A design method for resisting flow-induced wear of a distributor plate in a catalytic gasoline desulfurization reactor

Abstract

The present invention discloses a method for designing the anti-flow-induced wear of the distribution plate of the catalytic gasoline desulfurization reactor. Select a number of adjustable variables; screen the adjustable variables to obtain the influencing variables; sample the influencing variables to construct different design schemes, and numerically simulate and calculate to obtain the corresponding flow-induced wear rate and gas-solid contact efficiency; establish a proxy model according to the multivariate linear regression expression, substitute it into the solution to determine the proxy model; establish an objective function, solve the objective function to obtain the optimal value of each influencing variable, and use the optimal value of each influencing variable as the design result of the anti-flow-induced wear of the distribution plate of the catalytic gasoline desulfurization reactor. The method of the present invention selects significant variables through statistical analysis, obtains the sample value, and obtains the proxy model through linear regression, and combines the two into the objective function through linear weighting according to the relationship between the two, thereby realizing a fast and accurate design of the anti-flow-induced wear structure of the distribution plate of the catalytic gasoline desulfurization reactor.

Description

Technical Field

[0001] The invention relates to a method for designing and processing internal flow characteristics of a process equipment system in a petrochemical industry, and in particular to a method for designing the anti-flow-induced wear of a distribution plate of a catalytic gasoline desulfurization reactor based on an intelligent optimization algorithm. Background Art

[0002] The catalytic cracking gasoline adsorption desulfurization (S Zorb) technology is one of the core technologies for producing ultra-low sulfur clean gasoline in the petrochemical industry. It is mainly used for the desulfurization of all components of catalytic cracking gasoline, and has the advantages of high desulfurization depth, low hydrogen consumption, and low octane number loss. At present, more than 30 sets of industrial units have been built in China, with a total processing capacity of more than 40Mt / a. It is widely used and has significant economic and social benefits.

[0003] The raw material of the S Zorb unit is mainly catalytic cracking gasoline, and the desulfurization purpose is achieved mainly by selectively adsorbing sulfur atoms in sulfide-containing products by adsorbents. Its core process is mainly desulfurization reaction. The specific process part can be described as follows: the sulfur-containing gasoline from the catalytic unit or tank area enters the raw material buffer tank after passing through the filter, is pressurized by the reaction feed pump and mixed with circulating hydrogen, and then enters the adsorption feed heat exchanger in two ways to exchange heat with the product at the top of the reactor. The mixed hydrogen raw material after heat exchange is heated in the feed heating furnace, and enters the bottom of the desulfurization reactor after reaching a predetermined temperature and undergoes adsorption desulfurization reaction in the desulfurization reactor. The desulfurization reactor is equipped with an adsorbent, and the mixed hydrogen raw material flows from bottom to top inside the reactor to form a fluidized state inside the reactor. After the raw material is acted on by the adsorbent, the organic sulfide in it is removed. In order to prevent the adsorbent from being brought into the subsequent system, a filter and backflush measures are set at the top of the reactor to separate the adsorbent dust carried in the product and to clean (backflush) the filter online. In this process, the hydrogen-containing gasoline is heated in a heating furnace to a temperature higher than 400°C and a hydrogen flow rate greater than 6000m 3 / h. The flow direction of hydrogen-containing gasoline at the distribution plate below the bubble cap turns 180°, and the upper surface of the distribution plate is affected by the combined influence of high temperature and high pressure gas-solid two-phase flow (carrying adsorbent), and the flow-induced wear effect is obvious, resulting in flow-induced wear failure under long-term action.

[0004] In view of the fact that the reactor feed distributor is composed of a distribution plate and a bubble cap installed on the distribution plate, the use of the bubble cap can evenly distribute the reaction medium, improve its internal flow conditions, and achieve good contact with the catalyst, thereby achieving uniform distribution in the radial and axial directions. However, through the conventional trial and error method, the design efficiency is low, and it is difficult to accurately find better structural characteristic parameters to avoid or reduce the flow-induced wear rate and improve the gas-solid contact efficiency. To this end, the purpose of the present invention is to provide a catalytic gasoline desulfurization reactor distribution plate anti-flow-induced wear design method based on an intelligent optimization algorithm, and optimize the structural parameters of the distribution plate based on the intelligent optimization algorithm, thereby reducing the flow-induced wear rate on the distribution plate surface, improving the gas-solid contact efficiency of the gas-solid two-phase flow, and improving the desulfurization depth of catalytic cracking gasoline, so as to meet the national environmental protection requirements. Summary of the invention

[0005] In order to solve the problems existing in the background technology, the present invention proposes a design method for anti-flow-induced wear of a distribution plate of a catalytic gasoline desulfurization reactor based on an intelligent optimization algorithm.

