Refining production monitoring control method and system based on three-dimensional modeling optimization

Abstract

The invention discloses a refining production monitoring control method and system based on three-dimensional modeling optimization, and relates to the technical field of refining industry monitoring control, and the method comprises the steps: S1, constructing a three-dimensional real-time state field, S2, reconstructing a three-dimensional space distribution field, S3, carrying out three-dimensional dynamic evolution simulation, S4, recognizing three-dimensional features and carrying out abnormity early warning, and S5, generating and verifying a control strategy. A three-dimensional real-time state field is generated by collecting multi-dimensional time sequence data in real time in a mapping mode, a three-dimensional space distribution field with physical consistency and high precision is obtained through reconstruction, a fluid dynamic model is input to simulate the future evolution trend, abnormal areas are recognized based on gradient and uniformity rules, an early warning priority ranking table is generated, and the early warning effect is achieved. And a targeted collaborative optimization control strategy is generated in combination with an abnormal physical field type and a process mechanism rule, and a closed-loop verification effect is performed, so that the problems of traditional monitoring information splitting, control lag and strategy conflict are solved, and the operation safety and efficiency of the catalytic cracking device are improved.

Description

Technical Field

[0001] This invention relates to the field of monitoring and control technology in the refining and chemical industry, specifically to a refining and chemical production monitoring and control method and system based on three-dimensional modeling optimization. Background Technology

[0002] In the development of refining and chemical production monitoring and control, petroleum refining enterprises have gradually built relevant information technology platforms, improving the efficiency of production operation and management. Initially, 3D modeling technology was used to build static models of refining and chemical production scenarios, clearly presenting the layout and basic operating status of petroleum refining units. With the optimization of modeling technology, it has achieved docking with refining and chemical production process data, developing dynamic 3D models that can be updated in real time, which can synchronously reflect the real-time operating status of petroleum refining and chemical production. Subsequently, the optimization direction of modeling has continued to expand, the model detail fidelity and data response speed have been continuously improved, and the integration with the petroleum refining monitoring and control links has gradually deepened. Therefore, it is necessary to study a refining and chemical production monitoring and control method and system based on 3D modeling optimization.

[0003] Existing technologies, such as the invention patent application with announcement number CN105700517B, disclose an adaptive data-driven early fault monitoring method and device for refining and chemical processes. The method involves: extracting historical data of monitoring parameters of refining and chemical equipment, updating training data according to the sliding window step size, determining the SPE control limit through standardization and principal component analysis, comparing the SPE of the sample to be tested, and judging anomalies if the limit is exceeded, thus solving the problem of false alarms in static models. Existing technologies, such as the invention patent application with announcement number CN113720788B, disclose a method, device, system, and infrared imaging device for monitoring leaks in refining and chemical production. The method includes: controlling the infrared imaging device to rotate according to the control angle data, judging the type of the leak target area, acquiring images in the corresponding infrared band, and judging the leak. It can perform real-time roving monitoring, is anti-interference, has a fast response, and is highly efficient.

[0004] As can be seen from the above solutions, the current monitoring and control methods for refining and chemical production still have significant shortcomings. First, traditional monitoring methods rely on isolated point sensors, which make it difficult to characterize the real interaction and dynamic evolution of key parameters such as temperature, pressure, and catalyst distribution in three-dimensional space, resulting in difficulties in locating the root cause of anomalies. Second, existing control strategies are mostly reactive and lack the ability to predict the future operating trends of the unit, making it impossible to provide early warnings and proactive intervention for the evolution of abnormal states. Third, when multiple process parameters need to be adjusted, there is a lack of coordination mechanisms between control commands, which can easily lead to strategy conflicts and make it difficult to achieve precise control of overall optimal performance. This restricts further improvement in the safety, stability, and efficiency of catalytic cracking units. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the present invention aims to provide a refining and chemical production monitoring and control method and system based on three-dimensional modeling optimization.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The first aspect of the present invention provides a refining and chemical production monitoring and control method based on three-dimensional modeling optimization, including S1. Constructing a three-dimensional real-time state field: real-time acquisition of time-series data of temperature, pressure, catalyst reserves and key component concentrations in the reactor and regenerator of the catalytic cracking unit during the refining and chemical production process, and mapping them to a three-dimensional spatial grid model of the catalytic cracking unit to generate a three-dimensional real-time state field inside the catalytic cracking unit.

[0007] S2. Reconstructing the three-dimensional spatial distribution field: Spatial smoothing and physical consistency correction are performed on the three-dimensional real-time state field to reconstruct the three-dimensional spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator.

[0008] S3. Three-dimensional dynamic evolution simulation: The three-dimensional spatial distribution field is input into a preset fluid dynamics simulation model, and the three-dimensional dynamic evolution field of the catalyst fluidization state and reaction progress state in the future period is output.

[0009] S4. Identify 3D features and issue early warnings: Based on the gradient and uniformity judgment rules of the 3D dynamic evolution field, spatial feature identification is performed on the 3D dynamic evolution field and an early warning priority ranking table is generated.

[0010] S5. Control Strategy Generation and Verification: Based on the judgment result of S4, an optimized control strategy for the catalytic cracking unit is generated and executed. S1 to S3 are executed sequentially to obtain the three-dimensional dynamic evolution field after the control strategy is executed. The control effect is verified by comparing the three-dimensional dynamic evolution field before and after the control strategy is executed.

[0011] A second aspect of the present invention provides a system for a refining and chemical production monitoring and control method based on three-dimensional modeling optimization, comprising constructing a three-dimensional real-time state field module: used to acquire in real time the time-series data of temperature, pressure, catalyst reserves and key component concentrations in the reactor and regenerator of the catalytic cracking unit during the refining and chemical production process, and map them to a three-dimensional spatial grid model of the catalytic cracking unit to generate a three-dimensional real-time state field inside the catalytic cracking unit.

