Intelligent control method, device and equipment for aviation kerosene preparation system

Abstract

The invention provides an intelligent control method, device and equipment for an aviation kerosene preparation system, and the method comprises the steps: obtaining various sample data, the various sample data comprises physical property parameters, distillation range distribution parameters and component content parameters of sample oil, and process parameters of the aviation kerosene preparation system; according to the multiple sample data, a long-short-term memory network algorithm and a physical information neural network algorithm, a prediction algorithm model is established, and the prediction algorithm model is at least used for predicting the yield of aviation kerosene of the aviation kerosene preparation system; according to the prediction algorithm model and the multiple sample data, determining the prediction yield of the aviation kerosene in a first preset time period and multiple operation parameters of the aviation kerosene preparation system in the first preset time period; and adjusting the process parameters of the aviation kerosene preparation system according to the predicted yield of the aviation kerosene and the plurality of operation parameters. The method solves the technical problem of unstable aviation kerosene yield in the aviation kerosene preparation process of the aviation kerosene preparation system.

Description

Technical Field

[0001] This application relates to the field of aviation kerosene production, and more specifically, to an intelligent control method for an aviation kerosene preparation system, an intelligent control device for an aviation kerosene preparation system, an intelligent control equipment for an aviation kerosene preparation system, a computer-readable storage medium, and a processor. Background Technology

[0002] In the process of producing aviation kerosene from direct coal liquefaction, the oil undergoes processing steps such as hydrorefining, hydromodification, atmospheric distillation, and vacuum distillation. The quality of the produced aviation kerosene is highly dependent on the properties of the initial oil and changes in processing conditions. The production process is complex and influenced by numerous factors, all of which have a significant impact on the stable production of aviation kerosene. Summary of the Invention

[0003] The main objective of this application is to provide an intelligent control method, an intelligent control device, an intelligent control equipment, a computer-readable storage medium, and a processor for an aviation kerosene preparation system, so as to at least solve the technical problem of unstable aviation kerosene yield in the process of preparing aviation kerosene in the aviation kerosene preparation system in the prior art.

[0004] To achieve the above objectives, according to one aspect of this application, an intelligent control method for an aviation kerosene preparation system is provided, comprising: acquiring multiple sample data, the multiple sample data including: physical property parameters, distillation range distribution parameters, and component content parameters of sample oil, and process parameters of the aviation kerosene preparation system, wherein the sample oil includes diluted direct coal liquefaction oil, hydrogenation products of the direct coal liquefaction oil, and substances obtained after gas-liquid separation of the hydrogenation products, wherein the substances include aviation kerosene; establishing a prediction algorithm model based on the multiple sample data, a long short-term memory network algorithm, and a physical information neural network algorithm, wherein the prediction algorithm model is used at least to predict the yield of aviation kerosene in the aviation kerosene preparation system; determining the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period based on the prediction algorithm model and the multiple sample data, wherein the start time of the first preset time period is the time when the multiple sample data are acquired, and the operating parameters include at least the temperature, pressure, and gas concentration during the operation of the aviation kerosene preparation system; and adjusting the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and the multiple operating parameters.

[0005] According to another aspect of this application, an intelligent control device for an aviation kerosene preparation system is provided, comprising: an acquisition module for acquiring multiple sample data, the multiple sample data including: physical property parameters, distillation range distribution parameters, and component content parameters of sample oil, and process parameters of the aviation kerosene preparation system, the sample oil including diluted direct coal liquefaction oil, hydrogenation products of the direct coal liquefaction oil, and substances after gas-liquid separation of the hydrogenation products; an establishment module for establishing a prediction algorithm model based on the multiple sample data, a long short-term memory network algorithm, and a physical information neural network algorithm, the prediction algorithm model being used at least to predict the yield of aviation kerosene in the aviation kerosene preparation system; a determination module for determining, based on the prediction algorithm model and the multiple sample data, the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period, the start time of the first preset time period being the time when the multiple sample data are acquired, and the types of operating parameters being the same as the types of process parameters; and a control module for adjusting the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and the multiple operating parameters.

[0006] According to another aspect of this application, an intelligent control device for an aviation kerosene preparation system is provided, comprising: a production process control unit, including a hydrogenation unit and a distillation unit, the distillation unit being connected to the hydrogenation unit, the hydrogenation unit performing hydrogenation reforming treatment on direct coal liquefaction oil to obtain hydrogenated products of the direct coal liquefaction oil, and the distillation unit performing gas-liquid separation on the hydrogenated products; a detection unit, connected to the production process control unit, for monitoring the process parameters of the aviation kerosene preparation system, and for detecting at least a sample oil to obtain various sample data, the sample oil including: diluted direct coal liquefaction oil, the hydrogenated products, and substances obtained after gas-liquid separation of the hydrogenated products; and an intelligent central unit, connected to the detection unit and the production process control unit, for executing an intelligent control method for the aviation kerosene preparation system.

[0007] According to another aspect of this application, a processor is provided for running a program, wherein the program executes the intelligent control method for the aviation kerosene preparation system.

[0008] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform the intelligent control method of the aviation kerosene preparation system.

[0009] By applying the technical solution of this application, the physical properties, distillation range distribution, and component content parameters of the sample oil, as well as the process parameters of the aviation kerosene preparation system, are obtained; a prediction algorithm model is established based on multiple sample data, a long short-term memory network algorithm, and a physical information neural network algorithm; based on the prediction algorithm model and multiple sample data, the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period are determined; and the process parameters of the aviation kerosene preparation system are adjusted based on the predicted yield of aviation kerosene and multiple operating parameters. This application derives a predictive algorithm model based on the physical properties, distillation range distribution, and component content parameters of the sample oil, as well as the process parameters of the aviation kerosene preparation system. This predictive algorithm model can predict the predicted yield of aviation kerosene over a future period, and multiple operating parameters of the aviation kerosene preparation system. Based on the predicted yield of aviation kerosene, it can determine whether there is a downward trend in the yield. When a decline is detected, the production process control system is adjusted in a timely manner to stabilize the yield of aviation kerosene. By adjusting the aviation kerosene preparation system through multiple predicted operating parameters, the internal production conditions of the aviation kerosene preparation system can be kept stable, thereby stabilizing the yield of aviation kerosene. Attached Figure Description

[0010] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0011] Figure 1 A hardware structure block diagram of a mobile terminal for performing an intelligent control method for an aviation kerosene preparation system according to an embodiment of this application is shown.

[0012] Figure 2 A flowchart illustrating an intelligent control method for an aviation kerosene preparation system according to an embodiment of this application is shown.

[0013] Figure 3 A structural block diagram of an intelligent control device for an aviation kerosene preparation system according to an embodiment of this application is shown;

[0014] Figure 4 A structural block diagram of an intelligent control device for an aviation kerosene preparation system according to an embodiment of this application is shown;

[0015] Figure 5 A structural diagram of an intelligent control device for another aviation kerosene preparation system provided according to an embodiment of this application is shown;

[0016] Figure 6 The diagram shows the structural block diagram of the sampling unit in the intelligent control equipment of the aviation kerosene preparation system;

[0017] Figure 7 A structural block diagram of the oil analysis unit in the intelligent control equipment of an aviation kerosene preparation system is shown.

