Emergency scheduling and production scheduling method and device for refinery production

Abstract

The application discloses a refinery production emergency scheduling and production scheduling method and device, and the method comprises the following steps: acquiring real-time data of core operation indexes of all devices in a pre-established refinery device relationship tree; the refinery device relationship tree is established by using a knowledge graph technology according to the types of devices in the refinery and the production relationship between the devices; the core operation index is an operation index with an influence degree on the refinery production higher than a preset influence degree; when it is judged that the real-time data is abnormal according to a preset abnormality judgment algorithm, and it is determined that automatic production scheduling calculation is enabled according to preconfigured production scheduling parameters corresponding to the core operation index, the automatic production scheduling calculation step is executed, a reasonable fault emergency scheduling and production scheduling scheme at the whole-plant level can be quickly generated, the emergency scheduling and production scheduling efficiency, initiative and timeliness are improved, and the production delay risk is reduced.

Description

Technical Field

[0001] This invention relates to the field of emergency production scheduling technology in refineries, and particularly to a method and apparatus for emergency production scheduling in refineries. Background Technology

[0002] This section is intended to provide background or context for embodiments of the present invention. The description herein is not intended to imply that it is prior art simply because it is included in this section.

[0003] As petrochemical enterprises develop towards integrated refining and chemical production, their processing scale is large, their equipment is numerous, and their processes are complex. At the same time, these plants are characterized by high temperature and pressure, as well as flammability and explosiveness, making them prone to production fluctuations and abnormal accidents.

[0004] However, existing emergency scheduling methods have some significant shortcomings in practical applications. Firstly, traditional scheduling methods typically involve manual troubleshooting and repair after a failure occurs. This troubleshooting and repair phase leads to prolonged production stoppages, impacting enterprise stability. Furthermore, the lack of systematic data analysis and scientific model support results in slow emergency response and an inability to promptly generate plant-wide material and energy balance scheduling plans. Moreover, in refinery emergency scheduling technology, it is impossible to make reasonable production arrangements based on the failure situation or to perform appropriate production line matching, thereby reducing production scheduling efficiency and increasing the risk of production delays. Therefore, how to quickly generate reasonable plant-wide emergency scheduling plans to improve the efficiency, proactivity, and timeliness of emergency scheduling and reduce the risk of production delays is an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a refinery emergency production scheduling method to quickly generate a reasonable plant-wide emergency production scheduling plan, improving the efficiency, initiative, and timeliness of emergency production scheduling, and reducing the risk of production delays. The method includes:

[0006] Obtain real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; the refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between units; the core operating indicators are operating indicators whose impact on refinery production is higher than the preset impact level.

[0007] When a real-time data anomaly is detected by a preset anomaly detection algorithm, and automatic production scheduling is enabled based on pre-configured production scheduling parameters corresponding to core operating indicators, the automatic production scheduling calculation steps are executed. The preset anomaly detection algorithm is determined based on the data type of the real-time data. The production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is detected to be anomaly.

[0008] Automatic production scheduling calculations include:

[0009] Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree;

[0010] Based on the abnormal device, the devices associated with the abnormal device, and the pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data.

[0011] After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated.

[0012] The results of the automatic production scheduling calculations were used as the emergency production scheduling plan for the refinery.

[0013] This invention also provides a refinery emergency production scheduling device for quickly generating reasonable plant-wide emergency production scheduling plans, improving the efficiency, initiative, and timeliness of emergency production scheduling, and reducing the risk of production delays. The device includes:

[0014] The acquisition module is used to acquire real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; the refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between units; the core operating indicators are operating indicators whose impact on refinery production is higher than the preset impact level.

[0015] The execution module is used to execute the automatic production scheduling calculation steps when the real-time data is determined to be abnormal according to the preset abnormality judgment algorithm and the automatic production scheduling calculation is enabled according to the pre-configured production scheduling parameters corresponding to the core operating indicators. The preset abnormality judgment algorithm is determined according to the data type of the real-time data. The production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is determined to be abnormal.