[0012] The 3D spatial distribution field reconstruction module is used to perform spatial smoothing and physical consistency correction on the 3D real-time state field, and reconstruct the 3D spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator.

[0013] Three-dimensional dynamic evolution simulation module: used to input the three-dimensional spatial distribution field into a preset fluid dynamics simulation model, and output the three-dimensional dynamic evolution field of the catalyst fluidization state and reaction progress state in the future period.

[0014] The 3D feature identification and anomaly warning module: Based on the gradient and uniformity judgment rules of the 3D dynamic evolution field, it performs spatial feature identification of the 3D dynamic evolution field and generates a warning priority ranking table.

[0015] Control strategy generation and verification module: Based on the judgment result of S4, it generates and executes the optimized control strategy of the catalytic cracking unit, and sequentially executes S1 to S3 to obtain the three-dimensional dynamic evolution field after the control strategy is executed. The control effect is verified by comparing the three-dimensional dynamic evolution field before and after the control strategy is executed.

[0016] The beneficial effects of the present invention are as follows: (1) The first part of the present invention: by constructing a high-fidelity three-dimensional real-time state field, the continuous and integrated characterization of multiple physical quantities such as temperature, pressure and catalyst distribution inside the catalytic cracking unit in three-dimensional space is realized for the first time. It can reveal the real fluidization state and reaction environment inside the reactor and regenerator in a panoramic way, providing an unprecedented complete and spatial data foundation for subsequent accurate diagnosis and optimization control, and solving the problem of information fragmentation from the source.

[0017] (2) The second part of the present invention: not only does it generate a physically reasonable three-dimensional spatial distribution field, but it can also dynamically predict the evolution trend of catalyst fluidization and reaction in future cycles, so that the system can leap from traditional post-event perception to pre-event prediction. It can detect potential risks such as local overheating and uneven fluidization in advance, significantly enhancing the early risk insight and operational safety margin of the catalytic cracking unit, and laying a solid foundation for the scientific and timely nature of optimization decision-making.

[0018] (3) The third part of the present invention: Based on the intelligent identification and priority ranking of the three-dimensional dynamic evolution field, a control strategy that accurately matches the abnormal space features can be generated. Its core advantage lies in its ability to perform multi-objective collaborative optimization. When faced with the coexistence of multiple abnormal fields, the system can perform strategy disambiguation and collaboration based on the process mechanism to avoid control command conflicts. Finally, the control effect is ensured through closed-loop verification, realizing the leap from single-point passive control to global active and intelligent optimization, and greatly improving the overall operating efficiency of the device. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0021] Figure 2 This is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Reference Figure 1 As shown, the present invention provides a refining and chemical production monitoring and control method based on three-dimensional modeling optimization, including S1. Constructing a three-dimensional real-time state field: real-time acquisition of time-series data of temperature, pressure, catalyst reserves and key component concentrations in the reactor and regenerator of the catalytic cracking unit during the refining and chemical production process, and mapping them to a three-dimensional spatial grid model of the catalytic cracking unit to generate a three-dimensional real-time state field inside the catalytic cracking unit.

[0024] The catalyst stockpile refers to the amount of catalyst held within the reactor and regenerator.

[0025] The key component concentrations include the concentrations of C4 and below light hydrocarbon components in the reactor, the concentrations of gasoline fraction marker components, and the concentrations of components characterizing coking tendency, as well as the oxygen, carbon monoxide, and carbon dioxide concentrations in the regenerator.

[0026] For example, the C4 and below light hydrocarbon components include propylene, butene, etc., the gasoline fraction marker components include isopentane, aromatics, etc., and the components characterizing coking tendency include polycyclic aromatic hydrocarbons, gums, etc.

[0027] It should be noted that the time-series data of temperature, pressure, and catalyst reserves are collected in real time from the sensor network deployed in the reactors and regenerators of the catalytic cracking unit in the core production of petroleum refining, and the time-series data of key component concentrations detected by online analyzers installed on the flue gas outlet pipeline of the regenerator and the product pipeline of the reactor are acquired simultaneously.

[0028] In a specific embodiment of the present invention, the method for generating the three-dimensional real-time state field inside the catalytic cracking unit is as follows: the time-series data of temperature, pressure, catalyst reserves and key component concentrations in the reactor and regenerator of the catalytic cracking unit are obtained and mapped to the corresponding spatial nodes of the three-dimensional spatial grid model of the catalytic cracking unit according to the spatial coordinates of the physical installation positions of the corresponding sensors. The entire domain of the three-dimensional spatial grid model is then extrapolated using the Kriging space interpolation algorithm to generate a continuously distributed temperature field, pressure field, catalyst density field and key component concentration field. These fields are then integrated into the three-dimensional spatial grid model to generate a three-dimensional real-time state field characterizing the real-time operating conditions inside the catalytic cracking unit.

[0029] It should be noted that the Kriging spatial interpolation algorithm is a standard method based on regionalized variables and spatial correlation principles for optimally extrapolating a continuous spatial field from discrete observation point data. The algorithm itself is a mature existing technology in the fields of spatial statistics and geological engineering. Its core lies in optimally extrapolating the continuous distribution of the entire three-dimensional spatial field based on the known spatial distribution structure and correlation of sample points, thereby providing technical support for the generation of a high-fidelity three-dimensional real-time state field in this invention.