[0018] The above figures include the following reference numerals:

[0019] 102. Processor; 104. Memory; 106. Transmission equipment; 108. Input / output devices; 410. Detection unit; 41. Sampling unit; 411. Sampling probe; 412. Automatic sampler; 413. Electronic tag writing unit; 414. Self-cleaning unit; 42. Oil analysis unit; 421. Viscosity sensor; 422. Multi-channel solvent injection pump; 423. Differential scanning calorimeter; 424. Sampling robotic arm; 425. RFID radio frequency identifier; 426. Fully automated distillation range analyzer; 427. Fully two-dimensional gas chromatograph; 50. 51. Production process control unit; 52. Hydrogenation refining unit; 53. Hydrogenation reforming unit; 54. Atmospheric distillation unit; 55. Vacuum distillation unit; 56. High-pressure separator; 57. Atmospheric separator; 58. Condenser; 60. Intelligent central unit; 71. Oil transportation unit; 72. Coal direct liquefaction oil storage tank; 73. Coal direct liquefaction oil transportation pipeline; 74. Coal direct liquefaction oil transportation pump; 85. Heating furnace; 86. Pneumatic transmission unit; 87. Circulating hydrogen compressor; 88. Hydrogen gas cylinder; 89. Hydrogen gas transportation pipeline; 90. Intelligent storage unit. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] As described in the background section, in the existing technology for producing aviation kerosene from direct coal liquefaction, there are many influencing factors, making it difficult to stably produce aviation kerosene. To solve the problem of unstable aviation kerosene yield in the process of producing aviation kerosene in an aviation kerosene preparation system, embodiments of this application provide an intelligent control method for an aviation kerosene preparation system, an intelligent control device for an aviation kerosene preparation system, an intelligent control equipment for an aviation kerosene preparation system, a computer-readable storage medium, and a processor.

[0024] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0025] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for an intelligent control method of an aviation kerosene preparation system according to an embodiment of this application. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0026] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the intelligent control method of the aviation kerosene preparation system in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thus implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0027] This embodiment provides an intelligent control method for an aviation kerosene preparation system that runs on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0028] Figure 2 This is a flowchart of an intelligent control method for an aviation kerosene preparation system according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0029] Step S1: Obtain various sample data, including: physical property parameters, distillation range distribution parameters, component content parameters of sample oil, and process parameters of aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of coal direct liquefaction oil, and substances after gas-liquid separation of hydrogenation products, including aviation kerosene.

[0030] Specifically, multiple sample data are acquired in real time. The sample oils mentioned above are substances produced by the aviation kerosene preparation system after viscosity adjustment. The substances obtained after gas-liquid separation of the hydrogenation products include aviation kerosene, light distillate oil, diesel oil, and heavy distillate oil. The physical properties, distillation range distribution parameters, and component content parameters of the sample oils are obtained through laboratory analysis. The physical properties include density, dynamic viscosity, freezing point, and smoke point. The distillation range distribution parameters include the initial boiling point, distillation temperatures at different distillation ranges (e.g., 10%, 30%, 50%, 70%, and 90%), and the final boiling point. The component content parameters include the cycloalkanes content, aromatics content, and C11+ heavy cycloalkanes threshold. The physical properties also include impurity content, such as sulfur and nitrogen content. The aforementioned process parameters include process unit parameters, fluid transport parameters, and equipment status parameters. Process unit parameters include: temperature, hydrogen partial pressure, liquid hourly space velocity (LHSV), hydrogen-to-oil ratio, dynamic data on axial temperature difference in the catalyst bed, vacuum level, and reflux ratio. Fluid transport parameters include the flow rate of direct coal liquefaction oil, the outlet temperature of the heater, the hydrogen circulation rate, and the operating status of the circulating hydrogen compressor. Equipment status parameters include the vibration spectrum of the reactor pressure vessel, pump and valve openings, and heat exchanger efficiency. These parameters allow for the construction of a "raw material-process-product" relationship network for the aviation kerosene production system, providing comprehensive data support for subsequent intelligent decision-making.

[0031] Step S2: Based on multiple sample data, long short-term memory network algorithm and physical information neural network algorithm, establish a prediction algorithm model. The prediction algorithm model is used to predict the yield of aviation kerosene in the aviation kerosene preparation system.

[0032] Specifically, using the aforementioned process parameters and test data, the preliminary prediction algorithm models established by the Long Short-Term Memory (LSTM) network algorithm and the Physics-Informed Neural Network (PINN) algorithm are trained to obtain the final prediction algorithm model. This prediction algorithm model can predict the optimal production process parameters and the yield of aviation kerosene, and optimize the production process parameters of the aviation kerosene preparation system and carry out further control.

[0033] Step S3: Based on the prediction algorithm model and multiple sample data, determine the predicted yield of aviation kerosene within the first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period. The start time of the first preset time period is the time when multiple sample data are obtained. The operating parameters include the temperature, pressure and gas concentration of the aviation kerosene preparation system during operation.

[0034] Specifically, the LSTM algorithm reads data from multiple samples, constructs a three-dimensional input vector, encodes features of time-series data such as the temperature sequence of the hydrogenation unit and hydrogen partial pressure fluctuations, controls the weight of historical information through the algorithm, captures the long-term correlation of process parameters, inputs the current 72-hour process parameter data, and outputs the predicted jet fuel yield within a first preset time period (e.g., 2 hours) in the future, identifying the inflection point of yield decline in advance. The PINN algorithm reads data from multiple samples, normalizes physical parameters of different dimensions to the same scale, and outputs the temperature field T(x, y, z), pressure field (x, y, z), and gas concentration field C(x, y, z) for the next 2 hours, with temperature error ±2℃, pressure error ±0.2MPa, and concentration error ±0.3wt%. The aforementioned first preset time period is a time obtained through multiple tests, and the prediction for the next 2 hours is relatively accurate.

[0035] Step S4: Adjust the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and multiple operating parameters;

[0036] Specifically, if the predicted yield of aviation kerosene does not show a downward inflection point, the parameters of the aviation kerosene preparation system can be adjusted based on multiple operating parameters. If the predicted yield of aviation kerosene does show a downward inflection point, the sample data of the current aviation kerosene preparation system is reacquired using the above method, and multiple operating parameters of the aviation kerosene preparation system within the first preset time period are predicted using the prediction algorithm model. The aviation kerosene preparation system is then adjusted based on the predicted operating parameters.

[0037] This embodiment utilizes a predictive algorithm model derived from the physical properties, distillation range distribution, and component content parameters of the sample oil, as well as the process parameters of the aviation kerosene preparation system. This model can predict the future yield of aviation kerosene and multiple operating parameters of the aviation kerosene preparation system. Based on the predicted yield, it can determine whether there is a downward trend in the aviation kerosene yield. If a decline is detected, the production process control system can be adjusted promptly to stabilize the aviation kerosene yield. Furthermore, the aviation kerosene preparation system can be adjusted based on the predicted operating parameters to maintain stable internal production conditions, thereby stabilizing the aviation kerosene yield.

[0038] Specifically, when the predicted yield of aviation kerosene reaches a decline inflection point, the entire aviation kerosene production system can be monitored. If the decline inflection point is detected to be related to temperature or pressure fluctuations in the hydrogenation unit, adjustment commands are quickly sent to the hydrogenation unit via the TSN network for fine-tuning of temperature and pressure. For example, when cycloalkanes fluctuations exceed ±3%, the system automatically triggers the temperature gradient compensation program in the hydrogenation unit, adjusting the axial temperature difference of the reactor bed to maintain the cycloalkanes ring-opening efficiency of the catalyst. The possible adjustment range is ±5℃ from the original level. For the distillation unit, the system automatically analyzes the mass spectrometry data of the top components and constructs a dynamic equation for tray efficiency versus reflux ratio. If the C11+ heavy cycloalkanes content in the aviation kerosene fraction is detected to be too high (>1.2%), the vacuum level of the vacuum tower is immediately adjusted to a lower level (e.g., below 10 kPa), and the side-stream recovery ratio is optimized to ensure that the final boiling point of aviation kerosene is stably controlled within the range of 270±2℃, thereby avoiding the negative impact of heavy cycloalkanes on yield and oil quality. Considering the impact of the hydrogen-to-oil ratio and flow rate on the hydrogenation unit, if the prediction algorithm model indicates that insufficient hydrogen-to-oil ratio or unstable flow rate is the main reason for the yield decline, the hydrogen-to-oil volume ratio can be adjusted through the TSN network (it can be increased to near the maximum threshold, such as close to 2000 v / v), and the flow rate of the direct coal liquefaction oil can be optimized (ensuring that the liquid hourly space velocity is within the optimal range, such as 0.3–4.0 h⁻¹). -1 To maintain optimal hydrogenation conversion and hydrogen recycling efficiency, the control system initiates a catalyst replacement process or compensates for catalyst activity loss by adjusting reaction conditions (such as increasing hydrogen partial pressure and optimizing temperature sequence) to improve oil conversion and restore yield.