[0006] The technical solution adopted by the present invention includes:

[0007] Step 1) Preliminarily selecting a number of adjustable variables and setting specific thresholds for each adjustable variable according to the original structural parameters of the catalytic gasoline desulfurization reactor;

[0008] Step 2) using a screening test design method to screen the adjustable variables of the distribution plate of the catalytic gasoline desulfurization reactor, and screening the retained adjustable variables as influencing variables;

[0009] Step 3) Latin hypercube sampling method is used to sample the influencing variables obtained in step 2) to construct different design schemes, and numerical simulation calculations are performed on different design schemes to obtain flow-induced wear rate E and gas-solid contact efficiency C E ;

[0010] Specifically, the given influencing variables are randomized, that is, the influencing variables in step 1) are divided into equal probability intervals, and random values ​​are generated from each interval, and the number of values ​​in each interval is proportional to the probability of occurrence of the values ​​in the interval.

[0011] 4) Establish gas-solid contact efficiency C E and the proxy model of flow-induced wear rate E, the gas-solid contact efficiency C corresponding to each design scheme E The flow-induced wear rate E and the influencing variables in each design scheme are substituted into the solution to obtain the coefficients in the proxy model, thereby determining the proxy model, that is, determining the gas-solid contact efficiency C E The relationship between the flow-induced wear rate E and each influencing variable;

[0012] 5) Taking the minimum flow-induced wear rate and stable gas-solid contact efficiency as the optimization goal, according to the gas-solid contact efficiency C obtained in step 4) E The objective function is established by the proxy model of flow-induced wear rate E, and the optimal value of each influencing variable is solved to obtain the optimal value of each influencing variable. The optimal value of each influencing variable is used as the design result of the anti-flow-induced wear of the distribution plate of the catalytic gasoline desulfurization reactor.

[0013] The adjustable variables selected in step 1) include eight variables, namely, the distance h1 from the lower end surface of the bubble to the upper surface of the distribution plate, the number of teeth m of the bubble, the radius r1 of the bubble cavity, the height h2 of the lifting tube on the distribution plate, the stacking height h3 of the adsorbent particles, the material inlet flow rate v, the sphericity j of the adsorbent particles, the particle size D of the adsorbent particles p .

[0014] The step 2) is specifically as follows:

[0015] The test groups of each adjustable variable are determined according to the screening test design method, and the flow-induced wear rate E and gas-solid contact efficiency C corresponding to different test groups are calculated using the gas-solid two-phase flow-induced wear characteristic calculation method. E ;

[0016] The flow-induced wear rate E and gas-solid contact efficiency C of different test groups were E The calculation results are used as samples for statistical analysis, and the arrangement distribution diagram of the standardized effect is made. The significance test value of each adjustable variable is calculated according to the following formula:

[0017]

[0018] Where, T xi , P xi They represent the first significance test value and the second significance test value of the ith adjustable variable respectively, and xi represents the ith adjustable variable; represents the sample mean, S * represents the sample standard deviation, n represents the sample size, μ represents the sample expectation, Represents the upper quantile of the t distribution;

[0019] Then make the following judgment:

[0020]

[0021] If the first significance test value T is satisfied xi ≠0, but the second significance test value P xi If the value is less than 0.05, the adjustable variable is retained as an influencing variable; otherwise, the adjustable variable is not retained as an influencing variable.

[0022] The step 3) is specifically as follows:

[0023] 3.1) Normalize each influencing variable to a 0-1 numerical interval, perform statistical analysis on each influencing variable, and divide the 0-1 numerical interval of the optimization variable into M non-overlapping intervals with equal probability;

[0024] 3.2) Randomly select values ​​for each influencing variable in each interval to form a sampling point. Each sampling point represents a design scheme with different influencing variable combinations. Multiple sampling points are taken as matrix elements and stored in a T×M sampling point matrix Q, which can be expressed as:

[0025]

[0026] Q TM ={x1~x N}

[0027] Among them, Q TM represents the Tth row and Mth column element of the sampling point matrix Q, x1 represents the value of the first influencing variable in the range of 0-1, and x N Indicates the value of the Nth influencing variable in the range of 0-1;

[0028] 3.3) Re-arrange the elements of each row in the sampling point matrix Q randomly to obtain the Latin hypercube sample matrix W, which is expressed as:

[0029]

[0030] Where: W is the Latin hypercube sample matrix; Q xNM represents the xth N The element in the row and column M, W NM Represents the Nth row and Mth column element of the sampling point matrix Q;

[0031] 3.4) The numerical simulation method is used to calculate and solve each sampling point in the Latin hypercube sample matrix to obtain the flow-induced wear rate E and gas-solid contact efficiency C corresponding to each design scheme. E .

