CN117594144A - Modeling method of crude oil direct catalytic cracking process at molecular level - Google Patents

Abstract

The invention provides a modeling method of a crude oil direct catalytic cracking process at a molecular level, which comprises the following steps: s110, establishing a structure-oriented lumped model of the mixed molecular level of the raw oil based on a molecular structure reconstruction technology; s120, establishing a crude oil direct catalytic cracking process model of a molecular level corresponding to the pilot plant based on the lumped reaction dynamics model; s130, carrying out repeatability and accuracy verification on a crude oil direct catalytic cracking process model; and S140, optimizing key process parameters by using a crude oil direct catalytic cracking process model. The invention establishes a lumped dynamic model of a mixed molecular level, and establishes a crude oil direct catalytic cracking model of a molecular level by utilizing a two-section riser catalytic cracking technology, thereby researching and optimizing the relationship between key process parameters of the crude oil direct catalytic cracking, and solving the problems that the accurate crude oil direct catalytic cracking model with high solving efficiency is difficult to establish at present.

Description

A modeling method for crude oil direct catalytic cracking process at the molecular level

Technical field

The invention belongs to the technical field of petrochemical production, and in particular relates to a modeling method of direct catalytic cracking process of crude oil at the molecular level.

Background technique

In recent years, as domestic demand for fuel oil has declined, more and more oil has been diverted to the production of chemical raw materials. At present, the most representative "oil conversion" process path mainly revolves around two core technologies: steam cracking and catalytic cracking. Most new refineries use hydrocracking units to convert heavy oil into naphtha as raw material for steam cracking to produce olefins. Some refineries also deploy heavy oil catalytic cracking units to produce olefins and aromatics. But no matter which method is used, the process route is greatly extended and the investment cost and operating cost are increased. One-step conversion of crude oil (OSCO) is an emerging process route that converts crude oil into chemicals in one step through a catalytic cracking unit. While maintaining the same olefin production, its reaction temperature is more than 200°C lower than steam cracking. , energy consumption and equipment investment are significantly reduced. In the future, direct catalytic cracking of crude oil will be one of the important ways for my country to achieve its carbon neutrality goal.

The raw materials for ordinary catalytic cracking are mostly heavy oils such as vacuum wax oil. The molecules have long carbon chains and low cracking activation energy. Cracking can occur at about 540°C to generate gasoline and diesel. Compared with the reaction process of ordinary catalytic cracking, direct catalytic cracking of crude oil contains additional light hydrocarbons in the raw material, namely naphtha, kerosene, and diesel fractions. These light hydrocarbons are mainly composed of small molecules of alkanes and cycloalkanes. They are relatively stable and have high cracking activation energy. They are difficult to react under ordinary catalytic cracking reaction conditions. Therefore, how to solve the problem of cracking reaction of raw materials is one of the current bottlenecks.

Chinese patent CN1118539C discloses a two-stage riser catalytic cracking technology, which mainly uses a two-stage riser combined with a conventional catalyst to separately react fresh raw materials and circulating oil that is more likely to produce coke, thereby improving the average performance of the catalyst in the riser. and single-pass conversion rate, which provides ideas for solving the problem of crude oil direct catalytic cracking raw material cracking reaction. Through the two-stage preheating of crude oil, the light and heavy components can be well separated, so that the light components that are difficult to crack can enter the second riser separately for reaction. Through harsh cracking conditions (such as higher reaction temperature, higher agent-to-oil ratio, etc.) to improve the cracking performance of light components, while the heavy components enter the first stage riser and react under conventional reaction conditions. Since the products of direct catalytic cracking of crude oil are mostly small molecular olefins and aromatics.