[0039] In the specific implementation process, step S2 above can be achieved through the following steps:

[0040] The prediction algorithm model also includes a screening model, which establishes a prediction algorithm model based on multiple sample data, long short-term memory network algorithms, and physical information neural network algorithms, including:

[0041] Based on the screening model, the input sample data from various sources are filtered. The screening model is as follows:

[0042] f t =σ(W f ·[h t-1 ,x t ]+b f ),

[0043] Among them, f t For the various sample data remaining after screening, W f Let b be the weight matrix. f Let h be the bias vector, σ be the sigmoid activation function, and h be the bias vector.t-1 To represent the hidden state at the previous time step (t-1), x t The above sample data;

[0044] Based on the data from the various samples remaining after screening and the Long Short-Term Memory (LSTM) network algorithm, a Long Short-Term Memory (LSTM) network model was established; the specific input data for the LSTM model included the reaction temperature T. reactor hydrogen partial pressure P H2 Liquid hourly space velocity (LHSV), hydrogen to direct coal liquefaction oil flow ratio (H2:Oil), aviation kerosene density (ρ), flow rate influence factor (η), 10% or 90% distillation temperature (T10% / T90%), and aviation kerosene component content (C). nap The input matrix X has (9, 72) rows and columns, which means 9 features and 72 time steps.

[0045] The LSTM model is trained using historical sample data as input. Training stops after the neural network converges to obtain the LSTM model, which can adapt to the fluctuations in the properties of coal direct liquefaction oil. The evolutionary algorithm can be obtained from the following formula:

[0046] i t =σ(W i ·[h t-1 ,x t ]+b i ),

[0047] K t =tanh(W K ·[h t-1 ,x t ]+b K ),

[0048] C t =f t ☉C t-1 +i t ☉K t ,

[0049] Among them, i t To control the update gate, σ is the sigmoid activation function, and W... i Let b be the weight matrix. i h is the bias vector. t-1 To represent the hidden state at the previous time step (t-1), x t For the above sample data, K t Let W be the candidate cell state, tanh be the core activation function, and W be the activating function. k Let b be the weight matrix. k C is the bias vector. t For the current cell state, C t-1Let f be the cell state at the previous moment, ⊙ be the Hadamard product element-wise multiplication, and f be the cell state at the previous moment. t This refers to the data from the various samples that remained after screening.

[0050] The algorithm outputs a predicted jet fuel yield for the next two hours, identifying the inflection point of yield decline in advance. The output algorithm can be derived from the following formula:

[0051] y=σ(W y ·h t +b y ),

[0052] Where y is the predicted jet fuel yield; σ is the sigmoid activation function; h t To represent the hidden state at the current moment; b y This is the bias vector.

[0053] Based on the energy conservation equation, mass conservation equation, and pyrolysis reaction rate equation, a physical information neural network model is established; the energy conservation equation is: The mass conservation equation is The pyrolysis reaction rate equation is

[0054] in, Let α be the rate of change of temperature with time, and ▽ be the thermal diffusivity. 2 T is the temperature gradient, β is the exothermic coefficient of the reaction, r(T,C) is the pyrolysis reaction rate, γ is the pressure influence factor, and P H2 For hydrogen partial pressure, Let D be the rate of change of concentration over time, and D be the mass diffusion coefficient. Let V be the concentration gradient, δ be the reaction consumption coefficient, η be the flow rate influence factor, and V be the flow rate influence factor. oil Let k0 be the feed velocity for direct coal liquefaction oil, and k0 be the pre-exponential factor. This is the Arrhenius equation, where Ea is the activation energy R. g Let T be the gas constant, C be the temperature, n be the gas concentration, and f be the reaction order. cat (η) is the catalyst activity function.

[0055] Input dimensions:

[0056] The above describes an 8-dimensional spatiotemporal system, where (x, y, z) are three-dimensional spatial coordinates (position within the reactor); t is time. V is the partial pressure of hydrogen. oil η is the feed flow rate. cat T is the activity coefficient. inlet This refers to the feed temperature;

[0057] Output dimension: output = [T, P, C], where T is the temperature field; P is the pressure field; and C is the gas concentration field.

[0058] A prediction algorithm model is established based on the Long Short-Term Memory (LSTM) network model and the Physical Information Neural Network (PIN) model. The LSTM model and the PINN model work together to construct a comprehensive prediction algorithm model. The LSTM component of the prediction algorithm model is responsible for predicting the yield of aviation kerosene, while the PINN component is responsible for predicting multiple operating parameters of the aviation kerosene preparation system. This combined model can simultaneously consider both the yield and operating parameters of aviation kerosene, providing more comprehensive prediction results and ensuring real-time optimization of production conditions. These operating parameters may also include the process parameters of the aviation kerosene preparation system.

[0059] After the prediction algorithm model finishes its predictions, the predicted operating parameters can be input into a virtual reactor built using ANSYS Fluent software to verify quality indicators and safety. If the prediction results pass verification, optimized parameters are generated. The LSTM algorithm is used to predict future process trends and generate constraint boundaries. The PINN algorithm is used to calculate the theoretical optimal solution. Candidate parameter combinations are tested in the virtual reactor model to screen for feasible solutions that simultaneously meet the following conditions: maximum bed temperature < catalyst tolerance limit, aviation kerosene final boiling point < 272℃, and hydrogen consumption increase < 5%.

[0060] In some optional implementations, the intelligent control method further includes: acquiring multiple process data and multiple test data for direct coal liquefaction oil, wherein the process data includes production process unit parameters, fluid transport parameters, and equipment status parameters, and the test data includes physical property parameters, distillation range distribution, and component content; and updating the prediction algorithm model based on the process data and test data. Every 8 hours, the acquired historical process data and multiple test data are used as input to train the prediction algorithm model, updating the weights of the LSTM algorithm model in the prediction algorithm model. Training stops after the prediction algorithm model converges to obtain an updated prediction algorithm model, making its prediction results more accurate.

[0061] This application also provides an intelligent control device for an aviation kerosene preparation system. It should be noted that the intelligent control device for the aviation kerosene preparation system in this application can be used to execute the intelligent control method for the aviation kerosene preparation system provided in this application. This device is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0062] The following describes the intelligent control device for the aviation kerosene preparation system provided in the embodiments of this application.

[0063] Figure 3 This is a schematic diagram of the intelligent control device for an aviation kerosene preparation system according to an embodiment of this application. Figure 3 As shown, the device includes: an acquisition module 10 for acquiring various sample data, including physical property parameters, distillation range distribution parameters, and component content parameters of sample oil, as well as process parameters of the aviation kerosene preparation system. The sample oil includes diluted direct coal liquefaction oil, hydrogenated products of direct coal liquefaction oil, and substances after gas-liquid separation of the hydrogenated products; an establishment module 20 for establishing a prediction algorithm model based on the various sample data, a long short-term memory network algorithm, and a physical information neural network algorithm. The prediction algorithm model is used to predict at least the yield of aviation kerosene in the aviation kerosene preparation system; a determination module 30 for determining the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period based on the prediction algorithm model and the various sample data. The start time of the first preset time period is the time when the various sample data are acquired, and the types of operating parameters are the same as the types of process parameters; and a control module 40 for adjusting the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and multiple operating parameters.

