, the simulated value of the component concentration is subtracted from the true value to obtain the residual sequence of the component concentration {Δs t}, and perform wavelet transform on the residual sequence, define the data state of the component concentration sampling value at time t as S, and construct the data state transfer matrix A = (a ij ) 2×2 , calculate the probability value p of the component concentration sampling value being 0 or 1 at time t t1 According to the calculation results, the abnormal situation of the component concentration at the real monitoring point in the tower at time t is determined, and the data sequence {s zt}, and obtain the reconstructed component concentration time series at the real monitoring point in the tower {szt '}, according to the obtained sequence {s zt '}Calculating the axial mixing coefficient in the extraction tower to be evaluated can determine the abnormal data in the component concentration sampling values, thereby constructing an accurate component concentration time series, reducing the deviation between the actual situation and the simulation results caused by system or sampling errors, and thus more accurately reflecting the actual distribution of axial mixing in the tower. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flow chart of the evaluation method of axial mixing in an extraction tower based on CFD simulation of the present invention;

[0035] Figure 2 It is a structural block diagram of the evaluation system of axial mixing in the extraction tower based on CFD simulation of the present invention;

[0036] Figure 3 Schematic diagram of the residence time distribution curve of the components in the tower. DETAILED DESCRIPTION

[0037] The present invention will be further described below in conjunction with specific implementation methods, wherein the accompanying drawings are only used for exemplary descriptions and represent only schematic diagrams rather than actual drawings, and should not be understood as limiting the present invention. In order to better illustrate the specific implementation methods of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product.

[0038] The implementation process of the technical solution of the present invention includes the following steps:

[0039] Step 1: Obtain the geometric parameters and processing parameters of the extraction tower to be evaluated, and construct the geometric and physical model of the extraction tower;

[0040] It should be noted that the geometric parameters mainly include the tower diameter, shaft diameter, internal structure form and size of the extraction tower; the processing parameters mainly include flow value, speed value, fluid type, etc.

[0041] Step 2: According to the constructed model, CFD flow field simulation is performed in the extraction tower. During the simulation process, the model adopts the assumption of two-dimensional rotation axis symmetry, the computational domain is discretized using quadrilateral grids, the simulated fluid medium adopts velocity boundary conditions at both the tower top inlet and the tower bottom outlet, and no-slip boundary conditions are adopted for all the walls in the tower. After the CFD flow field in the tower is stable, the pulse signal is simulated by injecting pure components within a fixed time interval Δt to obtain the component concentration time series {s mt}, t = 1, 2, ..., N;

[0042] Step 3: Run the extraction tower to be evaluated with the same processing parameters to obtain the time series of component concentrations at the actual monitoring points in the tower {szt}, where the sampling step length of the component concentration time series at the real monitoring point in the tower is equal to the step length of the component concentration time series at the simulated monitoring point;

[0043] Step 4: Subtract the simulated value of the component concentration from the true value to obtain the residual sequence of the component concentration {Δs t}, where Δs t =s mt -s zt , and perform wavelet transform on the residual sequence;

[0044] It should be noted that the residual sequence of component concentrations itself contains the systematic error of measurement. The wavelet decomposition of the residual sequence is used to distinguish the systematic error from the gross error. When the signal is relatively smooth, it is manifested as a small number of high-frequency components, while the gross error contains relatively more high-frequency components. That is, in the process of wavelet decomposition, as the scale increases, the amplitude corresponding to the systematic error in the residual sequence is relatively small, while the high-frequency detail component corresponding to the abnormal data is relatively large, so that the gross error information characteristics are reflected in the decomposition coefficient at the largest scale, thereby avoiding the interference of noise and systematic errors and further improving the recognition accuracy of the data.

[0045] Step 5: define the data state of the component concentration sampling value at time t as S, where S = {0, 1}, S = 1 indicates that the data state is normal; S = 0 indicates that the data state is abnormal;

[0046] Step 6: According to the definition of data state, construct the data state transfer matrix A = (a ij ) 2×2 , where a ij represents the probability value of data transitioning from state i to state j in one step, a ij =P(s t =js t-1 =i); i, j∈S, then there is:

[0047] Step 7: Calculate the probability value p of the component concentration sampling value being 0 or 1 at time t t1 , the calculation formula is:

[0048]

[0049] Where N[·|·] is the Gaussian distribution function; W(t,f) is the wavelet coefficient at time t; W a is the average value of normal wavelet coefficients; W v is the variance of normal wavelet coefficients; T is the Gaussian distribution function relationship;

[0050] Step 8: Determine the abnormal situation of the component concentration at the real monitoring point in the tower at time t based on the calculation results, identify and eliminate the data sequence {s zt}, and obtain the reconstructed component concentration time series at the real monitoring point in the tower {s zt '},in,