Although the traditional lumped kinetic model can achieve high accuracy in product distribution, it is difficult to meet the needs of fully describing light hydrocarbon fractions, such as naphtha, kerosene and diesel. This means that if direct catalytic cracking of crude oil is used The traditional lumped kinetic model will be difficult to describe the generation process of small molecular olefins and aromatics. On the contrary, the catalytic cracking reaction kinetic model at the molecular level has a more excellent ability to predict the distribution of crude oil cracking products and calculate the properties of crude oil cracking products. However, as the fraction becomes heavier, the corresponding molecules will be geometric, which greatly improves This increases the complexity of the model and reduces the solution efficiency. How to establish a mixed molecular lumped dynamics model that has molecular-level prediction capabilities while retaining high solution efficiency is one of the current bottlenecks.

Contents of the invention

The object of the present invention is to provide a modeling method for the direct catalytic cracking process of crude oil at the molecular level. The direct catalytic cracking model of crude oil established by the modeling method in the present invention has high accuracy and high solving efficiency.

The present invention provides a modeling method for the direct catalytic cracking process of crude oil at the molecular level, which includes the following steps:

S110. Establish a structure-oriented lumped model at the mixed molecular level of raw oil based on molecular structure reconstruction technology;

The mixed molecular level is as follows: the light fraction is described using real molecules, and the heavy oil fraction is described using virtual components;

The boiling point of the light fraction is between the initial boiling point and the staged preheating cutting temperature, and the boiling point of the heavy oil fraction is between the staged preheating cutting temperature and the final boiling point;

S120. Establish a direct catalytic cracking process model of crude oil at the molecular level corresponding to the pilot plant based on the lumped reaction kinetic model;

S130. Verify the repeatability and accuracy of the crude oil direct catalytic cracking process model;

S140. Use the crude oil direct catalytic cracking process model to optimize key process parameters.

Preferably, the step S110 includes:

S111. Generate raw material oil molecules based on the chemical analysis data of the raw material oil based on the Monte Carlo sampling algorithm;

S112. Determine the fraction type and the distribution type and parameters of the probability density function of different molecules based on the properties of different groups;

S113. Generate a set of virtual molecules based on the probability density function sampling and calculate their properties;

S114. When the difference between the properties of the virtual molecule set and the analysis data results does not meet the preset conditions, the distribution function will continue to be optimized through the global optimization algorithm until the difference between the properties of the virtual molecule set and the analysis data results. The difference between satisfies the preset conditions;

S115. When the difference between the properties of the virtual molecule set and the analysis data results meets the preset conditions, the virtual molecule set is retained to obtain a structure-oriented lumped model at the mixed molecule level of the raw oil mixed fraction.

Preferably, in step S111, the chemical analysis data of the feed oil includes one or more of elemental composition, molar mass, distillation range, PONA, sulfur content and nitrogen content.

Preferably, in step S113, the properties of the virtual molecule set include density, initial distillation point, final distillation point, 10% distillation point, 30% distillation point, 50% distillation point, 70% distillation point and 90% distillation point. One or more of the % distillation points;

The preset conditions described in steps S114 and S115 are 3 to 10%.

Preferably, in step S120, the lumped reaction kinetics model adopts the twenty-one lumped reaction kinetics model.

Preferably, the step S120 specifically includes:

S121. Based on the hierarchical preheating treatment technology, the light fraction and heavy oil fraction of the feed oil are separated and entered into the corresponding reactor reaction zone for reaction;

S122. Establish a reaction model based on the direct catalytic cracking pilot test data of the feed oil;

S123. Establish a product separation model based on the product characteristics of the reaction model;

S124. When the difference between the product distribution results of the separation model and the pilot data results meets the preset conditions, retain the correction factor set;

S125. When the difference between the product distribution results of the separation model and the pilot data results does not meet the preset conditions, correct the reaction model until the difference between the product distribution results of the separation model and the pilot data results satisfies Preset conditions.

Preferably, the reactor in step S121 is a two-stage riser catalytic cracking reactor;

The separated light fraction enters the first stage riser reaction, and the heavy oil fraction enters the second stage riser reaction.

Preferably, the direct catalytic cracking pilot data of feed oil described in step S122 includes the initial working conditions adopted in the actual production process and the distribution of process products corresponding to the initial working conditions, and the process products include dry gas. , one or more of liquefied gas, gasoline, diesel, oil slurry and coke;

The preset conditions described in steps S124 and S125 are 3 to 10%.