[0064] By acquiring physical properties, distillation range distribution, and component content parameters of the sample oil, as well as the process parameters of the aviation kerosene preparation system, the module establishes a predictive algorithm model. Based on the sample data and this predictive algorithm model, the module can predict the future yield of aviation kerosene over a certain period, and determine multiple operating parameters of the aviation kerosene preparation system. Based on the predicted yield, the module can determine if there is a downward trend in the aviation kerosene yield. If a decline is detected, the control module promptly adjusts the production process control unit to stabilize the aviation kerosene yield. Furthermore, the control module can adjust the aviation kerosene preparation system based on the predicted operating parameters to maintain stable internal production conditions, thereby stabilizing the aviation kerosene yield.

[0065] An optional scheme includes a module comprising a first screening module, a first sub-module, a second sub-module, and a third sub-module. The first screening module is used to screen multiple input sample data according to a screening model, where the screening model is: f t =σ(W f ·[h t-1 ,x t ]+b f ), where f t Data from multiple samples remaining after screening; W f Let b be the weight matrix.f Here, σ is the bias vector; h is the sigmoid activation function; t-1 To represent the hidden state at the previous time step (t-1); x t The above sample data is used as the basis for the following steps: The first sub-module is used to establish a Long Short-Term Memory (LSTM) network model based on the various sample data remaining after screening and the LTM network algorithm; the second sub-module is used to establish a physical information neural network model based on the energy conservation equation, the mass conservation equation, and the pyrolysis reaction rate equation; and the third sub-module is used to establish a prediction algorithm model based on the LSM network model and the physical information neural network model.

[0066] In one optional embodiment, the intelligent control device further includes a first acquisition module and an update module. The first acquisition module is used to acquire multiple process data and multiple test data of coal direct liquefaction oil. The process data includes production process unit parameters, fluid transport parameters, and equipment status parameters. The test data includes physical property parameters, distillation range distribution, and component content. The update module is used to update the prediction algorithm model based on the process data and test data.

[0067] In an optional solution, the intelligent control device also includes a communication module. The communication module adopts the Open Platform Communication Unified Architecture (OPCUA) and Time-Sensitive Network Protocol (TSN) to achieve deterministic communication with the PLC and online instruments. The transmission delay of key instructions is <10ms. A quantum encryption gateway is deployed to perform end-to-end encryption of control instructions, blocking 99.99% of attacks.

[0068] Given that the module predicts multiple operating parameters of the aviation kerosene preparation system and the yield of aviation kerosene:

[0069] In one optional scheme, the intelligent control device also includes a full-process collaborative management and control module. The communication module encodes the optimized parameters into OPCUA messages and sends them to the full-process collaborative management and control module via TSN protocol for information transmission. The optimized parameters are then sent to the hydrorefining unit, hydromodification unit, atmospheric distillation unit, and vacuum distillation unit via PLC. This achieves a temperature control accuracy of ±1.5℃ for the hydrorefining unit, a hydrogen-to-oil ratio adjustment response time of <15s, and a distillation column reflux ratio adjustment of 0.1 / step. When a deviation in the jet fuel distillation range is detected, a three-level adjustment strategy is automatically activated: fine-tuning the vacuum distillation column side-stream extraction ratio by ±3%, compensating for the heater power by ±5kW, and switching the catalyst bed, with a response time of <8min.

[0070] In one alternative, the intelligent control device further includes an autonomous evolutionary learning platform module. The communication module encodes the optimization parameters into OPCUA messages and sends them to the autonomous evolutionary learning platform module via TSN for real-time feature data extraction, enabling autonomous iterative optimization. Preprocessing is performed through edge computing nodes to construct a composite feature space containing time-domain statistics, frequency-domain features, and physical constraints. Subsequently, a hybrid evolutionary strategy is adopted to run genetic algorithms and gradient optimization in parallel within the hypernetwork architecture. Through crossover mutation, a model structure with both physical interpretability and data-driven capabilities is generated. Pareto front analysis is used to screen the optimal model in the multi-objective space. After generating the optimal model, it is applied to the determination module to generate the optimal parameters.

[0071] In one optional scheme, the intelligent control device also includes an intelligent operation module for experimental instruments. The communication module encodes the optimized parameters into OPCUA messages and sends them to the intelligent operation module for experimental instruments via the TSN protocol to control the operation of the experimental equipment. This enables automatic sampling, separation, and viscosity detection of aviation kerosene samples. Emergency stop protection is achieved through a SIL3-level PLC, which automatically stops operation when a leak or collision with the robotic arm is detected.

[0072] In one optional solution, the intelligent control device also includes a human-machine collaboration module. This module employs a hierarchical architecture combining multimodal fusion and cognitive enhancement to achieve deep collaboration between human intuition and machine intelligence. It enables AR visualization, allowing for 3D perspective viewing of the device through AR glasses. Key parameters are displayed as holographic floating windows, supporting gesture-based adjustment of settings. The PTP protocol aligns multi-source data time nodes, ensuring a latency of <10ms between the AR overlay and sensor data. SLAM is used to establish a device coordinate system, mapping vibration sensor data to a transformation matrix in AR space. A causal reasoning engine performs root cause analysis on alarm events and pushes tiered response plans to mobile terminals.

[0073] The intelligent control device for the aviation kerosene preparation system includes a processor and a memory. The aforementioned acquisition module, establishment module, determination module, and control module are all stored as program units in the memory. The processor executes these program units stored in the memory to achieve the corresponding functions. All of the above modules reside in the same processor; alternatively, the modules may be located in different processors in any combination.

[0074] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters can address the technical problem of unstable aviation kerosene yield in existing aviation kerosene preparation systems.

[0075] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0076] According to one embodiment of this application, an intelligent control device for an aviation kerosene preparation system is provided, such as... Figure 4 As shown, it includes: a production process control unit 50, comprising a hydrogenation unit and a distillation unit, the distillation unit being connected to the hydrogenation unit, the hydrogenation unit performing hydrogenation and upgrading treatment on direct coal liquefaction oil to obtain hydrogenated products of direct coal liquefaction oil, and the distillation unit performing gas-liquid separation on the hydrogenated products; a detection unit 410, connected to the production process control unit 50, monitoring the process parameters of the aviation kerosene preparation system, and at least detecting sample oil to obtain various sample data, the sample oil including: diluted direct coal liquefaction oil, hydrogenated products, and substances obtained after gas-liquid separation of the hydrogenated products; and an intelligent central unit 60, connected to the detection unit 410 and the production process control unit 50, used to execute the intelligent control method of the aviation kerosene preparation system.

[0077] The hydrogenation product of direct coal liquefaction oil is obtained through the hydrogenation unit in the production process control unit, and the gas-liquid separation substance of the hydrogenation product is obtained through the distillation unit. The physical property parameters, distillation range distribution parameters, and component content parameters of the hydrogenation product and the gas-liquid separation substance are obtained by the detection unit. The detection unit monitors the aviation kerosene preparation system to obtain the process parameters of the aviation kerosene preparation system. The above parameters are combined into sample data. The intelligent central unit executes the intelligent control method of the aviation kerosene preparation system based on the sample data. In the process of executing the method, a prediction algorithm model is obtained based on the physical property parameters, distillation range distribution parameters, and component content parameters of the sample oil and the process parameters of the aviation kerosene preparation system. The prediction algorithm model can predict the predicted yield of aviation kerosene in the future period and multiple operating parameters of the aviation kerosene preparation system. Based on the predicted yield of aviation kerosene, it can be determined whether there is a downward trend in the yield. When a decline is detected, the production process control unit is adjusted in time to stabilize the yield of aviation kerosene. The aviation kerosene preparation system is adjusted based on multiple predicted operating parameters to maintain stable internal production conditions and thus stabilize the aviation kerosene yield.