[0032] In step 4), the gas-solid contact efficiency C is established according to the multivariate linear regression expression. E and the proxy model of flow-induced wear rate E, respectively, are expressed as:

[0033]

[0034] Where: b 1,0 ~b 1,i Indicates the gas-solid contact efficiency C E The coefficients of each proxy model, b 2,0 ~b 2,iThe coefficients in the proxy model representing the flow-induced wear rate E are:

[0035] In the step 4), the coefficients of each item in the proxy model are obtained by solving the least square method.

[0036] In the step 5), according to the gas-solid contact efficiency C E The proxy model of flow-induced wear rate E uses a weighted linear combination to establish the objective function:

[0037]

[0038] x I £x£x J

[0039] F1(x)=|f1(x)-0.9612

[0040] F2(x)=f2(x)

[0041]

[0042] In the formula, x I and x J are the minimum and maximum values ​​of the influencing variables, respectively, w j represents the weighted value of the jth proxy model, j = 1, 2, Fj(x) represents the function expression of the jth proxy model, and F1(x) represents the gas-solid contact efficiency C E The proxy model function expression of F2(x) is the proxy model function expression of flow-induced wear rate E; f1(x) is the gas-solid contact efficiency C E f2(x) is a proxy value of flow-induced wear rate E.

[0043] In the step 5), the objective function is specifically solved by using a particle swarm optimization algorithm to obtain the best value of the influencing variable.

[0044] The distribution plate of the catalytic gasoline desulfurization reactor is a distribution plate in the catalytic gasoline desulfurization reactor, on which adsorbent particles are piled and arranged, and the distribution plate flows through the sulfur-containing gas-phase gasoline, with the sulfur-containing gas-phase gasoline as the gas-phase fluid and the adsorbent particles as the solid-phase particles. When the sulfur-containing gas-phase gasoline is mixed with the adsorbent particles, a gas-solid two-phase flow is formed;

[0045] The bottom of the catalytic gasoline desulfurization reactor is provided with a material inlet for the entry of sulfur-containing gaseous gasoline. A distribution plate is arranged in the tower above the material inlet. Through holes are evenly arranged on the distribution plate. A lifting pipe is arranged and installed in each through hole. The top of the upper end of the lifting pipe is connected to a bubble cap. Adsorbent particles are accumulated on the distribution plate.

[0046] The lower end of the bubble is provided with an opening, the upper end of the lifting tube is connected to the inner top surface of the bubble via support rods arranged at intervals along the circumference, and an annular gap is provided between the opening at the lower end of the bubble and the lifting tube, so that the upper end of the lifting tube is connected to the inner cavity of the bubble via the interval between the support rods, and then connected to the outside via the annular gap.

[0047] The lower end surface of the bubble cap is provided with a plurality of grooves evenly spaced along the circumference, and the sulfur-containing gas phase gasoline first flows out from the grooves. The opening of each groove forms a tooth structure on the lower end surface of the bubble cap, and the grooves serve as tooth grooves of the tooth structure.

[0048] The beneficial effects of the present invention are:

[0049] The present invention provides an optimization design method for the flow-induced wear resistance of the distribution plate of the catalytic gasoline desulfurization reactor based on the simulated data through an intelligent algorithm, and the significant variables are screened out through statistical analysis, and the linear regression is used to obtain the C E The proxy model of E and E is combined into an objective function in a linear weighted manner according to the relationship between the two, which realizes the fast and accurate anti-flow-induced wear structure design of the distribution plate of the catalytic gasoline desulfurization reactor, and can provide an important reference for the optimization of the structure and process parameters of the catalytic gasoline desulfurization reactor. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Attached Figure 1 It is a front view of the distribution plate structure of the catalytic gasoline desulfurization reactor according to an embodiment of the present invention.

[0051] Attached Figure 2 It is an axonometric view of the distribution plate structure of the catalytic gasoline desulfurization reactor according to an embodiment of the present invention.

[0052] Attached Figure 3 It is a schematic diagram of the structure of a single blister according to an embodiment of the present invention.

[0053] Attached Figure 4 It is a schematic diagram of the unfolded structure of the blister according to an embodiment of the present invention.