Preferably, the calibration reaction model in step S125 specifically includes:

Adjust the reaction kinetic parameters in the direct catalytic cracking model of crude oil at the molecular level in a preset value increase or decrease manner based on the values in the empirical formula;

The preset value-added or devalued range is 0.02 to 0.08.

Preferably, in step S140, the key process parameters of the model are optimized based on the life cycle optimization strategy and the multi-objective optimization algorithm.

The present invention provides a molecular level modeling method for the direct catalytic cracking process of crude oil, which includes the following steps: S110. Establish a structure-oriented lumped model of the mixed molecule level of the raw oil based on the molecular structure reconstruction technology; the mixed molecule level is : Use real molecules to describe the light fraction, and use virtual components to describe the heavy oil fraction; the boiling point of the light fraction is between the initial boiling point and the fractional preheating cutting temperature, and the boiling point of the heavy oil fraction is between the fractional preheating temperature and the initial boiling point. between the thermal cutting temperature and the final boiling point;; S120. Based on the lumped reaction kinetic model, establish a direct catalytic cracking process model of crude oil at the molecular level corresponding to the pilot plant; S130, conduct feasibility studies on the direct catalytic cracking process model of crude oil. Repeatability and accuracy verification; S140. Use the crude oil direct catalytic cracking process model to optimize key process parameters. The present invention establishes a lumped kinetic model at the mixed molecule level based on the molecular reconstruction technology, and uses two-stage riser catalytic cracking technology to establish a molecular-level direct catalytic cracking model of crude oil, thereby determining the relationship between key process parameters of direct catalytic cracking of crude oil. The relationship is studied and optimized to solve the current problem of difficulty in establishing an accurate and efficient direct catalytic cracking model of crude oil at the mixed molecular level.

Compared with the prior art, the present invention has the following beneficial effects.

1. This invention is based on molecular reconstruction technology and uses crude oil classification preheating to establish a structure-oriented aggregation at the mixed molecular level. The naphtha, kerosene, and diesel fractions are represented by molecules, which solves the problem that traditional aggregation cannot fully describe the light hydrocarbon fractions. problem while retaining the traditional lumped form of heavy oil fractions, solving the problems of complex molecular-level lumped dynamics models and low solution efficiency.

2. The present invention establishes a crude oil direct catalytic cracking device model based on the two-stage riser catalytic cracking technology and the 21-stage collective reaction kinetics model, conducts partitioned reactions of different fractions of crude oil, and uses the model to explore the impact of different reaction conditions on product distribution in different reaction areas. It solves the time-consuming and labor-intensive problem of traditional experimental methods when exploring complex variable relationships.

3. Based on the life cycle strategy, the present invention obtains the best parameter combination of operable variables to achieve the optimization goal through automatic search of optimization algorithms and manual intervention. It is of great significance to assist engineers in operations and improve enterprise efficiency.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are only These are embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on the provided drawings without exerting creative efforts.

Figure 1 is a schematic diagram of molecular level modeling of the crude oil staged catalytic cracking process in Example 1 of the present invention;

Figure 2 is a schematic flow chart of the modeling method of the direct catalytic cracking process of crude oil at the molecular level in the present invention;

Figure 3 is a schematic flow chart of step S110 in the present invention;

Figure 4 is a schematic flowchart of step S120 of the present invention.

Detailed ways

The invention provides a modeling method for the direct catalytic cracking process of crude oil at the molecular level, which includes the following steps:

S110. Establish a structure-oriented lumped model at the mixed molecular level of raw oil based on molecular structure reconstruction technology;

The mixed molecular level is as follows: the light fraction is described using real molecules, and the heavy oil fraction is described using virtual components;

In the present invention, the boiling point of the light fraction is between the initial boiling point and the graded preheating cutting temperature, and the boiling point of the heavy oil fraction is between the graded preheating cutting temperature and the final boiling point; the division of graded preheating At different temperatures, the division ranges of light and heavy fractions are different. The light fraction is the gas phase produced in the hierarchical preheating system, covering naphtha, kerosene, and diesel fractions, and the heavy oil fraction is the liquid phase produced in the last stage of the hierarchical preheating system, which is a fraction larger than diesel; Figure 1 Example A two-stage preheating system is used, and the cutting temperatures are 210°C and 290°C, so the range of the light fraction is from the initial boiling point to 290°C, and the range of the heavy fraction is from 290°C to the final boiling point.