[0030] It should be noted that the pre-constructed three-dimensional spatial mesh model of the catalytic cracking unit is a digital model that accurately reflects the internal geometry and spatial dimensions of key equipment such as reactors and regenerators, established by three-dimensional laser scanning of the point cloud data of the catalytic cracking unit. This model was constructed before the implementation of the monitoring and control methods, and its construction process itself adopts mature three-dimensional modeling and mesh generation technology in this field.

[0031] The three-dimensional real-time state field provides the initial spatial distribution for subsequent simulation steps.

[0032] S2. Reconstructing the three-dimensional spatial distribution field: Spatial smoothing and physical consistency correction are performed on the three-dimensional real-time state field to reconstruct the three-dimensional spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator.

[0033] In a specific embodiment of the present invention, the method for spatial smoothing and physical consistency correction of the three-dimensional real-time state field is as follows: the data of each spatial node in the three-dimensional real-time state field is smoothed by using a Gaussian filtering algorithm, and physical constraints are constructed based on the mass and energy conservation relationship of the catalytic cracking device to correct the physical consistency of the data of each spatial node in the smoothed three-dimensional real-time state field.

[0034] It should be noted that smoothing the data of each spatial node in the three-dimensional real-time state field is to suppress local data fluctuations caused by sensor noise. Correcting the data of each spatial node in the smoothed three-dimensional real-time state field is to ensure that the catalyst density field conforms to the law of mass conservation, and that the temperature field and the concentration field of key components conform to the law of energy conservation and material balance, thereby generating a more physically reasonable three-dimensional spatial distribution field.

[0035] In one specific embodiment, physical constraints are constructed based on the mass and energy conservation relationship of the catalytic cracking unit. The specific construction method is as follows: for the control volume of each spatial node in the three-dimensional spatial grid model of the reactor and regenerator, the balance relationship of the catalyst mass flow rate flowing into and out of the spatial node is defined as the mass conservation constraint of the catalyst density field, the balance relationship between the reaction heat effect and the heat inflow and outflow is defined as the energy conservation constraint of the temperature field, and the balance relationship between the generation and consumption rate of key components is defined as the component material balance constraint of the concentration field. These algebraic relationships based on physical principles are used to form a set of equations or inequalities that must be satisfied when correcting the corresponding field data.

[0036] In a specific embodiment of the present invention, the method for reconstructing the three-dimensional spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator is as follows: the three-dimensional real-time state field after spatial smoothing and physical consistency correction is used as the basic data source, the spatial node grid of the catalyst density, temperature, pressure and key component concentration data is refined and interpolated using a cubic spline interpolation algorithm, and the optimal three-dimensional spatial distribution that satisfies all physical constraints is solved by a mathematical optimization algorithm to generate a three-dimensional spatial distribution field with physical consistency.

[0037] It should be noted that cubic spline interpolation is a mature mathematical tool that connects known data points by constructing a piecewise cubic polynomial function and ensuring that the function curve has continuous first and second derivatives at the connection points, thereby achieving smoothness and continuity of the interpolation results. This algorithm is a well-known existing technology in the fields of computational mathematics and numerical analysis. This invention applies it to the refinement interpolation of three-dimensional spatial meshes, aiming to utilize its excellent smoothness and proximity characteristics to obtain a high-quality data field that better conforms to the continuous change law of the physical field. This algorithm is mature and will not be described in detail here.

[0038] It should be noted that mathematical optimization algorithms are a class of general computational methods that find the optimal solution of a specific objective function through iterative calculation. They encompass a variety of mature algorithms such as gradient descent, conjugate gradient, and Lagrange multiplier method. These algorithms are standard existing technologies in fields such as operations research and computational mathematics. This invention applies them to solve for three-dimensional spatial distributions that satisfy physical constraints, aiming to transform physical conservation relationships into mathematical constraints and utilize the optimization capabilities of the algorithm to obtain the physically most reasonable spatial field distribution.

[0039] S3. Three-dimensional dynamic evolution simulation: The three-dimensional spatial distribution field is input into a preset fluid dynamics simulation model, and the three-dimensional dynamic evolution field of the catalyst fluidization state and reaction progress state in the future period is output.

[0040] In a specific embodiment of the present invention, the output of the three-dimensional dynamic evolution field of the catalyst fluidization state and the reaction progress state in the future period is as follows: a three-dimensional spatial distribution field with physical consistency is input as the initial condition to a preset fluid dynamics simulation model. By solving the unsteady-state numerical calculation of the mass, momentum, energy and component conservation equations, the three-dimensional dynamic evolution field presented in a discrete time step sequence in the future period is output. The three-dimensional dynamic evolution field includes the three-dimensional dynamic evolution field of the catalyst fluidization state and the three-dimensional dynamic evolution field of the reaction progress state. The three-dimensional dynamic evolution field of the catalyst fluidization state includes the catalyst particle velocity vector field, the catalyst volume fraction distribution field and the gas-solid interphase force field. The three-dimensional dynamic evolution field of the reaction progress state includes the temperature field, pressure field, key component concentration field and chemical reaction rate distribution field of the spatial nodes.

[0041] It should be noted that the three-dimensional spatial distribution field not only possesses physical consistency but also high precision. The high precision of the three-dimensional spatial distribution field stems from multiple safeguards in its construction process. Its data foundation is a three-dimensional real-time state field that has undergone spatial smoothing and physical consistency correction. This correction process has effectively suppressed sensor noise and eliminated unreasonable physical fluctuations. The cubic spline interpolation algorithm used has excellent smoothness and continuity approximation characteristics, which can retain the true characteristics of the original data to the maximum extent while refining the mesh, avoiding additional errors introduced by improper interpolation methods. Furthermore, the optimal distribution that satisfies all physical constraints is solved through mathematical optimization algorithms. This process essentially deeply couples experimental observation data with physical mechanisms, making the final three-dimensional spatial distribution field not only mathematically smooth and continuous and physically realistic and reliable, but also achieving high precision in approximating the true state inside the device, providing a reliable initial field for subsequent dynamic evolution simulation.