[0016] Automatic production scheduling calculations include:

[0017] Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree;

[0018] Based on the abnormal device, the devices associated with the abnormal device, and the pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data.

[0019] After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated.

[0020] The results of the automatic production scheduling calculations were used as the emergency production scheduling plan for the refinery.

[0021] Compared with existing emergency production scheduling solutions for refineries, this invention obtains real-time data of core operating indicators for all units in a pre-established refinery unit relationship tree. The refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between them. Core operating indicators are those whose impact on refinery production exceeds a preset impact level. When a preset anomaly detection algorithm determines that the real-time data is abnormal, and the pre-configured scheduling parameters corresponding to the core operating indicators determine whether to activate automatic production scheduling calculation, the automatic production scheduling calculation steps are executed. The preset anomaly detection algorithm is determined based on the data type of the real-time data. The scheduling parameters are used to determine whether to activate automatic production scheduling calculation when the real-time data of the core operating indicators is determined to be abnormal. Automatic production scheduling calculation includes: determining the units associated with the abnormal units corresponding to the abnormal real-time data based on the refinery unit relationship tree; determining the units associated with the abnormal units based on the abnormal units and the units associated with the abnormal units; and... A pre-configured table defines the calculation model, calculation model parameters, calculation duration, and recalculation interval. The table includes the correspondence between all units in the refinery's unit relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The goal of the mathematical programming model is to minimize the impact of abnormal real-time data. After the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model for the calculation duration, a recalculation interval is performed after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again for the calculation duration, generating an automatic production scheduling calculation result. This automatic production scheduling calculation result is used as the refinery's emergency production scheduling plan, which can quickly generate a reasonable plant-wide emergency production scheduling plan, improving the efficiency, initiative, and timeliness of emergency production scheduling and reducing the risk of production delays. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0023] Figure 1 This is a flowchart of an emergency production scheduling method for refineries provided in an embodiment of the present invention;

[0024] Figure 2 This is a flowchart illustrating a specific example of an emergency production scheduling method for refineries provided in this embodiment of the invention.

[0025] Figure 3 This is a schematic diagram of the structure of an emergency production scheduling device for a refinery provided in an embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram of a specific example of a refinery emergency production scheduling device provided in an embodiment of the present invention;

[0027] Figure 5 This is a schematic diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0029] The acquisition, storage, use, and processing of data in this application comply with relevant laws and regulations.

[0030] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0031] In the description of this specification, the terms "comprising," "including," "having," and "containing" are open-ended terms, meaning that they include but are not limited to. The terms "an embodiment," "a specific embodiment," "some embodiments," and "for example," etc., refer to specific features, structures, or characteristics described in connection with that embodiment or example that are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. The order of steps involved in the various embodiments is used to illustrate the implementation of this application, and the order of steps is not limited and can be adjusted appropriately as needed.

[0032] Based on existing technologies, the urgent problem to be solved is how to quickly generate plant-wide material and energy balance production scheduling plans and improve the initiative and timeliness of emergency scheduling.

[0033] In view of the problems existing in the prior art, the purpose of this invention is to automatically trigger an emergency dispatch plan when the core operating indicators of the monitoring device are detected to exceed a preset threshold. This involves retrieving the preset calculation model and parameters of the relevant device and performing multi-dimensional production scheduling simulation calculations based on the model algorithm. The resulting optimized production scheduling plan will be pushed to the plan's sender via in-system messaging and instant messaging, ensuring that key users receive it immediately and providing data support and plan reference for subsequent scheduling decisions.

[0034] Figure 1 This is a flowchart of a refinery emergency production scheduling method provided in an embodiment of the present invention, such as... Figure 1 As shown, the method may include:

[0035] Step 101: Obtain real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; the refinery unit relationship tree is established using knowledge graph technology based on the type of units in the refinery and the production relationship between units; the core operating indicators are operating indicators whose impact on refinery production is higher than the preset impact level.

[0036] Step 102: When the real-time data is determined to be abnormal according to the preset anomaly judgment algorithm, and the automatic production scheduling calculation is enabled according to the pre-configured production scheduling parameters corresponding to the core operating indicators, the automatic production scheduling calculation step is executed; the preset anomaly judgment algorithm is determined according to the data type of the real-time data; the production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is determined to be abnormal.