S120. Establish a direct catalytic cracking process model of crude oil at the molecular level corresponding to the pilot plant based on the lumped reaction kinetic model;

S130. Verify the repeatability and accuracy of the crude oil direct catalytic cracking process model;

S140. Use the crude oil direct catalytic cracking process model to optimize key process parameters.

It should be noted that, as long as there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other. In addition, unless explicitly limited or contradicted by the context, the specific steps included in the methods described in the present invention are not necessarily limited to the described order, but may be performed in any order or in parallel.

FIG. 2 shows a flowchart of a modeling method 100 for a direct catalytic cracking process of crude oil at a molecular level according to an exemplary embodiment of the present invention.

As shown in Figure 2, the execution of the modeling method 100 for the direct catalytic cracking process of crude oil at the molecular level includes the following steps:

S110. Establish a structure-oriented lumped model at the mixed molecular level based on molecular structure reconstruction technology;

S120. Establish a molecular-level crude oil direct catalytic cracking process model corresponding to the pilot plant based on the lumped kinetic model;

S130. Verify the repeatability and accuracy of the crude oil direct catalytic cracking process model; and

S140. Use the crude oil direct catalytic cracking process model to optimize key process parameters.

It should be understood that the steps shown in the molecular level crude oil direct catalytic cracking process modeling method 100 are not exclusive, and the molecular level crude oil direct catalytic cracking process modeling method 100 may also include additional steps that are not shown. steps and/or steps shown may be omitted and the scope of the invention is not limited in this respect. Steps S110 to S140 are described in detail below with reference to FIGS. 3 to 4 .

S110

For the reaction kinetics model of catalytic cracking, traditional lumped kinetics cannot fully describe light hydrocarbon fractions, such as naphtha, kerosene, and diesel. The molecular-level lumped kinetics can accurately predict the chemical properties and distribution of the products, but the model The complexity is high and the solution efficiency is low. The present invention adopts structure-oriented aggregation at the mixed molecule level, that is, lighter fractions such as naphtha and diesel are described with real molecules, while heavy oil fractions are still described according to traditional virtual components, based on product chemical properties and product distribution. The divided twenty-one integrated reaction kinetics model was used to establish a crude oil direct catalytic cracking unit model corresponding to the pilot plant.

In some implementations, step S110 includes:

S111. Generate raw oil molecules based on the chemical analysis data of the raw oil (element composition, molar mass, distillation range, etc.) based on the Monte Carlo sampling algorithm;

S112. Determine the distribution types and parameters of the probability density functions of different molecules based on the properties of different groups, thereby determining the probability of occurrence of different molecules, and then perform sampling through orthogonal analysis;

S113. Generate a set of virtual molecules based on the distribution function sampling and calculate their properties;

S114. When the difference between the properties of the virtual molecule set and the analysis data results does not meet the preset conditions, the distribution function will continue to be optimized through the global optimization algorithm until the properties of the virtual molecule set are consistent with the analysis data results. The difference between them satisfies the preset conditions;

S115. When the difference between the properties of the virtual molecule set and the analysis data results meets the preset conditions, retain the virtual molecule set to obtain a structure-oriented lumped model of the raw oil mixed fraction at the molecular level.

Specifically, in step S111, crude oil analysis data, such as elemental composition, molar mass, distillation range, PONA, sulfur content, nitrogen content, etc., are obtained through the laboratory, and molecular sampling and molecular reconstruction are performed through the Monte Carlo sampling algorithm. Finally, Generate thousands of molecules. The algorithm samples based on the probability density function of each molecule, compressing the thousands of generated molecules into hundreds while ensuring that the relevant properties are consistent with laboratory data.