[0078] In the above embodiments, such as Figure 4 and Figure 5As shown, the aforementioned intelligent control equipment also includes an oil conveying unit 70 and a pneumatic transmission unit 80. The oil conveying unit 70 includes a coal direct liquefaction oil storage tank 71, a coal direct liquefaction oil transport pipeline 72, a coal direct liquefaction oil transport pump 73, and a heating furnace 74 connected in sequence. The coal direct liquefaction oil storage tank 71 is used to store coal direct liquefaction oil; the coal direct liquefaction oil transport pipeline 72 is used to transport coal direct liquefaction oil to the production process control unit 50; the coal direct liquefaction oil transport pump 73 is used to transport coal direct liquefaction oil; and the heating furnace 74 is used to heat the coal direct liquefaction oil. The pneumatic transmission unit 80 includes a circulating hydrogen compressor 81, a hydrogen gas cylinder 82, and a hydrogen gas transport pipeline 83. The circulating hydrogen compressor 81 is connected to the hydrogen gas cylinder 82 and the hydrogen refueling unit through the hydrogen gas transport pipeline 83. The circulating hydrogen compressor 81 is mainly used to compress and circulate hydrogen gas to ensure its continuous flow in the intelligent control equipment. The hydrogen gas cylinder 82 is used to store hydrogen gas. The hydrogen gas transport pipeline 83 is used to transport hydrogen gas to the hydrogen refueling unit.

[0079] In the above embodiments, such as Figure 4 As shown, the intelligent control equipment also includes an intelligent storage unit 90, which comprises a sampling robotic arm and a sample storage chamber. The sampling robotic arm is connected to the sampling unit to receive sampled oil. The sample storage chamber automatically scans electronic tags to register and store the sampled oil and backup oil, and uploads the information of the sampled oil and backup oil to the intelligent central unit 60. The intelligent storage unit 90 uses fire-resistant materials as the main material of the storage room, and the oil bottles are made of steel, with each being stored in an isolated manner. The sample storage room is equipped with a sprinkler pump and a carbon dioxide fire extinguishing unit.

[0080] In some alternative implementations, such as Figure 5As shown, the hydrogenation unit includes a hydrogenation refining unit 51 and a hydrogenation reforming unit 52, and the distillation unit includes an atmospheric distillation unit 53 and a vacuum distillation unit 54. Specifically: the hydrogenation refining unit 51 is connected to the hydrogenation reforming unit 52. The hydrogenation refining unit 51 removes impurities from the direct coal liquefaction oil to obtain hydrogenated refined oil and hydrogenated refined recycle gas; the hydrogenation reforming unit 52 reforms the hydrogenated refined oil and hydrogenated refined recycle gas to obtain hydrogenated reformed oil and hydrogenated reformed recycle gas, which are hydrogenation products; the atmospheric distillation unit 53 is connected to both the hydrogenation reforming unit 52 and the vacuum distillation unit 54. The atmospheric distillation unit 53 performs gas-liquid separation on the hydrogenated reformed oil to obtain heavy fraction oil; the vacuum distillation unit 54 performs gas-liquid separation on the heavy fraction oil to obtain aviation kerosene and diesel fuel. The hydrorefining unit uses a Ni-Mo / Al2O3 catalyst to convert impurities such as sulfur and nitrogen in coal liquefaction oil into easily removable forms, resulting in relatively pure hydrorefined oil. The hydromodification unit uses a Ni-W / Al2O3 catalyst to adjust oil properties, such as reducing viscosity and increasing cycloalkane ring-opening efficiency, to obtain hydromodified oil more suitable for distillation. The atmospheric and vacuum distillation units allow for gradual separation and purification. By controlling pressure, temperature, and reflux ratio, effective oil separation is achieved, ensuring high purity and quality of aviation kerosene and diesel. Simultaneously, the heavy fraction can be recycled as feedstock for subsequent hydrorefining units, improving feedstock utilization.

[0081] In the above embodiments, the hydrorefining unit 51 uses a hydrorefining catalyst for hydrorefining. The catalyst support is γ-Al₂O₃, and the active metal component of the catalyst is Ni-Mo, specifically Ni-Mo / Al₂O₃. The process conditions are: temperature 300–450°C, hydrogen partial pressure 6.0–30.0 MPa, and liquid hourly space velocity 0.3–4.0 h⁻¹. -1 The hydrogen-to-oil ratio is 500–2000. Hydrotreating unit 52 uses a hydrotreating catalyst for hydrotreating, adjusting the properties of the hydrorefined oil from hydrorefining unit 51 to obtain hydrotreated oil and hydrotreated recycle gas. The hydrotreating catalyst support is γ-Al₂O₃, and the active metal component of the hydrotreating catalyst is Ni-W, specifically Ni-W / Al₂O₃. The process conditions are: temperature 300–450℃, hydrogen partial pressure 6.0–20.0 MPa, and liquid hourly space velocity 0.3–3.0 h⁻¹. -1The hydrogen-to-oil ratio is 500–3000. Atmospheric distillation unit 53 separates and purifies the hydrotreated oil from hydrotreated unit 52, obtaining atmospheric distillation recycle gas, light component fractions, and heavy component oils. The process conditions are: pressure 0.1 MPa, residual heat of the oil at the hydrotreated unit outlet 300–450°C, top temperature controlled at 100–180°C, bottom temperature controlled at 250–300°C, and dynamic reflux ratio 1.5–3.0. Vacuum distillation unit 54 further separates and purifies the heavy component oil from atmospheric distillation unit 53, obtaining vacuum distillation recycle gas, aviation kerosene, and diesel oil. The process conditions are: pressure 5–20 kPa, top temperature controlled at 180–250°C, and bottom temperature limited to below 350°C.

[0082] In the above embodiments, such as Figure 5 As shown, the production process control unit also includes a high-pressure separator 55 and an atmospheric pressure separator 56. The high-pressure separator 55 is connected to the hydrotreating unit 52 and the atmospheric pressure separator 56, respectively. The atmospheric pressure separator 56 is connected to the atmospheric distillation unit 53. The direct coal liquefaction oil is stored in the direct coal liquefaction oil storage tank 71. After passing through the direct coal liquefaction oil transport pipeline 72 and entering the direct coal liquefaction oil transport pump 73, it is heated in the heater 74. After passing through the direct coal liquefaction oil transport pipeline 72, it enters the hydrorefining reactor (hydrorefining unit 51) through the top. Hydrogen gas is stored in hydrogen gas cylinder 82 and enters the hydrorefining reactor (hydrorefining unit 51) and the hydroreforming reactor (hydroreforming unit 52) ​​through hydrogen gas transport pipeline 83. It exits the hydrorefining reactor at the bottom through a pipeline, enters the hydroreforming reactor at the top through a pipeline, exits the hydroreforming reactor at the bottom through a pipeline, and enters the high-pressure separator 55. The separated gas then enters the circulating hydrogen compressor 81 through hydrogen gas transport pipeline 83. The hydrogen gas is used for the reaction. The separated oil enters the atmospheric pressure separator 56, and the separated gas is discharged through the tail gas pipeline. The oil product enters the atmospheric pressure distillation unit 53 through the direct coal liquefaction oil transport pipeline 72.

[0083] In the distillation unit, the oil leaving the hydrotreating reactor via the direct coal liquefaction oil transport pipeline 72 enters the atmospheric distillation column (atmospheric distillation unit 53). The light fraction at the top of the distillation column is piped into the condenser 57 for condensation and separation. The light fraction gas from the top enters the circulating hydrogen compressor 81 via the hydrogen gas transport pipeline 83 for reuse in the reaction cycle. The oil at the bottom is pumped and piped to the vacuum distillation column (vacuum distillation unit 54). The liquid at the bottom of the vacuum distillation column is piped out as heavy distillate oil product and enters the hydrorefining reactor for recycling of the reaction oil. The light fraction at the top is piped out and stored in the light distillate oil storage tank. The middle fraction is piped out as aviation kerosene and diesel products and stored in aviation kerosene and diesel storage tanks, respectively.