[0054] Attached Figure 5 It is an arrangement diagram of the standardized effect of flow-induced wear rate E of the present invention.

[0055] Attached Figure 6 is the gas-solid contact efficiency C of the present invention E Pareto plot of standardized effects.

[0056] Attached Figure 7 It is a relationship curve diagram of the dimensionless diameter and the dimensionless flow-induced wear rate of the present invention.

[0057] Attached Figure 8 It is a relationship curve diagram of the dimensionless sphericity and the dimensionless flow-induced wear rate of the present invention.

[0058] Attached Fig. 9 It is a relationship curve diagram between the dimensionless flow velocity and the dimensionless flow-induced wear rate of the present invention.

[0059] Attached Fig.10 It is a relationship curve diagram of the dimensionless stacking height and the dimensionless flow-induced wear rate of the present invention.

[0060] Attached Fig.11 It is a relationship curve diagram of the dimensionless particle size and the dimensionless gas-solid contact efficiency of the present invention.

[0061] Attached Fig.12 It is a relationship curve diagram of the dimensionless sphericity and the dimensionless gas-solid contact efficiency of the present invention.

[0062] Attached Fig.13 It is a relationship curve diagram between the dimensionless flow velocity and the dimensionless gas-solid contact efficiency of the present invention.

[0063] Attached Fig.14 It is a relationship curve diagram of the dimensionless stacking height and the dimensionless flow-induced wear rate of the present invention.

[0064] Explanation of the numbers in the figure: 1. bubble cap, 2. upper surface of distribution plate, 3. lower surface of distribution plate, 4. reactor, 5. lifting pipe, 6. material inlet. DETAILED DESCRIPTION

[0065] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0066] The catalytic gasoline desulfurization reactor distribution plate is a distribution plate in the catalytic gasoline desulfurization reactor 4, on which adsorbent particles are piled and arranged, and the distribution plate flows through the sulfur-containing gas-phase gasoline, with the sulfur-containing gas-phase gasoline as the gas-phase fluid and the adsorbent particles as the solid-phase particles. When the sulfur-containing gas-phase gasoline is mixed with the adsorbent particles, a gas-solid two-phase flow is formed;

[0067] like Figure 1 and Figure 2 As shown, a material inlet 6 for sulfur-containing gaseous gasoline to enter is provided at the bottom of the catalytic gasoline desulfurization reactor 4, a distribution plate is arranged in the tower above the material inlet 6, the distribution plate has an upper surface 2 and a lower surface 3 of the distribution plate, through holes are evenly provided on the distribution plate, and a lifting pipe 5 is arranged and installed in each through hole, the lifting pipe 5 passes through the through hole from top to bottom and is welded and fixed to the lower surface 3 of the distribution plate of the distribution plate, the top of the upper end of the lifting pipe 5 is connected to the bubble 1 through a circumferentially uniformly distributed support plate, and the bubble 1 can be lifted and moved under the drive of the lifting pipe 5; many adsorbent particles are accumulated on the upper surface 2 of the distribution plate of the distribution plate, and accumulated between the bubbles 1, and the accumulation height can be higher than the bubble 1.

[0068] The specific structure of blister 1 is as follows Figure 3 and Figure 4 shown.

[0069] The lower end of the bubble cap 1 has an opening, and the upper end of the lifting tube 5 is connected to the inner top surface of the bubble cap 1 via support rods arranged at intervals along the circumference. There is an annular gap between the opening at the lower end of the bubble cap 1 and the lifting tube 5, so that the upper end of the lifting tube 5 is connected to the inner cavity of the bubble cap 1 via the interval between the support rods, and then connected to the outside via the annular gap.

[0070] After the sulfur-containing gaseous gasoline enters the material inlet 6 of the catalytic gasoline desulfurization reactor 4, it enters each lifting pipe 5 through the through holes on the distribution plate in turn. After the sulfur-containing gaseous gasoline flows out of the upper end of the lifting pipe and is blocked by each bubble cap 1, it impacts the adsorbent particles located on the upper surface 2 of the distribution plate after turning 180 degrees. The sulfur-containing gaseous gasoline entrains the adsorbent particles to form a fluidized gas-solid two-phase flow in the flow space above the distribution plate. The fluidized gas-solid two-phase flow causes flow wear to the distribution plate of the catalytic gasoline desulfurization reactor during the back-mixing process. The multi-component medium of the gas-solid two-phase flow entrains the non-spherical adsorbent particle group to impact the upper surface of the adsorption tower plate, causing flow wear to the upper surface of the distribution plate.