[0042] It should be noted that the three-dimensional dynamic evolution field of catalyst fluidization state is used to characterize the motion and spatial distribution of catalyst particles. The catalyst particle motion velocity vector field and catalyst volume fraction distribution field comprehensively characterize the catalyst's circulation dynamics, fluidization quality, and gas-solid two-phase contact efficiency, providing a direct basis for predicting and optimizing the system's hydrodynamic state. The three-dimensional dynamic evolution field of reaction progress state is used to characterize the progress, intensity, thermal effect, and hydrodynamic stability of the chemical reaction. The changes in the key component concentration field reflect the spatiotemporal distribution of reactant consumption and product generation in real time to quantify the degree of reaction progress. The reaction rate field reveals the instantaneous chemical reaction intensity at each spatial node. The temperature field dynamically characterizes the generation, transfer, and distribution of reaction heat. The pressure field characterizes the system pressure drop and flow stability, providing a basis for evaluating reactor transfer efficiency and equipment safety status. The combination of these four elements completely reproduces the reaction intensity, conversion efficiency, and thermodynamic evolution of cracking and regeneration reactions in three-dimensional space, laying the foundation for achieving precise control based on reaction mechanisms.

[0043] It should be noted that the pre-set fluid dynamics simulation model is a numerical model built on the principles of computational fluid dynamics and chemical reaction engineering, capable of coupling gas-solid two-phase flow dynamics with catalytic cracking and coke combustion reaction dynamics. This model simulates the physicochemical processes inside the device by solving a set of partial differential equations that conserve mass, momentum, energy, and chemical composition. The construction principle and method of such models are known existing technologies in the relevant technical field. Those skilled in the art can flexibly configure them based on conventional experience and specific scenarios without affecting the implementation and reproduction of the technical solution of this invention. This invention uses it as the core prediction engine and dynamically couples it with the high-precision three-dimensional spatial distribution field generated in the aforementioned steps to achieve high-fidelity prediction of future operating conditions in a closed-loop monitoring and control system.

[0044] S4. Identify 3D features and issue early warnings: Based on the gradient and uniformity judgment rules of the 3D dynamic evolution field, spatial feature identification is performed on the 3D dynamic evolution field and an early warning priority ranking table is generated.

[0045] In a specific embodiment of the present invention, the gradient and uniformity determination rules of the three-dimensional dynamic evolution field include gradient anomaly determination conditions: for the scalar field in the three-dimensional dynamic evolution field, the scalar field includes a temperature field, a pressure field, a key component concentration field, and a chemical reaction rate distribution field. If, within a connected spatial region of any of the scalar physical fields, the average value of the spatial gradient modulus of the physical quantity calculated based on all spatial grid nodes within that region is greater than the preset gradient threshold of the physical quantity within a target time period, then it is determined that there is a gradient anomaly in that region.

[0046] Uniformity anomaly determination criteria: For the vector physical field in the three-dimensional dynamic evolution field, the vector physical field includes the catalyst particle velocity vector field, the catalyst volume fraction distribution field and the gas-solid phase interaction force field. Within the preset three-dimensional spatial analysis area, if the coefficient of variation of the vector magnitude of all spatial grid nodes of any vector physical field is greater than the preset uniformity threshold of the vector field within the target time period, then the analysis area is determined to have a uniformity anomaly.

[0047] It should be noted that the connected spatial region refers to a continuous three-dimensional spatial subset formed by grid nodes in the three-dimensional dynamic evolution field whose abnormal physical quantity values ​​are greater than the corresponding warning thresholds stored in the local database and which are connected to each other in space through adjacency. The preset three-dimensional spatial analysis region refers to a three-dimensional spatial range predefined based on the process structure and monitoring requirements of the catalytic cracking unit, including but not limited to the dense phase bed region of the regenerator, the reaction section of the reactor riser, and the inlet region of the cyclone separator group.

[0048] It should be noted that the key component concentration fields include the oxygen concentration field, carbon monoxide concentration field, and carbon dioxide concentration field in the regenerator; as well as the hydrogen concentration field, C4 and below light hydrocarbon component concentration field, gasoline fraction marker component concentration field, and component concentration field characterizing coking tendency in the reactor.

[0049] It should be noted that the calculation is based on the data structure of the three-dimensional dynamic evolution field itself and is implemented using a well-known mathematical method. The average value of the spatial gradient modulus is obtained by calculating the spatial gradient vector of the physical quantity at each spatial grid node using the finite difference method and taking the modulus, and then arithmetically averaging the modulus values ​​of all nodes in the connected region. The coefficient of variation of the vector modulus is obtained by calculating the ratio of the standard deviation to the average value of the physical quantity values ​​of all spatial nodes in any vector physical field within the preset three-dimensional spatial analysis region. The finite difference method and the calculation of the coefficient of variation are standard existing techniques in the fields of numerical analysis and statistics, and will not be described in detail here.

[0050] It should be noted that gradient anomaly detection is mainly applied to temperature field, pressure field, key component concentration field, and chemical reaction rate field. The spatial gradient of these scalar fields increases dramatically, directly and sensitively indicating anomalies that require immediate attention, such as local overheating, reactant concentration fronts, or sudden changes in the reaction zone. Uniformity anomaly detection is mainly applied to catalyst volume fraction distribution field, catalyst velocity field, and interphase force field. The uniformity of these fields is key to evaluating the overall fluidization quality, gas-solid contact efficiency, and flow stability of the system. By calculating the coefficient of variation of the vector modulus in a specific process region, the uniformity of distribution can be effectively quantified, and fluidization faults such as channeling and flow deviation can be diagnosed.