[0084] In some alternative implementations, such as Figures 5 to 7 As shown, the detection unit 410 includes a sampling unit 41 and an oil analysis unit 42. The sampling unit 41 is connected to the production process control unit 50 and the oil analysis unit 42, respectively. The sampling unit 41 includes multiple sampling probes 411 and an automatic sampler 412. The sampling probes 411 sample the product oil, which includes direct coal liquefaction oil, hydrotreated oil, hydrotreated and upgraded oil, heavy fraction oil, aviation kerosene, and diesel oil. The automatic sampler 412 is connected to the sampling probes 411 and distributes the product oil according to a preset ratio into test bottles, storage bottles, and backup bottles. The product analysis unit 42 includes a viscosity sensor 421, a multi-channel solvent injection pump 422, and a differential scanning calorimeter 423. The viscosity sensor 421 is connected to both the sampling unit 41 and the multi-channel solvent injection pump 422. The differential scanning calorimeter 423 is connected to the multi-channel solvent injection pump 422. The viscosity sensor 421 is used to detect the viscosity of the product oil. The multi-channel solvent injection pump 422 adjusts the viscosity of the product oil according to a preset concentration to obtain sample oil. The differential scanning calorimeter 423 is used to acquire the physical properties of the sample oil, including at least the freezing point and smoke point. The sampling unit 41 deploys sampling probes 411 at key nodes to collect oil samples in real time. An automatic sampler 412 distributes the samples proportionally (e.g., 60%, 30%, and 10%) to testing, storage, and backup bottles to ensure the representativeness of the samples. The viscosity sensor 421 and multi-channel solvent injection pump 422 of the oil analysis unit 42 work together to automatically adjust the viscosity of the sample oil to a suitable concentration for testing (e.g., 300 mPa·s), improving the accuracy and efficiency of the test. The differential scanning calorimeter 423 is used to determine the freezing point, smoke point, and other physical properties of the sample oil, assisting in the optimization of conditions in the hydrogenation unit and ensuring the low-temperature performance and combustion characteristics of aviation kerosene.

[0085] In the above embodiments, such as Figure 5 and Figure 6As shown, the sampling probe 411 performs point sampling of direct coal liquefaction oil, hydrorefining unit product oil, hydromodification unit product oil, light oil, aviation kerosene, and diesel oil in the oil delivery unit 70, hydrorefining unit 51, hydromodification unit 52, atmospheric distillation unit 53, and vacuum distillation unit 54 to obtain product oil, adapting to different oil viscosities and pressure conditions; the automatic sampler 412, equipped with a high-precision mass flow meter, distributes the product oil to test bottles (60%), storage bottles (30%), and backup bottles (10%) according to a preset ratio. The sampling unit also includes an electronic tag writing unit 413 and a self-cleaning unit 414. The electronic tag writing unit 413 engraves a unique ID (a 128-bit matrix code containing a timestamp and process unit code) on the sample bottle using laser-induced breakdown spectroscopy technology and binds it to the electronic tag. The self-cleaning unit 414 performs nitrogen purging during the sampling interval and pulse cleaning of the pipeline at a pressure of 0.6 MPa to avoid cross-contamination.

[0086] In the above embodiments, such as Figure 7 As shown, the oil analysis unit also includes a sampling robotic arm 424, which is connected to the sampling unit 41 to perform operations such as opening, transferring, and cleaning sample bottles. An RFID radio frequency identifier 425 transfers the sampled oil through the sampling robotic arm 424, scans the electronic tag to identify and record the sampled oil information, and automatically uploads the oil source and sampling time to the intelligent central unit. A viscosity sensor 421 transfers the scanned sampled oil through the sampling robotic arm 424, detects the oil sample viscosity online, and for high-viscosity oils, a multi-channel solvent injection pump 422 automatically injects tetrahydrofuran diluent, which is adjusted in real time by the viscosity sensor 421 and uploaded to the intelligent central unit.

[0087] like Figure 7 As shown, the viscosity-adjusted oil is cooled at a constant rate in a differential scanning calorimeter 423 until it completely solidifies. During solidification, the oil releases latent heat, and an exothermic peak appears on the curve displayed by the differential scanning calorimeter 423. The exothermic peak is integrated using the differential scanning calorimeter software to calculate the enthalpy of crystallization and determine the freezing point. The oil is then heated at a constant rate in the differential scanning calorimeter 423 until it reaches a high temperature. The oil's oxidation reaction releases heat, and an exothermic peak appears on the aforementioned curve. By comparing the above curve with a standard smoke point test, the correspondence between the decomposition peak's starting temperature and the smoke point is determined. The smoke point of the oil is then determined and uploaded to the intelligent central unit 60.

[0088] In some alternative implementations, such as Figure 7As shown, the oil analysis unit also includes a fully automated distillation range analyzer 426 and a two-dimensional gas chromatograph 427. The fully automated distillation range analyzer 426 is connected to a multi-channel solvent injection pump 422. The fully automated distillation range analyzer 426 is used to monitor the distillation range distribution of the products from the hydrogenation and distillation units. The two-dimensional gas chromatograph 427 is used to obtain the component content parameters of the sample oil, which include at least the aromatic content, hydrogenated aromatic content, and alkane content. The fully automated distillation range analyzer 426 monitors the distillation range distribution of the sampled oils from the hydrorefining unit, hydromodification unit, atmospheric distillation unit, and vacuum distillation unit in real time and automatically uploads the data to the intelligent central unit. The distillation range of the oil after viscosity adjustment is detected in the fully automated distillation range analyzer 426. The oil sample is taken, zeolite is added to prevent bumping, and the oil is heated. The vapor liquefies through a condenser, and the droplets trigger a sensor to record the corresponding temperature. A temperature-distillation volume curve is plotted in real time, and key distillation points are marked. When the residual liquid in the flask stops distilling, the final boiling point is recorded and uploaded to the intelligent central unit. The distillation range analyzer can monitor the distillation range distribution of the product oil in real time, including the initial boiling point, distillation temperatures and final boiling points at different distillation range distributions (e.g., 10%, 30%, 50%, 70%, and 90%). This data is crucial for adjusting the vacuum, temperature, and reflux ratio of the rectification unit, helping to maintain the high quality and high yield of aviation kerosene. A full two-dimensional gas chromatograph 427 is used to detect the cycloalkanes, aromatics, and sulfur and nitrogen compounds in the viscosity-adjusted oil sample. The cycloalkanes, aromatics, and sulfur and nitrogen compounds in the oil sample are also detected using a full two-dimensional gas chromatograph 427 coupled with a time-of-flight mass spectrometer (e.g., GC×GC-TOFMS). Hydrocarbon compounds in petroleum products exhibit a regular distribution according to chemical category on a two-dimensional plane of a two-dimensional gas chromatography-mass spectrometry (GC×GC) system. Their proportions are calculated through image analysis. Alkanes, cycloalkanes, and aromatics elute in order of boiling point. The distribution regions of alkanes, cycloalkanes, and aromatics are determined using standards or literature data. GC×GC data processing software is used to integrate all peaks within a specific region, calculating the percentage of each category's peak area relative to the total hydrocarbon peak area. This peak area proportion directly reflects the aromatic content. Sulfur and nitrogen compounds are distributed in specific regions on the two-dimensional plane. Qualitative analysis is performed by comparing retention times with standards and matching mass spectrometry libraries. The total sulfur content is calculated by integrating the peak areas and comparing the results with a standard curve. All of the above data is simultaneously uploaded to the intelligent central unit.

[0089] The input parameters for the intelligent central unit of the LSTM model are shown in Table 1.