[0051] The principles for determining abnormal situations are:

[0052] Define the function θ t (1) θ t (0), where θ t (1) = a i1 ×p t1 θ t (0) = a i0 ×p t0 ;

[0053] When θ t (1)≥θ t (0), there is s t =1, indicating that the component concentration at the actual monitoring point in the tower is a normal data value;

[0054] When θ t (1)<θ t (0), there is s t =0, indicating that the component concentration at the actual monitoring point in the tower is an abnormal data value;

[0055] sequence{s zt The acquisition process of '} includes the following steps:

[0056] Step 811: Get data sequence {s zt Abnormal data in} zq , where q∈N;

[0057] Step 812: Extract data sequence {s zt Abnormal data in zq The first m data of are s z(q-1) 、s z(q-2) ,...,s z(q-m) , where qm≥1;

[0058] Step 813: Substitute the extracted data into the calculation formula: In the formula, s zq ' is abnormal data s zq The reconstructed data value; λ is a constant coefficient;

[0059] Step 814: Transform the data sequence {s zt Abnormal data in} zq Replaced with the reconstructed data value szq ', that is, to obtain the sequence {s zt '};

[0060] Step 9: According to the obtained sequence {s zt '}Calculate the axial mixing coefficient in the extraction tower to be evaluated. The calculation formula for the axial mixing coefficient is:

[0061]

[0062] Where Pe is the axial mixing coefficient; u is the fluid velocity in the tower; L is the characteristic length; t a Expressed as the average residence time; σ 2 It is expressed as the variance of residence time distribution; D is the diffusion coefficient of the fluid in the tower; the calculation formula for the average residence time is: In the formula, s k Component concentration sampling value; t k The sampling value of the component concentration is s k The calculation formula of the residence time distribution variance is:

[0063] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for evaluating axial mixing in an extraction column based on CFD simulation, characterized in that: include: The geometric parameters and processing parameters of the extraction tower to be evaluated are obtained, and the geometric and physical models of the extraction tower are constructed. The CFD flow field in the extraction tower is simulated according to the constructed model. After the CFD flow field in the tower is stable, the pulse signal is simulated by injecting pure components within a fixed time interval Δt to obtain the component concentration time series {s mt }, t = 1, 2, ..., N; The extraction tower to be evaluated is operated with the same processing parameters to obtain the time series of component concentrations at the real monitoring points in the tower. zt }, the simulated value of the component concentration is subtracted from the true value to obtain the residual sequence of the component concentration {Δs t }, where Δs t =s mt -s zt , and perform wavelet transform on the residual sequence; Define the data state of the component concentration sampling value at time t as S, where S = {0, 1}, S = 1 means the data state is normal; S = 0 means the data state is abnormal, and construct the data state transfer matrix A = (a ij ) 2×2 , where a ij represents the probability value of data transitioning from state i to state j in one step, a ij =P(s t =js t-1 =i); i, j∈S; Calculate the probability value p that the component concentration sampling value state is 0 or 1 at time t t1 According to the calculation results, the abnormal situation of the component concentration at the real monitoring point in the tower at time t is determined, and the data sequence {s zt }, and obtain the reconstructed component concentration time series at the real monitoring point in the tower {s zt '}; According to the obtained sequence {s zt '}Calculate the axial mixing coefficient in the extraction column to be evaluated.

2. The method for evaluating axial mixing in an extraction tower based on CFD simulation according to claim 1, characterized in that: Probability value p t1 The calculation formula is: Where N[·|·] is the Gaussian distribution function; W(t,f) is the wavelet coefficient at time t; W a is the average value of normal wavelet coefficients; W v is the variance of normal wavelet coefficients; T is the Gaussian distribution function relationship.

3. The method for evaluating axial mixing in an extraction tower based on CFD simulation according to claim 1, characterized in that: The principles for determining abnormal situations are: function definition t (1)、θ t (0), among which, i t (1)=a i1 ×p t1 ;θ t (0)=a i0 ×p t0 ; When θ t (1)≥θ t (0), there is s t =1, indicating that the component concentration at the actual monitoring point in the tower is a normal data value; When θ t (1)<θ t (0), there is s t =0, indicating that the component concentration at the actual monitoring point in the tower is an abnormal data value.

4. The method for evaluating axial mixing in an extraction tower based on CFD simulation according to claim 1, characterized in that: The axial mixing coefficient is calculated as: Where Pe is the axial mixing coefficient; u is the fluid velocity in the tower; L is the characteristic length; t a Expressed as the average residence time; σ 2 It is expressed as the residence time distribution variance; D is the diffusion coefficient of the fluid in the tower.