[0051] In a specific embodiment of the present invention, the method for spatial feature identification and early warning priority ranking table generation of the three-dimensional dynamic evolution field is as follows: extract gradient anomaly regions and uniformity anomaly regions identified based on gradient anomaly judgment conditions and uniformity anomaly judgment conditions from the three-dimensional dynamic evolution field, and identify their three-dimensional spatial location, spatial volume, the types of abnormal physical fields contained therein, the proportion of each type of abnormal physical field in the corresponding anomaly region, and the intensity of their abnormal physical quantities.

[0052] The proportion of each abnormal physical field type in the corresponding abnormal region and the intensity of its abnormal physical quantity are multiplied together to form an early warning index. Based on the early warning index, an early warning priority ranking table is generated for all identified abnormal physical fields.

[0053] It should be noted that for gradient anomaly regions, the intensity of anomaly physical quantities is equal to the average value of the spatial gradient modulus of the physical quantities in the anomalous scalar physical field calculated on all spatial grid nodes within the connected region. For homogeneity anomaly regions, it is equal to the coefficient of variation of the physical quantity values ​​on all nodes in the anomalous vector physical field identified within the analysis region.

[0054] It should be noted that the higher the warning index, the higher the priority of the corresponding abnormal physical field warning.

[0055] S5. Control Strategy Generation and Verification: Based on the judgment result of S4, an optimized control strategy for the catalytic cracking unit is generated and executed. S1 to S3 are executed sequentially to obtain the three-dimensional dynamic evolution field after the control strategy is executed. The control effect is verified by comparing the three-dimensional dynamic evolution field before and after the control strategy is executed.

[0056] In a specific embodiment of the present invention, the method for generating and executing the optimized control strategy of the catalytic cracking unit is as follows: generating a corresponding optimized control strategy based on the identified abnormal physical field type.

[0057] When the identified abnormal physical field type is the catalyst particle velocity field or the catalyst volume fraction distribution field, the direction and amplitude of each operating parameter of the catalyst circulation system are comprehensively adjusted according to the deviation angle between the catalyst particle movement direction and the preset fluidization direction, the deviation value of the average movement amplitude, and the distribution range.

[0058] When the identified abnormal physical field type is a temperature field or a chemical reaction rate distribution field, the amplitude of each operating parameter in the reactor is adjusted in a coordinated manner according to the high temperature point distribution of the temperature field and the intensity of the abnormal physical quantities of the reaction rate distribution field.

[0059] When the identified abnormal physical field type is a pressure field, the pressure balance parameters of the reactor and regenerator or the pressure setting of the fluidized medium system are adjusted according to the distribution of high-pressure or low-pressure points and their gradients.

[0060] When the identified abnormal physical field type is a gas-solid phase interaction force field, the fluidized medium system's flow velocity and pressure parameters are adjusted according to the abnormal amplitude deviation and spatial gradient distribution of the force field.

[0061] When the identified abnormal physical field type is a key component concentration field, its specific key components are obtained to determine the direction of control operation.

[0062] When multiple abnormal physical fields exist simultaneously in a catalytic cracking unit, the control operation direction corresponding to the highest priority abnormal physical field is executed first based on the early warning priority ranking table. The control operation directions of the remaining abnormal physical fields are then coordinated and adjusted according to the process mechanism rules to generate a conflict-free optimized control strategy, which is then sent to the field actuators for execution through the distributed control system.

[0063] The fluidizing medium refers to the gas used to fluidize catalyst particles in a catalytic cracking unit. In a regenerator, it mainly refers to the air used for coking. In a reactor, it mainly refers to the feedstock oil and gas and the steam used for lifting and atomization. The fluidizing medium system refers to the process equipment system that provides, transports, distributes and regulates the above-mentioned fluidizing medium, and usually includes a main blower, a booster compressor, a fluidizing air distribution network, a steam network, and corresponding regulating valves and control systems.

[0064] It should be noted that process mechanism rules refer to a set of logical judgment and decision rules constructed based on the inherent physical and chemical principles and specific process knowledge within the catalytic cracking unit. Its core function is to ensure that when the system needs to deal with multiple interrelated abnormal physical fields, the multiple control operation commands generated are physically feasible and process-coordinated, rather than contradictory or canceling each other, thereby generating a conflict-free optimized control strategy.

[0065] In one specific embodiment, the control operation direction is determined based on the specific key components obtained from the key component concentration field. The specific method is as follows: when the abnormal component is a hydrogen concentration field, the hydrogen-to-carbon ratio of the reactor feed is adjusted based on the deviation direction and value of its concentration from the target value; when the abnormal component is a light hydrocarbon component concentration field of C4 and below, the reactor feed preheating temperature or catalyst-to-oil ratio is adjusted based on the concentration deviation value; when the abnormal component is a gasoline fraction indicator component concentration field, the reactor reaction depth or catalyst activity is adjusted based on the concentration deviation value; when the abnormal component is a component concentration field characterizing coking tendency, the terminator injection amount or quench medium flow rate is adjusted based on the concentration deviation value from the warning value; when the abnormal component is an oxygen concentration field, the total air supply of the regenerator is adjusted based on the concentration deviation direction and value; when the abnormal component is a carbon monoxide concentration field, the combustion medium flow rate or distribution is adjusted based on the concentration deviation value; and when the abnormal component is a carbon dioxide concentration field, the regenerator coking operation intensity is adjusted based on the concentration deviation value.