In step S112, the distribution type and parameters of the probability distribution function of the molecule are initialized based on the group properties of the molecule.

In step S113, the initialized probability density function parameters, Monte Carlo sampling algorithm and orthogonal analysis are used for sampling to obtain a set of virtual molecules, and detailed properties are calculated.

In step S114, when the difference between the properties of the virtual molecule set and the analysis data results does not meet the preset conditions, it means that the initialized probability density function parameters cannot make the properties of the molecules before and after reconstruction consistent. Then the distribution function can be continuously optimized through the global optimization algorithm until the difference between the properties of the virtual molecule set and the analysis data results meets the preset conditions.

Specifically, the properties to be compared before and after molecular reconstruction can be density, initial distillation point, final distillation point, 10% distillation point, 30% distillation point, 50% distillation point, 70% distillation point, 90% distillation point. Come out and wait. As an example, the above difference satisfying the preset condition means that the difference between the two is less than the preset value, and the preset value range is approximately 3% to 10%. It should be pointed out that the above data are only for explaining the technical solution of the present invention and do not constitute a limitation on the protection scope of the present invention.

In step S115, when the difference between the properties of the virtual molecule set and the analysis data results meets the preset conditions, it means that the current probability density function parameters can make the properties of the molecules before and after reconstruction consistent. Data can be summarized to obtain structure-oriented lumped models at the molecular level.

S120

After establishing the structure-oriented lumped model at the mixed molecule level based on the molecular structure reconstruction technology in step S110, in step S120, a molecular-level crude oil direct catalytic cracking process model corresponding to the pilot plant is established based on the lumped kinetic model.

In some implementations, step S120 includes:

S121. Based on the hierarchical preheating treatment technology, different fractions of the raw oil are entered into the corresponding reactor reaction zone for reaction;

S122. Establish a reaction model based on the initial operating conditions of the direct catalytic cracking pilot test data of crude oil and the default reaction kinetic parameters;

S123. Establish a product separation model based on the product characteristics of the reaction model;

S124. When the difference between the product distribution results of the separation model and the pilot data results meets the preset conditions, retain the correction factor set;

S125. When the difference between the product distribution results of the separation model and the pilot data results does not meet the preset conditions, correct the reaction model until the difference between the product distribution results of the separation model and the pilot data results satisfies Preset conditions.

Specifically, industrial data or pilot-scale data of direct catalytic cracking of crude oil can be collected in the actual production activities of industrial units, including the initial working conditions adopted in the actual production process and the process products corresponding to the initial working conditions. distributed. Process products can include dry gas, liquefied gas, gasoline, diesel, oil slurry, coke, etc.

In step S121, the hierarchical preheating treatment technology may be a three-stage preheating treatment technology. The first stage is preheated to 180°C to separate the naphtha fraction. The second stage is preheated to 230°C to separate the kerosene fraction. The first stage is preheated to 260°C, and the diesel fraction is separated; naphtha, kerosene, and diesel fractions enter the first-stage riser reaction, and the heavy oil fraction enters the second-stage riser reaction.

In step S122, input crude oil catalytic cracking industrial data or initial operating conditions in pilot plant data into the crude oil direct catalytic cracking model, and set default kinetic parameters, that is, unseparated reaction products can be predicted.

In step S123, corresponding product separation models can be established based on the characteristics of the reaction products. For example, the products of the direct catalytic cracking process of crude oil are mostly small molecular olefins. Different product separation models such as sequential separation, pre-deethanization, and pre-depropanization can be established.

In step S124, when the difference between the distribution result of the process product predicted by the product separation model and the distribution result of the process product in the industrial data or pilot data meets the preset conditions, it means that the established crude oil is directly The catalytic cracking model can accurately predict the process products of the direct catalytic cracking reaction of crude oil. At this time, the preformed simulation working conditions can be input into the crude oil direct catalytic cracking model to predict the distribution of process products under the simulated working conditions.