[0071] The implementation process of the present invention mainly includes two aspects: one is the screening of influencing variables, and the other is intelligent optimization and optimization. The specific steps are explained in the following steps.

[0072] The steps for screening influencing variables are as follows:

[0073] Step 1): According to the original structure and typical operating parameters of the catalytic gasoline desulfurization reactor, the adjustable variables are preliminarily selected and the specific thresholds of the adjustable variables are set, wherein the adjustable variables mainly include eight influencing variables, namely, the distance h1 from the lower end surface of the bubble cap to the upper surface of the distribution plate, the number of bubble cap teeth m, the bubble cap radius r1, the height of the lifting tube on the distribution plate h2, the particle stacking height h3, the material inlet flow rate v, the sphericity j of the adsorbent particles, and the particle size D of the adsorbent particles. p .

[0074] Step 2): According to the original structure and typical operating parameters of the catalytic gasoline desulfurization reactor, the following constraints are imposed on the influencing variables, as shown in Table 1.

[0075] Table 1 Constraint thresholds of influencing variables (unit: mm)

[0076]

[0077] Step 3):

[0078] The test groups of each adjustable variable are determined according to the screening test design method, and the flow-induced wear rate E and gas-solid contact efficiency C corresponding to different test groups are calculated using the gas-solid two-phase flow-induced wear characteristic calculation method. E ;

[0079] The flow-induced wear rate E and gas-solid contact efficiency C of different test groups were E The calculation results are used as samples for statistical analysis, and the arrangement distribution diagram of the standardized effect is made (such as Figure 5 The significance test value of each adjustable variable is calculated according to the following formula to determine the high and low levels of each adjustable variable:

[0080]

[0081] Then make the following judgment:

[0082]

[0083] If the first significance test value T is satisfied xi ≠0, but the second significance test value P xi If the value is less than 0.05, the adjustable variable is retained as an influencing variable; otherwise, the adjustable variable is not retained as an influencing variable.

[0084] Table 2 Experimental plan and results

[0085]

[0086]

[0087] To clarify whether each influencing variable is statistically significant, firstly, it is necessary to determine whether the T test and P test results meet the significance evaluation criteria based on step 5); if so, the influencing variable is a significant variable, otherwise, it is an insignificant variable.

[0088] Based on the test scheme determined in Table 2 and the significance judgment criteria of each influencing variable established in step 5), the critical value T of the T test result of the flow-induced wear rate E is determined. C1 =3.180, as long as the T test results of each influencing variable are greater than their critical values ​​and meet P xi The test results show that the influencing variable is a significant factor, and the arrangement diagram of the standardized effect of flow-induced wear rate E is drawn as follows: Figure 5 As shown; after calculation, P of x1, x2, x3, x4 Xi The test results are 0.0020, 0.0310, 0.0030, and 0.0410, respectively. Therefore, x1, x2, x3, and x4 in Table 2 are listed as significant variables and are discussed and analyzed in detail.

[0089] In the specific implementation, the following four influencing variables are obtained by screening the eight adjustable variables mentioned above, namely, the stacking height h3 of the adsorbent particles, the material inlet flow rate v, and the sphericity of the adsorbent particles. The particle size of adsorbent particles D p .

[0090] Similarly, the experimental results of each group were obtained by numerical calculation, and the effects of various factors on the gas-solid contact efficiency C in the reactor were statistically analyzed. E Based on the significance judgment criteria of each influencing variable established in step 5), the gas-solid contact efficiency C is determined E The critical value of the T test result is T C2 =3.182, as long as the T test results of each influencing variable are greater than their critical values ​​and meet P xi The test results show that the influencing variable is a significant factor, and the gas-solid contact efficiency C is plotted. E The arrangement diagram of the standardized effects is as follows Figure 6 As shown; after calculation, P of x1, x2, x3 xi The test results are 0.0034, 0.0350, and 0.0100, respectively. Therefore, x1, x2, and x3 in Table 2 are listed as significant variables and are discussed and analyzed in detail. xi The test result is 0.064, which is relatively close to 0.05. In order to improve the optimization accuracy, it is also taken into consideration.

[0091] The steps of intelligent optimization are as follows:

[0092] Step 1): Use Latin hypercube sampling method to sample and calculate the target value. E The influencing variables of the input variables are divided into equal probability intervals and random values ​​are generated from each interval. The number of values ​​in each interval is proportional to the probability of the value in the interval appearing.