[0066] For example, when the anomalous component of the critical component concentration field is the hydrogen concentration field, if the hydrogen concentration field is too high, it indicates that hydrogen transfer or cracking reaction is excessive, and the hydrogen-to-carbon ratio of the reactor feed or the reaction severity should be reduced. When the anomalous component of the critical component concentration field is the concentration field of light hydrocarbon components of C4 and below, if the concentration of light hydrocarbon components is too low, it indicates that the reaction depth is insufficient, and the feedstock preheating temperature should be increased or the catalyst-to-oil ratio should be reduced. When the anomalous component of the critical component concentration field is the concentration field of gasoline fraction indicator components, if the gasoline fraction concentration is too low, the reaction temperature or catalyst activity should be increased. When the anomalous component of the critical component concentration field is the concentration field of components characterizing coking tendency. If the concentration of the coking tendency component is too high, the flow rate of the terminator or quenching medium should be increased to stop the excessive reaction. If the abnormal component of the key component concentration field is the oxygen concentration field, and the oxygen concentration in the regenerator is too high, it indicates that the air supply is excessive and the main air volume should be reduced. If the abnormal component of the key component concentration field is the carbon monoxide concentration field, and the carbon monoxide concentration is too high, it indicates that the combustion is incomplete and the amount of carbon monoxide oxidizer injected should be increased to promote its conversion into carbon dioxide. If the abnormal component of the key component concentration field is the carbon dioxide concentration field, and the carbon dioxide concentration is too low, it indicates that the coking intensity is insufficient and the regeneration temperature or the main air volume should be increased to enhance coking.

[0067] For example, when the identified abnormal physical field type is the catalyst particle motion velocity field or the catalyst volume fraction distribution field, if it is detected that in a region occupying 30% of the total area in the upper part of the riser, the catalyst volume fraction distribution non-uniformity index reaches 0.25, which is greater than the non-uniformity index threshold of 0.15, and the average deviation angle between the particle motion velocity vector and the axial fluidization direction in this region reaches 15°, then the catalyst circulation amount from the regenerator to the reactor is increased by 5% and the catalyst circulation amount from the reactor to the regenerator is decreased by 3% to balance the catalyst circulation amount distribution.

[0068] For example, when the identified abnormal physical field type is a temperature field or a chemical reaction rate distribution field, if the temperature of a local high-temperature point at the bottom of the reactor reaches 520°C, which is higher than the normal high-temperature point of 500°C, and the abnormal intensity of the chemical reaction rate in that area exceeds the standard by 30%, then the feed flow rate will be reduced by 8% and the feed preheating temperature will be increased by 10°C to balance the reaction intensity distribution.

[0069] For example, when the identified abnormal physical field type is a pressure field, if an abnormal pressure gradient is detected in a local area of ​​the dense phase bed of the regenerator, and the pressure deviation between the high-pressure point and the average pressure reaches 5 kPa and the spatial range of the high-pressure area accounts for 15% of the cross-section of the dense phase bed, then the control operation direction is to adjust the pressure setting of the fluidizing medium system, and increase the opening of the main air distribution ring pipe branch valve of the area corresponding to the high-pressure point by 3% to clear the fluidizing channel and balance the bed pressure distribution.

[0070] For example, when an abnormal force field between the gas and solid phases is detected, if the abnormal force field deviation of the dense phase bed in the regenerator is detected to be 15% and the spatial gradient distribution is uneven, the outlet pressure of the main fan will be increased by 5 kPa and the pressure drop setting of the distribution plate will be adjusted by 3 kPa to optimize the fluidization state.

[0071] For example, when multiple abnormal physical fields exist simultaneously, they are processed in sequence according to the warning priority sorting table: if the oxygen concentration field with the highest priority is abnormal and the main air volume needs to be reduced by 6%, while the temperature field with the second highest priority is abnormal and the feed temperature needs to be increased by 5°C, then the system will prioritize the reduction of air volume, and adjust the temperature increase operation to 3% to achieve coordinated control, provided that the temperature abnormality is not aggravated.

[0072] In a specific embodiment of the present invention, the method for verifying the control effect by comparing the three-dimensional dynamic evolution field before and after the execution of the control strategy is as follows: after the control strategy is executed, return to S1 to obtain a new round of time series data, reconstruct the three-dimensional spatial distribution field after execution, and further execute S3, using the three-dimensional spatial distribution field after execution as the initial condition input to the preset fluid dynamics simulation model, and output the three-dimensional dynamic evolution field after execution.

[0073] The three-dimensional dynamic evolution field after the control strategy is executed is compared with the three-dimensional dynamic evolution field before execution. For gradient anomaly regions and uniformity anomaly regions identified in the three-dimensional dynamic evolution field before execution, their status in the three-dimensional dynamic evolution field after execution is checked. If the average value of the spatial gradient modulus of physical quantities in the gradient anomaly region is less than the preset gradient threshold and the coefficient of variation of physical quantities in the uniformity anomaly region is less than the preset uniformity threshold, the control strategy is deemed effective; otherwise, the control strategy is deemed ineffective.

[0074] It should be noted that the three-dimensional dynamic evolution field is the future operating condition evolution trajectory predicted by the current three-dimensional spatial distribution field and the physical mechanism model. Directly comparing the three-dimensional spatial distribution field before and after execution can only evaluate the immediate static effect of the control strategy. However, comparing the three-dimensional dynamic evolution field before and after execution can evaluate the impact of the control strategy on the future dynamic behavior of the system. If the three-dimensional dynamic evolution field after execution shows that the original abnormal trend has disappeared or has been significantly improved, it indicates that the control strategy has not only corrected the current state, but also guided the system into a safer, more stable and more optimized operating trajectory in the future cycle, thus achieving a more forward-looking control effect verification.