As an example, the above difference satisfying the preset condition means that the difference between the two is less than the preset value. The preset value ranges from 3% to 10% depending on the different components. For example, for dry gas, the preset value is approximately 5%, for liquefied gas, the preset value is approximately 10%, for gasoline, the preset value is approximately 1%, and for diesel, the preset value is approximately 0.5 %, for oil slurry, the preset value is approximately 2%, and for coke, the preset value is approximately 3%. It should be pointed out that the above data are only for explaining the technical solution of the present invention and do not constitute a limitation on the protection scope of the present invention.

The parameters included in the preformed simulated operating conditions may be the same as the parameters included in the initial operating conditions in the industrial or pilot data. In addition, the values of the parameters in the simulated working conditions can be expanded based on the values of the parameters in the initial working conditions. Take the temperature of the upper chamber of the reactor in an industrial device as an example. In the initial working conditions in the industrial data, the temperature of the upper chamber of the reactor ranges from 500°C to 505°C. Then, the temperature of the upper chamber of the reactor in the simulated working conditions is The value range can be extended to 480℃~530℃.

In step S125, if the difference between the distribution result of the process product predicted by the product separation model and the distribution result of the process product in the industrial or pilot data does not meet the preset conditions, it means that the established crude oil directly catalyzes The cracking model is still not able to accurately predict the process products of the direct catalytic cracking reaction of crude oil. At this time, the reaction kinetic parameters can be readjusted until the difference between the distribution results of the process products predicted by the product separation model and the distribution results of the process products in the industrial or pilot data meets the preset conditions.

Specifically, when the difference between the distribution result of the process product and the distribution result of the process product in the industrial data or pilot data does not meet the preset conditions, for example, when it exceeds the above-mentioned preset value, the The crude oil direct catalytic cracking model is calibrated until the difference between the distribution results of the process products and the distribution results of the process products in the industrial big data meets the preset conditions, and then the simulation conditions are input to the crude oil direct catalytic cracking model. in the catalytic cracking model.

In some embodiments, the step of calibrating the crude oil direct catalytic cracking model in step S125 includes: adjusting reaction kinetic parameters in the crude oil direct catalytic cracking model.

As an example, in the above-mentioned crude oil direct catalytic cracking model, the values of the reaction kinetic parameters can be based on the values in the empirical formula. When the distribution results of the process products are consistent with the process products in the industrial data or pilot plant data, When the difference between the distribution results does not meet the preset conditions, it means that for the reaction kinetic parameters, the values in the empirical formula cannot fit the current industrial data or pilot data well. Therefore, it is necessary to The reaction kinetic parameters in the crude oil direct catalytic cracking model are adjusted to achieve the purpose of calibrating the crude oil direct catalytic cracking model.

In some embodiments, after adjusting the reaction kinetic parameters in a preset value increase or decrease manner based on the values in the empirical formula, the crude oil direct catalytic cracking industry data or the crude oil direct catalytic cracking model is input into the crude oil direct catalytic cracking model. The initial operating conditions in the test data are used to re-predict the distribution of process products. When the difference between the re-predicted distribution results of process products and the distribution results of process products in industrial data or pilot data still does not meet the preset conditions, the reaction kinetic parameters need to be adjusted again until the process products are re-predicted. The difference between the distribution result and the distribution result of the process product in the industrial data or pilot test data meets the preset conditions.

As an example, the preset value range of increase or decrease may be 0.02 to 0.08.

S130

After the direct catalytic cracking process model of crude oil at the molecular level corresponding to the pilot plant is established in step S120, the process model is used to verify repeatability and accuracy in step S130.