5. The method for evaluating axial mixing in an extraction tower based on CFD simulation according to claim 4, characterized in that: The calculation formula for the average residence time is: In the formula, s k Component concentration sampling value; t k The sampling value of the component concentration is s k The dwell time of The calculation formula for the variance of the residence time distribution is:

6. The method for evaluating axial mixing in an extraction tower based on CFD simulation according to claim 1, characterized in that: sequence{s zt The acquisition process of '} includes the following steps: Get data sequence {s zt Abnormal data in} zq , where q∈N; Extract data sequence {s zt Abnormal data in zq The first m data of z(q-1) 、s z(q-2) ,...,s z(q-m) , where qm≥1; Substitute the extracted data into the calculation formula: In the formula, s zq ' is abnormal data s zq The reconstructed data value; λ is a constant coefficient; The data sequence {s zt Abnormal data in} zq Replaced with the reconstructed data value s zq ', that is, to obtain the sequence {s zt '}.

7. Evaluation system for axial mixing in an extraction column based on CFD simulation, characterized in that: It includes tower data acquisition module, CFD flow field simulation module, component data acquisition module, data fitting processing module, abnormal data determination module, and in-tower axial mixing evaluation module; The tower data acquisition module is used to obtain the geometric parameters and processing parameters of the extraction tower to be evaluated, and to construct the geometric and physical model of the extraction tower; The CFD flow field simulation module is used to simulate the CFD flow field in the extraction tower according to the constructed model. After the CFD flow field in the tower is stable, the pulse signal is simulated by injecting pure components within a fixed time interval Δt to obtain the component concentration time series {s mt }, t = 1, 2, ..., N; The component data acquisition module is used to obtain the component concentration time series {s zt }; The data fitting processing module is used to perform difference processing between the simulated value and the real value of the component concentration to obtain the residual sequence {Δs t }, where Δs t =s mt -s zt , and perform wavelet transform on the residual sequence; The abnormal data determination module is used to determine the abnormal situation of the component concentration at the real monitoring point in the tower at time t, identify and eliminate the data sequence {s zt }, the specific implementation process of the abnormal data determination module includes the following steps: The data state of the component concentration sampling value at time t is defined as S, where S = {0, 1}, S = 1 indicates that the data state is normal; S = 0 indicates that the data state is abnormal; Construct data state transition matrix A = (a ij ) 2×2 , where a ij represents the probability value of data transitioning from state i to state j in one step, a ij =P(s t =js t-1 =i); i, j∈S; Calculate the probability value p that the component concentration sampling value state is 0 or 1 at time t t1 , the calculation formula is: Where N[·|·] is the Gaussian distribution function; W(t,f) is the wavelet coefficient at time t; W a is the average value of normal wavelet coefficients; W v is the variance of normal wavelet coefficients; T is the Gaussian distribution function relationship; The principles for determining data anomalies are: function definition t (1)、θ t (0), among which, i t (1)=a i1 ×p t1 ;θ t (0)=a i0 ×p t0 ; When θ t (1)≥θ t (0), there is s t =1, indicating that the component concentration at the actual monitoring point in the tower is a normal data value; When θ t (1)<θ t (0), there is s t =0, indicating that the component concentration at the actual monitoring point in the tower is an abnormal data value; The tower axial mixing assessment module is used to obtain the reconstructed component concentration time series at the real monitoring point in the tower. zt '}, and according to the obtained sequence {s zt '}Calculate the axial mixing coefficient Pe in the extraction tower to be evaluated, where the sequence {s zt The acquisition process of '} includes the following steps: Get data sequence {s zt Abnormal data in} zq , where q∈N; Extract data sequence {s zt Abnormal data in zq The first m data of z(q-1) 、s z(q-2) ,...,s z(q-m) , where qm≥1; Substitute the extracted data into the calculation formula: In the formula, s zq ' is abnormal data s zq The reconstructed data value; λ is a constant coefficient; The data sequence {s zt Abnormal data in} zq Replaced with the reconstructed data value s zq ', that is, to obtain the sequence {s zt '}; The calculation formula of Pe is: Where Pe is the axial mixing coefficient; u is the fluid velocity in the tower; L is the characteristic length; t a is the average residence time, s k Component concentration sampling value; t k The sampling value of the component concentration is s k The residence time of 2 Expressed as the variance of the residence time distribution, D is the diffusion coefficient of the fluid in the tower.

8. The evaluation system for axial mixing in an extraction tower based on CFD simulation according to claim 7, characterized in that: The system comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the steps of the method according to any one of claims 1 to 6 are implemented when the processor executes the program.

Images