[0075] Reference Figure 2As shown, the present invention provides a system for a refining and chemical production monitoring and control method based on three-dimensional modeling optimization, including a three-dimensional real-time state field module: used to acquire in real time the time-series data of temperature, pressure, catalyst reserves and key component concentrations in the reactor and regenerator of the catalytic cracking unit during the refining and chemical production process, and map them to the three-dimensional spatial grid model of the catalytic cracking unit to generate a three-dimensional real-time state field inside the catalytic cracking unit.

[0076] The 3D spatial distribution field reconstruction module is used to perform spatial smoothing and physical consistency correction on the 3D real-time state field, and reconstruct the 3D spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator.

[0077] Three-dimensional dynamic evolution simulation module: used to input the three-dimensional spatial distribution field into a preset fluid dynamics simulation model, and output the three-dimensional dynamic evolution field of the catalyst fluidization state and reaction progress state in the future period.

[0078] The 3D feature identification and anomaly warning module: Based on the gradient and uniformity judgment rules of the 3D dynamic evolution field, it performs spatial feature identification of the 3D dynamic evolution field and generates a warning priority ranking table.

[0079] Control strategy generation and verification module: Based on the judgment result of S4, it generates and executes the optimized control strategy of the catalytic cracking unit, and sequentially executes S1 to S3 to obtain the three-dimensional dynamic evolution field after the control strategy is executed. The control effect is verified by comparing the three-dimensional dynamic evolution field before and after the control strategy is executed.

[0080] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0081] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A refining production monitoring control method based on three-dimensional modeling optimization, characterized in that, Comprise the following steps: S1. Constructing a three-dimensional real-time state field: real-time acquisition of the time series data of temperature, pressure, catalyst inventory and key component concentration in the reactor and regenerator of the catalytic cracking unit in the refining production process, and mapping it to the three-dimensional spatial grid model of the catalytic cracking unit to generate a three-dimensional real-time state field inside the catalytic cracking unit; S2. Reconstructing the three-dimensional spatial distribution field: spatial smoothing and physical consistency correction of the three-dimensional real-time state field to reconstruct the three-dimensional spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator; S3. Three-dimensional dynamic evolution simulation: input the three-dimensional spatial distribution field into the preset fluid dynamics simulation model, and output the three-dimensional dynamic evolution field of the catalyst fluidization state and reaction state in the future period; S4. Identify three-dimensional features and abnormal early warning: based on the gradient and uniformity judgment rule of the three-dimensional dynamic evolution field, the spatial feature recognition and the generation of the warning priority ranking table are carried out on the three-dimensional dynamic evolution field; S5. Control strategy generation and verification: according to the judgment result of S4, the optimization control strategy of the catalytic cracking unit is generated and executed, and the three-dimensional dynamic evolution field after the execution of the control strategy is obtained by sequentially executing S1 to S3, and the control effect is verified by comparing the three-dimensional dynamic evolution field before and after the execution of the control strategy.

2. The method according to claim 1, wherein, The specific method for generating a three-dimensional real-time state field inside the catalytic cracking unit is: The obtained time series data of temperature, pressure, catalyst inventory and key component concentration in the reactor and regenerator of the catalytic cracking unit is mapped to the corresponding spatial nodes of the pre-constructed three-dimensional spatial grid model of the catalytic cracking unit according to the physical installation position spatial coordinates of the corresponding sensors, and the three-dimensional spatial grid model is deduced through the Kriging spatial interpolation algorithm to generate continuous distribution temperature field, pressure field, catalyst density field and key component concentration field, which are integrated in the three-dimensional spatial grid model to generate a three-dimensional real-time state field representing the real-time working condition inside the catalytic cracking unit.

3. The method according to claim 2, wherein, The specific method for spatial smoothing and physical consistency correction of the three-dimensional real-time state field is: Gaussian filtering algorithm is used to smooth the data of each spatial node in the three-dimensional real-time state field, and physical consistency correction is performed on the data of each spatial node in the smoothed three-dimensional real-time state field based on the mass and energy conservation relationship of the catalytic cracking unit.

4. The method according to claim 3, wherein, The specific method for reconstructing the three-dimensional spatial distribution field of catalyst density, temperature, pressure and key component concentration in the reactor and regenerator is: The three-dimensional real-time state field after spatial smoothing and physical consistency correction is used as the basic data source, and the three-dimensional spatial distribution field with physical consistency is generated by solving the optimal three-dimensional spatial distribution that satisfies all physical constraints through mathematical optimization algorithm.

5. The method according to claim 4, wherein, The output of the three-dimensional dynamic evolution field of the catalyst fluidization state and the reaction state in the future period: The three-dimensional spatial distribution field with physical consistency is input as an initial condition into a preset fluid dynamics simulation model, and a three-dimensional dynamic evolution field in a future period is output by solving the non-steady-state numerical calculation of the mass, momentum, energy and component conservation equation set in a discrete time step sequence, the three-dimensional dynamic evolution field including a catalyst fluidization state three-dimensional dynamic evolution field and a reaction state three-dimensional dynamic evolution field, wherein the catalyst fluidization state three-dimensional dynamic evolution field includes a catalyst particle movement velocity vector field, a catalyst volume fraction distribution field and a gas-solid phase interaction force field, and the reaction state three-dimensional dynamic evolution field includes a temperature field, a pressure field, a key component concentration field and a chemical reaction rate distribution field of a spatial node.