[0090] Table 1

[0091]

[0092] The aforementioned equipment uses a sampling unit to collect real-time data on coal direct liquefaction oil, hydrorefining unit product oil, hydroreforming unit product oil, light oil, aviation kerosene, and diesel oil from the hydrotreating reactor (hydrotreating unit) and distillation column (distillation unit). It relies on an oil analysis unit to perform high-frequency detection of key indicators such as oil viscosity, sulfur and nitrogen content, distillation range distribution, and alkane content, and synchronizes the data stream to the intelligent central unit, enabling the intelligent central unit to execute intelligent control methods. Considering the characteristics of coal liquefaction oil—a high cycloalkanes content of 85%–95% and an aromatics content as low as 1%–5%—the equipment's built-in hydrogenation catalyst activity compensation algorithm automatically triggers the temperature gradient compensation program in the hydroreforming unit when it detects fluctuations in feedstock alkane levels exceeding ±3%. By adjusting the axial temperature difference of the reactor bed, it maintains the cycloalkanes ring-opening efficiency of the catalyst. In the distillation stage, the equipment combines mass spectrometry analysis data of the top components of the vacuum distillation column to construct a dynamic equation for tray efficiency and reflux ratio. When the sum of C11 and heavy cycloalkanes in the jet fuel fraction exceeds 1.2%, the vacuum level is adjusted to below 15 kPa and the side stream recovery ratio is optimized to stably control the final boiling point within the range of 270 ± 2℃. Through closed-loop data verification between the production process control unit and the oil product analysis unit, the intelligent control equipment can achieve self-correction of process parameters within 8 minutes, with a jet fuel yield standard deviation ≤ 0.3 wt% and density fluctuation range compressed to 0.0005 g / cm³. 3 The system ensures that even with fluctuations of ±5% in feedstock composition, it can still continuously output No. 3 jet fuel that meets the GB6537-2018 standard.

[0093] The properties of the coal direct liquefaction oils used are listed in Table 2. The catalyst used in the hydrorefining unit is Ni-Mo / Al2O3, and the catalyst used in the hydromodification unit is Ni-W / Al2O3. These catalysts are loaded into the hydrorefining and hydromodification units respectively, with a catalyst volume ratio of 1:1. The coal direct liquefaction oil with the properties shown in Table 1 is mixed with hydrogen, preheated in a heater, and then enters the hydrorefining unit to contact the hydrorefining catalyst. The hydrorefined oil is then mixed with hydrogen again and enters the hydromodification unit to contact the hydromodification catalyst. After separation and fractionation in the distillation unit, gaseous products, diesel fraction, and aviation kerosene fraction are obtained. The hydrogen-rich gas is returned to the reaction unit for further hydrogenation.

[0094] When direct coal liquefaction oil is pumped from the oil conveying unit to the hydrogenation unit, the sampling unit collects oil samples at each node in real time using a high-pressure sampling probe. Simultaneously, the oil analysis unit performs viscosity testing, distillation range analysis, and component determination. The obtained data is uploaded to the intelligent central unit, which integrates production process parameters, equipment status, and oil quality data in real time. Based on the LSTM algorithm, it predicts the jet fuel yield trend for the next two hours and constructs a three-dimensional temperature field model of the reactor using the PINN algorithm. After verification through a virtual reactor, optimized parameters are generated. The adaptive decision engine transmits control commands through the TSN network. At each production unit, when a fluctuation in cycloalkanes exceeds ±3%, the bed gradient temperature control program in the hydrotreating unit is triggered. In the vacuum distillation stage, the vacuum level (<15 kPa) and final boiling point control (270 ± 2℃) are dynamically optimized based on online mass spectrometry data. Simultaneously, the oil analysis unit achieves full lifecycle traceability of samples through laser coding and RFID binding. This ultimately forms an intelligent control closed loop of data acquisition, model prediction, parameter optimization, and execution feedback, ensuring that jet fuel yield remains stable at 39.0 ± 0.3 wt% and density fluctuation ≤ 0.0005 g / cm³ under conditions of ±5% fluctuation in feedstock composition. 3 .

[0095] Examples 1, 2, and 3 in Tables 2-5 all used the optimal reaction conditions. The properties of the feedstock oil, optimized process conditions, and aviation kerosene product are listed in Tables 2 (Properties of Direct Coal Liquefaction Oil), 3 (Reaction Conditions of the Hydrorefining Unit), 4 (Reaction Conditions of Hydroreforming), and 5 (Properties of Aviation Kerosene), respectively. Table 2 shows that the direct coal liquefaction oil has high S and N content. Tables 3 and 4 show that the equipment can control the aviation kerosene production conditions within the optimal reaction conditions. Table 5 shows that the final aviation kerosene fraction has very low sulfur and nitrogen content and low aromatic content, achieving the purpose of hydrogenation. The final aviation kerosene density is 0.8818 g / cm³. 3 With a calorific value of over 45.6 MJ / kg and a yield of around 39.0 wt%, the goal of stable and high-quality aviation kerosene production has been achieved.

[0096] Note: Aviation kerosene yield (wt%) = aviation kerosene mass / coal direct liquefaction oil mass.

[0097] Table 2

[0098]

[0099]

[0100] Table 3

[0101] process parameters Example 1 Example 2 Example 3 Hydrogen partial pressure, MPa 8.0～20.0 8.0～19.0 11.0～20.0 Reaction temperature, °C 320～400 330～400 320～390 <![CDATA[WHSV of the main agent, h -1 > 0.4～3.5 0.4～3.5 0.4～3.5 Hydrogen-to-oil ratio, v / v 600～1000 600～1000 600～1000

[0102] Table 4

[0103] process parameters Example 1 Example 2 Example 3 Hydrogen partial pressure, MPa 8.0～18.0 7.0～18.0 7.0～18.0 Reaction temperature, °C 320～400 330～400 320～400 <![CDATA[Main agent space velocity, h -1 > 0.3～2.8 0.3～2.8 0.3～2.8 Hydrogen-to-oil ratio, v / v 600～1000 600～1000 600～1000

[0104] Table 5

[0105]

[0106]

[0107] This application provides a computer-readable storage medium including a stored program, wherein an intelligent control method for controlling the device where the computer-readable storage medium is located to execute an aviation kerosene preparation system is provided when the program is running.

[0108] Specifically, the intelligent control methods for aviation kerosene preparation systems include:

[0109] Step S1: Obtain various sample data, including: physical property parameters, distillation range distribution parameters, component content parameters of sample oil, and process parameters of aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of coal direct liquefaction oil, and substances after gas-liquid separation of hydrogenation products, including aviation kerosene.

[0110] Step S2: Based on multiple sample data, long short-term memory network algorithm and physical information neural network algorithm, establish a prediction algorithm model. The prediction algorithm model is used to predict the yield of aviation kerosene in the aviation kerosene preparation system.

[0111] Step S3: Based on the prediction algorithm model and multiple sample data, determine the predicted yield of aviation kerosene within the first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period. The start time of the first preset time period is the time when multiple sample data are obtained. The operating parameters include at least the temperature, pressure and gas concentration of the aviation kerosene preparation system during operation.

[0112] Step S4: Adjust the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and multiple operating parameters.

[0113] This application provides a processor for running a program, wherein the program executes an intelligent control method for an aviation kerosene preparation system.

[0114] Specifically, the intelligent control methods for aviation kerosene preparation systems include:

[0115] Step S1: Obtain various sample data, including: physical property parameters, distillation range distribution parameters, component content parameters of sample oil, and process parameters of aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of coal direct liquefaction oil, and substances after gas-liquid separation of hydrogenation products, including aviation kerosene.

[0116] Step S2: Based on multiple sample data, long short-term memory network algorithm and physical information neural network algorithm, establish a prediction algorithm model. The prediction algorithm model is used to predict the yield of aviation kerosene in the aviation kerosene preparation system.

[0117] Step S3: Based on the prediction algorithm model and multiple sample data, determine the predicted yield of aviation kerosene within the first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period. The start time of the first preset time period is the time when multiple sample data are obtained. The operating parameters include at least the temperature, pressure and gas concentration of the aviation kerosene preparation system during operation.

[0118] Step S4: Adjust the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and multiple operating parameters.

[0119] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program having at least the following method steps:

[0120] Step S1: Obtain various sample data, including: physical property parameters, distillation range distribution parameters, component content parameters of sample oil, and process parameters of aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of coal direct liquefaction oil, and substances after gas-liquid separation of hydrogenation products, including aviation kerosene.