[0093] 1.1) Normalize each influencing variable to a 0-1 numerical interval, combine the influencing variable screening results of step 2), perform statistical analysis on each influencing variable, draw a permutation distribution diagram of the standardized effect, and divide the 0-1 numerical interval of the optimization variable into M non-overlapping intervals with equal probability;

[0094] 1.2) Randomly select values ​​for each influencing variable in each interval to form a sampling point. Each sampling point represents a design scheme with different influencing variable combinations. Multiple sampling points are taken as matrix elements and stored in a T×M sampling point matrix Q, which can be expressed as:

[0095]

[0096] Q TM ={x1~x N}

[0097] Among them, Q TMrepresents the Tth row and Mth column element of the sampling point matrix Q, x1 represents the value of the first influencing variable in the range of 0-1, and x N Indicates the value of the Nth influencing variable in the range of 0-1;

[0098] 1.3) Re-arrange the elements of each row in the sampling point matrix Q randomly to generate a random combination of influencing variables and obtain the Latin hypercube sample matrix W, which is expressed as:

[0099]

[0100] Where: Subscript x 11 、x 12 、x 1M represents the random permutation of the row vectors of the matrix, W is the Latin hypercube sample matrix, and the column vector is the M-dimensional sample; represents the xth N The element in the row and column M, W NM Represents the Nth row and Mth column element of the sampling point matrix Q;

[0101] 1.4) The numerical simulation method is used to calculate and solve each sampling point in the Latin hypercube sample matrix to obtain the flow-induced wear rate E and gas-solid contact efficiency C corresponding to each design scheme. E .

[0102] Step 2): Establish the gas-solid contact efficiency C E and the proxy model of flow-induced wear rate E, the gas-solid contact efficiency C corresponding to each design scheme E The flow-induced wear rate E and the influencing variables in each design scheme are substituted into the solution to obtain the coefficients in the proxy model, thereby determining the proxy model, that is, determining the gas-solid contact efficiency C E The relationship between the flow-induced wear rate E and each influencing variable;

[0103] The gas-solid contact efficiency C was established according to the multivariate linear regression expression. E and the proxy model of flow-induced wear rate E, respectively, are expressed as:

[0104] C E =b 10 +b 11 x1+b 12 x2+...+b 1i x i ...+b 1N x N

[0105] E=b 20 +b 21 x1+b 22 x2+...+b 2i xi ...+b 2N x N

[0106] x=(x1,x2,x3,x4)

[0107] Where: b 10 ~b 1i Indicates the gas-solid contact efficiency C E The coefficients of each proxy model, b 20 ~b 2i The coefficients in the proxy model representing the flow-induced wear rate E, x i is the i-th influencing variable.

[0108] The coefficients of each proxy model are obtained by solving the problem using the least squares method.

[0109] Based on the above formula, we can find the relationship between the dependent variable and multiple independent variables. However, in most cases, many independent variables are not completely linear with the dependent variable, and the form of the independent variable needs to be modified. Taking the flow-induced wear rate as an example, the relationship between any influencing variable and the target value of the flow-induced wear rate is plotted separately and its distribution state is observed, as shown in the figure below. Figure 7 , Figure 8 , Fig. 9 , Fig.10 shown.

[0110] Similarly, the relationship between any influencing variable and the target value of gas-solid contact efficiency is plotted separately and its distribution is observed, as shown in Fig.11 , Fig.12 , Fig.13 , Fig.14 It should be noted that the horizontal axis in the figure represents the dimensionless influencing variable. Taking the dimensionless stacking height as an example, it represents the ratio of the stacking height at the current moment to the maximum stacking height; the vertical axis in the figure represents the dimensionless flow-induced wear rate or the dimensionless gas-solid contact efficiency, which represents the ratio of the value corresponding to the current moment to the maximum value, and is a dimensionless number.

[0111] Step 3): Minimum flow-induced wear rate and stable gas-solid contact efficiency are the optimization goals. According to the gas-solid contact efficiency C obtained in step 4), E The objective function is established by the proxy model of flow-induced wear rate E, and the optimal value of each influencing variable is solved to obtain the optimal value of each influencing variable. The optimal value of each influencing variable is used as the design result of the anti-flow-induced wear of the distribution plate of the catalytic gasoline desulfurization reactor.