In some embodiments, operating conditions in multiple sets of crude oil direct catalytic cracking industrial data or pilot data are used, and the crude oil direct catalytic cracking process model established in step S120 is used to predict the distribution of process products, and the distribution of process products is compared with the industrial data or pilot plant data. Conduct comparative error analysis on the distribution results of process products in the test data. If the differences between the distribution results of the process products predicted under multiple sets of operating conditions and the distribution results of the process products in the industrial data or pilot test data all meet the preset conditions, it means that the process model established in step S120 has Repeatability and accuracy. If there is a difference between the distribution results of the process products predicted under multiple sets of operating conditions and the distribution results of the process products in the industrial data or pilot data and cannot meet the preset conditions, it means that the process model established in step S120 Not repeatable or accurate. Step S120 needs to be repeated until the established process model is repeatable and accurate.

S140

After verifying the repeatability and accuracy of the crude oil direct catalytic cracking process model in step S130, in step S140, the operating conditions of the industrial device or pilot plant are verified through the crude oil direct catalytic cracking process model. optimize.

In some embodiments, the crude oil direct catalytic cracking process model is used, and within the safe range of the operable variables, the best parameter combination of the operable variables that achieves the optimization goal is obtained through automatic search of the optimization algorithm and manual intervention.

In some embodiments, a super multi-objective evolutionary algorithm based on maximum vector angle selection and reference vector adaptation is used to achieve the purpose of optimizing the best parameter combination of the operable variables of the target. For example, the aggregation function in the algorithm dynamically balances the convergence and diversity of the population according to the target number and evolutionary generations, where the convergence criterion is measured by the distance between the individual and the ideal point, and the diversity criterion is measured by the distance between the individual and the reference vector. Vectors measured in angles. This can effectively improve the algorithm's ability to solve ultra-multi-objective optimization problems.

In other embodiments, a multi-objective particle swarm algorithm with adaptive angle division can also be used to achieve the purpose of optimizing the best parameter combination of the operable variables of the target. For example, first of all, the uniformity of population distribution is improved through the guidance of boundary particles in the initial stage. Then, in the target space, the angle is adaptively adjusted based on the number of particles to divide the area. Then, according to the distribution of particles in the area, the optimal particles are selected and the external archive set is maintained to maintain good population diversity and improve the coverage of the optimal solution set. In addition, for areas without particle distribution, particles in adjacent areas are used to enhance the search and promote the uniformity of the optimal solution set.

Example 1

Taking the crude oil staged catalytic cracking process as an example (Figure 1), the crude oil enters the preheating system and is heated to about 170-210°C. The gas and liquid phase products enter the first-stage flash evaporation device to achieve flash separation. The gas product (naphtha) is sent to the second riser reactor of the riser reactor unit at both ends, and the liquid product is sent to the secondary preheating system and heated to 210-290°C. After the liquid product is flashed through the second-stage flash tank, gas-liquid separation is achieved, and the gas product (part of the light diesel oil) is sent to the second-stage riser reactor. Liquid products (including heavy diesel and heavy oil) enter the first riser reactor. The molecular characterization method is used in Aspen HYSYS to determine the different molecular structures in the light fractions. The regression method is enabled and the regression parameters are set. After running, the real molecular models of naphtha and diesel can be obtained. The heavy oil is processed by the traditional characterization method of Aspen HYSYS. The distillation range is cut into narrow distillation segments, and the narrow distillation segments are treated as a virtual component for collective classification. Combined with the pilot plant data, the preheating system, reaction regeneration system and separation system were built in Aspen HYSYS, the pilot plant operating parameters were input, and the model convergence was adjusted to obtain a direct catalytic cracking model of crude oil at the molecular level.

Carry out case analysis of the model under different working conditions, compare the product distribution with the product distribution under actual working conditions, and verify the repeatability and accuracy of the model. Python programming is used to establish a multi-objective optimization algorithm program. Engineering operating parameters such as preheating temperature and reaction temperature are used as independent variables, and life cycle assessment indicators are used as dependent variables. The model is optimized with Aspen HYSYS to obtain operations that can produce the greatest life cycle benefits. parameter.

The above are only preferred embodiments of the present invention. It should be noted that those skilled in the art can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications can also be made. should be regarded as the protection scope of the present invention.