6. The method according to claim 1, wherein, The gradient and uniformity determination rule of the three-dimensional dynamic evolution field includes a gradient abnormality determination condition: for a scalar field in the three-dimensional dynamic evolution field, including a temperature field, a pressure field, a key component concentration field and a chemical reaction rate distribution field, if the average value of the spatial gradient modulus of any one of the scalar physical fields in a connected space region is greater than the preset gradient threshold of the physical quantity in the target time length, it is determined that the region has gradient abnormality; a uniformity abnormality determination condition: for a vector physical field in the three-dimensional dynamic evolution field, including a catalyst particle movement velocity vector field, a catalyst volume fraction distribution field and a gas-solid phase interaction force field, if the variation coefficient of the vector modulus of all spatial grid nodes of any one of the vector physical fields in a preset three-dimensional spatial analysis region is greater than the preset uniformity threshold of the vector field in the target time length, it is determined that the analysis region has uniformity abnormality.

7. The method according to claim 6, wherein, The method for identifying the spatial characteristics of the three-dimensional dynamic evolution field and generating a warning priority ranking table includes: extracting the gradient abnormality region and the uniformity abnormality region identified based on the gradient abnormality determination condition and the uniformity abnormality determination condition from the three-dimensional dynamic evolution field, and identifying the three-dimensional spatial position, the spatial volume, the type of abnormal physical field contained, the proportion of each abnormal physical field type in the corresponding abnormal region and the abnormal physical quantity intensity in the three-dimensional dynamic evolution field; multiplying the proportion of each abnormal physical field type in the corresponding abnormal region and the abnormal physical quantity intensity to obtain a warning index, and generating a warning priority ranking table for all identified abnormal physical fields according to the warning index.

8. The method according to claim 7, wherein, The method for generating an optimization control strategy for the catalytic cracking device and executing the optimization control strategy includes: generating a corresponding optimization control strategy according to the identified abnormal physical field type; when the identified abnormal physical field type is a catalyst particle movement velocity field or a catalyst volume fraction distribution field, adjusting the direction and amplitude of each operating parameter of the catalyst circulation system according to the deviation angle of the catalyst particle movement direction and the preset fluidization direction, the deviation value and the distribution range of the average movement amplitude; when the identified abnormal physical field type is a temperature field or a chemical reaction rate distribution field, adjusting the amplitude of each operating parameter in the reactor according to the high temperature point distribution of the temperature field and the abnormal physical quantity intensity of the reaction rate distribution field. When the identified abnormal physical field type is a pressure field, then the pressure balance parameters of the reactor and regenerator or the pressure setting of the fluidization medium system are adjusted according to the high-pressure point or low-pressure point distribution and gradient of the pressure field; When the identified abnormal physical field type is a gas-solid phase interaction force field, then the medium flow rate and pressure parameters of the fluidization medium system are adjusted according to the abnormal amplitude deviation and spatial gradient distribution of the interaction force field; When the identified abnormal physical field type is a key component concentration field, then the specific key component determination control operation direction is obtained; When multiple abnormal physical fields exist in the catalytic cracking device at the same time, the control operation direction corresponding to the abnormal physical field with the highest priority is preferentially executed based on a pre-warning priority ranking table, and the control operation directions of the remaining abnormal physical fields are coordinately adjusted to generate a conflict-free optimization control strategy according to the process mechanism rules, which is then issued to the field execution mechanism for execution through a distributed control system.

9. The method according to claim 8, wherein, The control effect is verified by comparing the three-dimensional dynamic evolution fields before and after the execution of the control strategy, and the specific method is as follows: After the execution of the control strategy, the new time series data is obtained in S1, the three-dimensional spatial distribution field after the execution is reconstructed, and S3 is further executed, in which the three-dimensional spatial distribution field after the execution is input as an initial condition into the preset fluid dynamics simulation model to output the three-dimensional dynamic evolution field after the execution; The three-dimensional dynamic evolution field after the execution of the control strategy is compared with the three-dimensional dynamic evolution field before the execution, the state of the gradient abnormal region and the uniformity abnormal region identified in the three-dimensional dynamic evolution field before the execution is checked in the three-dimensional dynamic evolution field after the execution, and if the average value of the spatial gradient modulus of the physical quantity in the gradient abnormal region is less than the preset gradient threshold value and the variation coefficient value of the physical quantity in the uniformity abnormal region is less than the preset uniformity threshold value, it is determined that the control strategy is effective, otherwise it is determined that the control strategy is insufficient.

10. A system for performing a method of monitoring and control of refining production based on three-dimensional modeling optimization according to any one of claims 1-9, characterized by It comprises: A three-dimensional real-time state field module is constructed, which is used to obtain the time series data of the temperature, pressure, catalyst inventory and key component concentration in the reactor and regenerator of the catalytic cracking device in the refining and chemical production process in real time, and map the time series data to a three-dimensional spatial grid model of the catalytic cracking device to generate a three-dimensional real-time state field inside the catalytic cracking device; A three-dimensional spatial distribution field module is reconstructed, which is used to perform spatial smoothing and physical consistency correction on the three-dimensional real-time state field to reconstruct a three-dimensional spatial distribution field of the catalyst density, temperature, pressure and key component concentration in the reactor and regenerator; A three-dimensional dynamic evolution simulation module is used to input the three-dimensional spatial distribution field into a preset fluid dynamics simulation model to output a three-dimensional dynamic evolution field of the catalyst fluidization state and reaction state in the future period; A three-dimensional feature identification and abnormal pre-warning module is used to perform spatial feature identification on the three-dimensional dynamic evolution field based on the gradient and uniformity determination rules of the three-dimensional dynamic evolution field and generate a pre-warning priority ranking table. The control strategy generation and verification module is used for generating the optimized control strategy of the catalytic cracking device according to the judgment result of S4 and executing, and sequentially executing S1 to S3 to obtain the three-dimensional dynamic evolution field after the execution of the control strategy, and verifying the control effect by comparing the three-dimensional dynamic evolution fields before and after the execution of the control strategy.

Images