[0112] According to the gas-solid contact efficiency C E The proxy model of flow-induced wear rate E uses a weighted linear combination to establish the objective function, that is, F j (x) and weighted value w iThe linear combination constitutes the objective function:

[0113]

[0114] x I £x£x J

[0115] F1(x)=|f1(x)-0.9612

[0116] F2(x)=f2(x)

[0117] f1(x)=b 10 +b 11 x1+b 12 x2+...+b 1i x i ...+b 1N x N

[0118] f2(x)=b 20 +b 21 x1+b 22 x2+...+b 2i x i ...+b 2N x N

[0119] In the formula, x I and x J are the minimum and maximum values ​​of the influencing variables, respectively, w j represents the weighted value of the j-th proxy model, j = 1, 2, Fj(x) represents the function expression of the j-th proxy model, F1(x) represents the proxy model function expression of the gas-solid contact efficiency CE, F2(x) represents the proxy model function expression of the flow-induced wear rate E; f1(x) represents the gas-solid contact efficiency C E f2(x) is a proxy value of flow-induced wear rate E.

[0120] Specifically, the objective function is solved using the particle swarm optimization algorithm to obtain the optimal value of the influencing variables.

[0121] The established objective function is incorporated into the particle swarm optimization algorithm (PSO) for optimization, and the genetic algorithm (GA) is compared and analyzed to obtain the best value of the influencing variables; the optimization results and target values ​​of the design variables are shown in Table 3.

[0122] Table 3 Design variable optimization results and target values

[0123]

[0124]

[0125] Compared with the original design, the flow-induced wear rate of the distribution plate after optimization by the intelligent algorithm has been greatly reduced, while the gas-solid contact efficiency in the reactor is stable at around 26%. Comparing the two algorithms, it can be seen that the gas-solid contact efficiency calculated by the genetic algorithm and the particle swarm algorithm is basically stable within 0.16%, but the flow-induced wear rate corresponding to the design variables obtained by the particle swarm algorithm has been reduced by nearly 30%, and the effect has been significantly improved. Therefore, the particle swarm algorithm is more suitable for the design optimization of the distribution plate of the catalytic gasoline desulfurization reactor against flow-induced wear. The optimal design variable parameters are the adsorbent particle size of 94.16 mm, the particle sphericity of 0.741, the inlet flow rate of 4.021 m / s, and the adsorbent particle stacking height of 190.27 mm.

[0126] The above-mentioned specific implementation is used to explain the relevant contents of the present invention rather than to limit the present invention. Any modification and change made to the present invention within the spirit of the present invention and the protection scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A design method for anti-flow-induced wear of a distribution plate of a catalytic gasoline desulfurization reactor, characterized in that: The specific steps of the method are as follows: Step 1) Preliminarily selecting a number of adjustable variables based on the original structural parameters of the catalytic gasoline desulfurization reactor; Step 2) using a screening test design method to screen the adjustable variables of the distribution plate of the catalytic gasoline desulfurization reactor, and screening the retained adjustable variables as influencing variables; Step 3) Latin hypercube sampling method is used to sample the influencing variables obtained in step 2) to construct different design schemes, and numerical simulation calculations are performed on different design schemes to obtain flow-induced wear rate E and gas-solid contact efficiency C E ; 4) Establish gas-solid contact efficiency C E and the proxy model of flow-induced wear rate E, the gas-solid contact efficiency C corresponding to each design scheme E Substitute the flow-induced wear rate E and the influencing variables in each design scheme into the solution to obtain the coefficients in the proxy model, thereby determining the proxy model; 5) Taking the minimum flow-induced wear rate and stable gas-solid contact efficiency as the optimization goal, according to the gas-solid contact efficiency C obtained in step 4) E The objective function is established by the proxy model of flow-induced wear rate E, and the optimal value of each influencing variable is solved to obtain the optimal value of each influencing variable. The optimal value of each influencing variable is used as the design result of the anti-flow-induced wear of the distribution plate of the catalytic gasoline desulfurization reactor.

2. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 1, characterized in that: The adjustable variables selected in step 1) include eight variables, namely, the distance h1 from the lower end surface of the bubble to the upper surface of the distribution plate, the number of teeth m of the bubble, the radius r1 of the bubble cavity, the height h2 of the lifting tube on the distribution plate, the stacking height h3 of the adsorbent particles, the material inlet flow rate v, the sphericity of the adsorbent particles The particle size of adsorbent particles D p .

3. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 1, characterized in that: The step 2) is specifically as follows: The test groups of each adjustable variable are determined according to the screening test design method, and the flow-induced wear rate E and gas-solid contact efficiency C corresponding to different test groups are calculated using the gas-solid two-phase flow-induced wear characteristic calculation method. E ; The flow-induced wear rate E and gas-solid contact efficiency C of different test groups were E The calculation results are used as samples for statistical analysis, and the arrangement distribution diagram of the standardized effect is made. The significance test value of each adjustable variable is calculated according to the following formula: Where, T xi , P xi They represent the first significance test value and the second significance test value of the ith adjustable variable respectively, and xi represents the ith adjustable variable; represents the sample mean, S * represents the sample standard deviation, n represents the sample size, μ represents the sample expectation, Represents the upper quantile of the t distribution; Then make the following judgment: If the first significance test value T is satisfied xi ≠0, but the second significance test value P xi If the value is less than 0.05, the adjustable variable is retained as an influencing variable; otherwise, the adjustable variable is not retained as an influencing variable.

4. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 1, characterized in that: The step 3) is specifically as follows: 3.1) Normalize each influencing variable to a 0-1 numerical interval, perform statistical analysis on each influencing variable, and divide the 0-1 numerical interval of the optimization variable into M non-overlapping intervals with equal probability; 3.2) Randomly select values ​​for each influencing variable in each interval to form a sampling point. Each sampling point represents a design scheme with different influencing variable combinations. Multiple sampling points are taken as matrix elements and stored in a T×M sampling point matrix Q, which can be expressed as: Q TM ={x1~x N } Among them, Q TM represents the Tth row and Mth column element of the sampling point matrix Q, x1 represents the value of the first influencing variable in the range of 0-1, and x N Indicates the value of the Nth influencing variable in the range of 0-1; 3.3) Re-arrange the elements of each row in the sampling point matrix Q randomly to obtain the Latin hypercube sample matrix W, which is expressed as: Where: W is the Latin hypercube sample matrix; represents the xth N The element in the row and column M, W NM Represents the Nth row and Mth column element of the sampling point matrix Q; 3.4) The numerical simulation method is used to calculate and solve each sampling point in the Latin hypercube sample matrix to obtain the flow-induced wear rate E and gas-solid contact efficiency C corresponding to each design scheme. E .

5. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 1, characterized in that: In step 4), the gas-solid contact efficiency C is established according to the multivariate linear regression expression. E and the proxy model of flow-induced wear rate E, respectively, are expressed as: Where: β 1,0 ~β 1,i Indicates the gas-solid contact efficiency C E The coefficients of each proxy model, β 2,0 ~β 2,i The coefficients in the proxy model representing the flow-induced wear rate E.

6. The method for designing a catalytic gasoline desulfurization reactor distribution plate against flow-induced wear according to claim 1, characterized in that: In the step 5), according to the gas-solid contact efficiency C E The proxy model of flow-induced wear rate E uses a weighted linear combination to establish the objective function: minF(x), x I ≤x≤x J F1(x)=|f1(x)-0.9612| F2(x)=f2(x) In the formula, x I and x J are the minimum and maximum values ​​of the influencing variables, respectively, w j represents the weighted value of the jth proxy model, j = 1, 2, F j (x) represents the functional expression of the jth proxy model, and F1(x) represents the gas-solid contact efficiency C E The proxy model function expression of F2(x) is the proxy model function expression of flow-induced wear rate E; f1(x) is the gas-solid contact efficiency C E f2(x) is a proxy value of flow-induced wear rate E.

7. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 1, characterized in that: The distribution plate of the catalytic gasoline desulfurization reactor is a distribution plate in the catalytic gasoline desulfurization reactor (4), adsorbent particles are stacked and arranged on the distribution plate, and the distribution plate flows through the sulfur-containing gas phase gasoline; The bottom of the catalytic gasoline desulfurization reactor (4) is provided with a material inlet (6) for the entry of sulfur-containing gaseous gasoline. A distribution plate is arranged in the tower above the material inlet (6). Through holes are evenly arranged on the distribution plate. A lifting pipe (5) is arranged and installed in each through hole. The top of the upper end of the lifting pipe (5) is connected to the bubble cap (1); the adsorbent particles are accumulated on the distribution plate.

8. The method for designing a distribution plate of a catalytic gasoline desulfurization reactor against flow-induced wear according to claim 7, characterized in that: The lower end of the bubble (1) is provided with an opening, and the upper end of the lifting tube (5) is connected to the inner top surface of the bubble (1) via support rods arranged at intervals along the circumference. An annular gap is provided between the opening at the lower end of the bubble (1) and the lifting tube (5), so that the upper end of the lifting tube (5) is connected to the inner cavity of the bubble (1) via the interval between the support rods, and then connected to the outside via the annular gap.

Images