Claims (10)

1. A modeling method of a crude oil direct catalytic cracking process at a molecular level, comprising the steps of:

s110, establishing a structure-oriented lumped model of the mixed molecular level of the raw oil based on a molecular structure reconstruction technology;

the level of the mixed molecules is as follows: describing the light fraction by using real molecules, and describing the heavy oil fraction by using virtual components;

the distillation point of the light fraction is between the initial distillation point and the staged preheating cutting temperature, and the distillation point of the heavy oil fraction is between the staged preheating cutting temperature and the final distillation point;

s120, establishing a crude oil direct catalytic cracking process model of a molecular level corresponding to the pilot plant based on the lumped reaction dynamics model;

s130, carrying out repeatability and accuracy verification on the crude oil direct catalytic cracking process model;

and S140, optimizing key process parameters by using the crude oil direct catalytic cracking process model.

2. The modeling method of claim 1, wherein the step S110 includes:

s111, generating raw oil molecules according to the raw oil chemical analysis data based on a Monte Carlo sampling algorithm;

s112, determining the distribution type and parameters of probability density functions of different molecules based on the properties of different groups;

s113, sampling and generating a group of virtual component groups based on the probability density function, and calculating the properties of the virtual component groups;

s114, when the difference value between the property of the virtual molecular set and the analysis data result does not meet the preset condition, the distribution function is continuously optimized through a global optimization algorithm until the difference value between the property of the virtual molecular set and the analysis data result meets the preset condition;

and S115, when the difference between the nature of the virtual molecular set and the analysis data result meets the preset condition, retaining the virtual molecular set to obtain a structure-oriented lumped model of the mixed molecular level of the raw oil mixed fraction.

3. The modeling method as claimed in claim 2, wherein in the step S111, the raw oil chemical analysis data includes one or more of elemental composition, molar mass, distillation range, PONA, sulfur content and nitrogen content.

4. A modeling method according to claim 3, wherein in the step S113, the properties of the virtual component group include one or more of density, initial point, final point, 10% point, 30% point, 50% point, 70% point, and 90% point;

the preset conditions in step S114 and step S115 are 3 to 10%.

5. The modeling method of claim 1, wherein in the step S120, twenty-one lumped reaction dynamics model is used as the lumped reaction dynamics model.

6. The modeling method as defined in claim 1, wherein the step S120 specifically includes:

s121, separating the light fraction and the heavy oil fraction of the raw oil based on a staged preheating treatment technology, and feeding the light fraction and the heavy oil fraction into a corresponding reactor reaction area for reaction;

s122, establishing a reaction model based on direct catalytic cracking pilot-scale data of the raw oil;

s123, establishing a product separation model based on the product characteristics of the reaction model;

s124, when the difference value between the product distribution result of the separation model and the pilot scale data result meets a preset condition, reserving a correction factor set;

and S125, correcting the reaction model when the difference value between the product distribution result of the separation model and the pilot test data result does not meet the preset condition, until the difference value between the product distribution result of the separation model and the pilot test data result meets the preset condition.

7. The modeling method as defined in claim 6, wherein the reactor in step S121 is a two-stage riser catalytic cracking reactor;

the separated light fraction enters a first section of riser for reaction, and the heavy oil fraction enters a second section of riser for reaction.

8. The modeling method as claimed in claim 7, wherein the direct catalytic cracking pilot-scale data of the raw oil in step S122 includes an initial condition adopted in an actual production process and a distribution of a process product corresponding to the initial condition, the process product including one or more of dry gas, liquefied gas, gasoline, diesel, slurry oil and coke;

the preset conditions in step S124 and step S125 are 3 to 10%.

9. The modeling method as defined in claim 8, wherein the correction of the positive response model in step S125 specifically includes:

based on the value in the empirical formula, adjusting the reaction kinetic parameters in the crude oil direct catalytic cracking model of the molecular level in a preset value increasing or decreasing mode;

the preset increment or decrement range is 0.02-0.08.

10. The modeling method of claim 1, wherein in step S140, the model key process parameters are optimized based on a life cycle optimization strategy and a multi-objective optimization algorithm.