[0037] Automatic production scheduling calculations include:

[0038] Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree;

[0039] Based on the abnormal device, the devices associated with the abnormal device, and the pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data.

[0040] After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated.

[0041] The results of the automatic production scheduling calculations were used as the emergency production scheduling plan for the refinery.

[0042] Compared with existing technologies, the emergency scheduling function of this invention monitors the core operating indicators of the equipment in real time. When a parameter exceeds a preset threshold, it automatically triggers an emergency scheduling plan, retrieves the preset configuration parameters and calculation models of the relevant equipment, and calculates a multi-dimensional emergency scheduling plan based on the model algorithm. The final optimized scheduling plan will be pushed to relevant responsible persons through in-station messages and instant messaging, ensuring that key users receive it as soon as possible. This provides data support and plan reference for subsequent scheduling decisions, quickly generates a reasonable plant-wide emergency scheduling plan, improves the efficiency, initiative, and timeliness of emergency scheduling, and reduces the risk of production delays.

[0043] In one embodiment, before acquiring real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree, the refinery production emergency scheduling method further includes: acquiring the types of units in the refinery and the production relationships between units; the production relationships between units include the impact relationship between units after the real-time data of a core operating indicator of a unit is abnormal; and using knowledge graph technology to establish a refinery unit relationship tree based on the types of units in the refinery and the production relationships between units.

[0044] In this embodiment, real-time data of the core operating indicators of all devices in the device relationship tree can be obtained. The device relationship tree can be established using a knowledge graph based on the type, production relationship, and influence relationship between devices, with devices as entity nodes, type as attribute, and production relationship and influence relationship as the relationship between nodes. The device relationship tree is used to determine the calculation model and the calculation model parameters. One approach is to determine it based on the type of devices. For example, devices A and C are devices of the same type with the same production function. If device A fails and stops production, the calculation model can use the device relationship tree to determine multiple devices that can compensate for the impact of device A's stoppage, namely devices B and C. Another approach is to determine it based on the production relationship or influence relationship between devices. For example, if device B is the downstream device of device A, and the real-time data of device A's core operating indicator is abnormal, causing a stoppage, then device B is affected by device A's stoppage. The calculation model parameters need to add device B's operating indicators and adjust device B, for example, by replacing device B's upstream device with device C.

[0045] In one embodiment, the core operating indicators are those whose impact on refinery production exceeds a preset level. These core operating indicators are also key performance indicators (KPIs) for plant operation, and a pre-trained core operating indicator identification model can be used to select them from all operating indicators. The core operating indicator identification model is pre-trained using a training set to classify the model. The test set is then input into the trained core operating indicator identification model to obtain the core operating indicators used for subsequent production scheduling calculations.

[0046] In this embodiment, the core operating indicators are obtained by inputting the test set into a pre-trained core operating indicator recognition model; the core operating indicator recognition model is obtained by training a classification model using a training set; the test set consists of the data of all operating indicators of all devices in the pre-established refinery unit relationship tree for the first historical time period and the refinery production failure information for the first historical time period; the training set consists of the data of all operating indicators of all devices in the pre-established refinery unit relationship tree for the second historical time period and the refinery production failure information for the second historical time period.

[0047] In this embodiment, the test set consists of data for the first historical time period of all operating indicators of all units in a pre-established refinery unit relationship tree, along with refinery production failure information for that first historical time period. The training set consists of data for the second historical time period of all operating indicators of all units in the pre-established refinery unit relationship tree, along with refinery production failure information for that second historical time period. For example, if the current month is December, the training set could be data for all operating indicators of all units in the refinery unit relationship tree from June to August of this year, and the test set could be data for all operating indicators of all units in the refinery unit relationship tree from September to November of this year. Specifically, the training and test sets can be labeled data. During training, the classification model learns the characteristics of the labeled indicator data in the training set, i.e., the characteristics of the core operating indicators. Data from recent months is input into the trained core operating indicator recognition model to ensure comprehensive and accurate acquisition of core operating indicator data for all units in the refinery, providing a reliable data foundation for subsequent anomaly detection and automatic production scheduling calculations. In one embodiment, if the data type of the real-time data is numerical, the preset anomaly detection algorithm is a dynamic threshold algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is Boolean, the preset anomaly detection algorithm is a classification algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is an image, the preset anomaly detection algorithm is one or any combination of an edge detection algorithm, a regression-based deep learning algorithm, and a classification-based deep learning algorithm.