[0121] Step S2: Based on multiple sample data, long short-term memory network algorithm and physical information neural network algorithm, establish a prediction algorithm model. The prediction algorithm model is used to predict the yield of aviation kerosene in the aviation kerosene preparation system.

[0122] Step S3: Based on the prediction algorithm model and multiple sample data, determine the predicted yield of aviation kerosene within the first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period. The start time of the first preset time period is the time when multiple sample data are obtained. The operating parameters include at least the temperature, pressure and gas concentration of the aviation kerosene preparation system during operation.

[0123] Step S4: Adjust the process parameters of the aviation kerosene preparation system based on the predicted yield of aviation kerosene and multiple operating parameters.

[0124] Obviously, those skilled in the art should understand that the modules or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0125] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0127] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0128] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0129] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0130] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0131] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0132] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0134] The above are merely preferred embodiments of the present application and are not intended to limit the present application. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. An intelligent control method for an aviation kerosene preparation system, characterized in that, include: Acquire various sample data, including: physical property parameters, distillation range distribution parameters, and component content parameters of the sample oil, as well as process parameters of the aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of the coal direct liquefaction oil, and substances after gas-liquid separation of the hydrogenation products, including aviation kerosene. Based on various sample data, long short-term memory network algorithm, and physical information neural network algorithm, a prediction algorithm model is established. The prediction algorithm model is used at least to predict the yield of aviation kerosene in the aviation kerosene preparation system. Based on the prediction algorithm model and various sample data, the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period are determined. The start time of the first preset time period is the time when the various sample data are acquired. The operating parameters include at least the temperature, pressure, and gas concentration during the operation of the aviation kerosene preparation system. The process parameters of the aviation kerosene preparation system are adjusted based on the predicted yield of the aviation kerosene and multiple operating parameters.

2. The intelligent control method according to claim 1, characterized in that, The prediction algorithm model also includes a screening model. The step of establishing the prediction algorithm model based on various sample data, long short-term memory network algorithms, and physical information neural network algorithms includes: Based on the filtering model, the various input sample data are filtered, and the filtering model is as follows: f t =σ(W f ·[h t-1 ,x t ]+b f ), Among them, f t The data of the various samples remaining after screening; W f Let b be the weight matrix. f Here, σ is the bias vector; h is the sigmoid activation function; t-1 To represent the hidden state at the previous time step (t-1); x t The sample data; Based on the various sample data remaining after screening and the Long Short-Term Memory Network algorithm, a Long Short-Term Memory Network model is established; A physical information neural network model is established based on the energy conservation equation, the mass conservation equation, and the pyrolysis reaction rate equation. The prediction algorithm model is established based on the Long Short-Term Memory Network model and the Physical Information Neural Network model.

3. The intelligent control method according to claim 1, characterized in that, The intelligent control method further includes: Multiple process data and multiple test data of the coal direct liquefaction oil are obtained. The process data includes production process unit parameters, fluid transport parameters and equipment status parameters. The test data includes physical property parameters, distillation range distribution and component content. The prediction algorithm model is updated based on the process data and the test data.

4. An intelligent control device for an aviation kerosene preparation system, characterized in that, include: The acquisition module is used to acquire various sample data, including: physical property parameters, distillation range distribution parameters, and component content parameters of the sample oil, as well as process parameters of the aviation kerosene preparation system. The sample oil includes diluted coal direct liquefaction oil, hydrogenation products of the coal direct liquefaction oil, and substances after gas-liquid separation of the hydrogenation products. A module is established to build a prediction algorithm model based on various sample data, a long short-term memory network algorithm, and a physical information neural network algorithm. The prediction algorithm model is used to predict at least the yield of aviation kerosene in the aviation kerosene preparation system. The determination module is used to determine the predicted yield of aviation kerosene within a first preset time period and multiple operating parameters of the aviation kerosene preparation system within the first preset time period based on the prediction algorithm model and multiple sample data. The start time of the first preset time period is the time when multiple sample data are acquired, and the types of operating parameters are the same as the types of process parameters. The control module is used to adjust the process parameters of the aviation kerosene preparation system based on the predicted yield of the aviation kerosene and multiple operating parameters.

5. An intelligent control device for an aviation kerosene preparation system, characterized in that, include: The production process control unit includes a hydrogenation unit and a distillation unit. The distillation unit is connected to the hydrogenation unit. The hydrogenation unit performs hydrogenation and upgrading treatment on coal direct liquefaction oil to obtain the hydrogenated product of the coal direct liquefaction oil. The distillation unit performs gas-liquid separation on the hydrogenated product. The detection unit is connected to the production process control unit to monitor the process parameters of the aviation kerosene preparation system and to detect at least the sample oil to obtain various sample data. The sample oil includes: diluted coal direct liquefaction oil, hydrogenation products, and substances obtained after gas-liquid separation of the hydrogenation products. The intelligent central unit is connected to the detection unit and the production process control unit, and is used to execute the intelligent control method of the aviation kerosene preparation system according to any one of claims 1 to 3.

6. The intelligent control device according to claim 5, characterized in that, The hydrogenation unit includes a hydrogenation refining unit and a hydrogenation reforming unit, and the distillation unit includes an atmospheric distillation unit and a vacuum distillation unit, wherein: The hydrorefining unit is connected to the hydroreforming unit. The hydrorefining unit is used to remove impurities from the direct coal liquefaction oil to obtain hydrorefined oil and hydrorefined circulating gas. The hydrotreating unit is used to treat the hydrorefined oil and the hydrorefined recycle gas to obtain hydrorefined oil and hydrorefined recycle gas, wherein the hydrorefined oil, the hydrorefined recycle gas, the hydrorefined oil and the hydrorefined recycle gas are the hydrotreating products. The atmospheric distillation unit is connected to the hydrotreating unit and the vacuum distillation unit respectively. The atmospheric distillation unit is used to perform gas-liquid separation on the hydrotreating oil to obtain heavy fraction oil. The vacuum distillation unit is used to perform gas-liquid separation on the heavy oil fraction to obtain aviation kerosene and diesel.

7. The intelligent control device according to claim 6, characterized in that, The detection unit includes a sampling unit and an oil analysis unit. The sampling unit is connected to both the production process control unit and the oil analysis unit, wherein: The sampling unit includes multiple sampling probes and an automatic sampler. The sampling probes sample the product oil, which includes the direct coal liquefaction oil, the hydrorefined oil, the hydrotreated oil, the heavy fraction oil, aviation kerosene, and the diesel oil. The automatic sampler is connected to the sampling probes and distributes the product oil into test bottles, storage bottles, and backup bottles according to a preset ratio. The oil analysis unit includes a viscosity sensor, a multi-channel solvent injection pump, and a differential scanning calorimeter. The viscosity sensor is connected to both the sampling unit and the multi-channel solvent injection pump. The differential scanning calorimeter is connected to the multi-channel solvent injection pump. The viscosity sensor is used to detect the viscosity of the product oil. The multi-channel solvent injection pump adjusts the viscosity of the product oil according to a preset concentration to obtain the sample oil. The differential scanning calorimeter is used to acquire the physical property parameters of the sample oil, which include at least the freezing point and the smoke point.

8. The intelligent control device according to claim 7, characterized in that, The oil analysis unit also includes a fully automated distillation range analyzer and a two-dimensional gas chromatograph. The fully automated distillation range analyzer is connected to the multi-channel solvent injection pump. The fully automated distillation range analyzer is used to monitor the distillation range distribution of the products of the hydrogenation unit and the distillation unit. The two-dimensional gas chromatograph is used to obtain the component content parameters of the sample oil. The component content parameters include at least: aromatic content, hydrogenated aromatic content, and alkane content.

9. A processor, characterized in that, The processor is used to run a program, wherein the program executes the intelligent control method for the aviation kerosene preparation system according to any one of claims 1 to 3.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the intelligent control method for the aviation kerosene preparation system according to any one of claims 1 to 3.

Images