[0048] In this embodiment, when real-time data is determined to be abnormal according to a preset anomaly judgment algorithm, and automatic production scheduling is enabled based on pre-configured production scheduling parameters corresponding to core operating indicators, the automatic production scheduling calculation steps are executed, including: comparing real-time data with a dynamic threshold determined using a dynamic threshold algorithm, and judging real-time data as abnormal based on the comparison result; inputting real-time data into a recognition model pre-trained using an artificial intelligence algorithm to output recognition results, and judging real-time data as abnormal based on the recognition results; classifying real-time data using a classification algorithm, and judging real-time data as abnormal based on the classification results; analyzing the statistical or texture features of real-time data using an edge detection algorithm, and judging real-time data as abnormal based on the analysis results; encoding normal images using a regression-based deep learning network to learn the features of normal images, and judging real-time data as abnormal based on the learned regression-based deep learning network; and / or, learning the features of normal and abnormal images using a classification-based deep learning network, and judging real-time data as abnormal based on the learned classification-based deep learning network; and determining automatic production scheduling is enabled based on pre-configured production scheduling parameters corresponding to core operating indicators, and executing the automatic production scheduling calculation steps.

[0049] In one embodiment, a relationship table between data types and preset anomaly detection algorithms can be pre-configured, allowing different algorithms to be used for anomaly detection based on data type. For example, for numerical data such as temperature and load: 1. A dynamic threshold algorithm can be used, with the dynamic threshold dynamically updated based on historical threshold data and an artificial intelligence or machine learning model; 2. Alternatively, historical data and an artificial intelligence or machine learning model can be directly used for anomaly detection. For Boolean data (only taking 0 and 1 values): Algorithm learning can also be used. Based on historical data, the model learns that when the device is operating normally, index a = 1; when index a = 0 is detected, the model determines that the data of index a is abnormal. For image data: Edge detection algorithms, regression-based deep learning algorithms, or classification-based deep learning algorithms are used. Edge detection algorithms analyze the statistical or texture features of real-time data, and real-time data anomalies are determined based on the analysis results; regression-based deep learning networks encode normal images, learn the features of normal images, and the learned regression-based deep learning network determines real-time data anomalies; and / or, classification-based deep learning networks learn the features of normal and abnormal images, and the learned classification-based deep learning network determines the real-time data anomalies. In addition, algorithms for other data types can be added. This step enables accurate anomaly detection for different types of real-time data, improving the accuracy and reliability of anomaly detection.

[0050] In one embodiment, when real-time data is determined to be abnormal according to a preset anomaly judgment algorithm, and automatic production scheduling is enabled based on pre-configured production scheduling parameters corresponding to core operating indicators, the automatic production scheduling calculation steps are executed, including: after determining that real-time data is abnormal according to the preset anomaly judgment algorithm, determining one or any combination of the abnormal real-time data collection method, collection device, data source, and detection environment; determining whether one or any combination of the abnormal real-time data collection method, collection device, data source, and detection environment is abnormal, and generating an anomaly judgment result; if the anomaly judgment result is abnormal, it indicates that the real-time data anomaly is caused by a detection problem and is not real, and the alarm parameter is adjusted to not alarm; if the anomaly judgment result is not abnormal, it indicates that the real-time data anomaly is caused by an undetected problem and is real, and the alarm parameter is adjusted to alarm; enabling automatic production scheduling calculation is enabled based on pre-configured production scheduling parameters corresponding to core operating indicators; and executing the automatic production scheduling calculation steps based on the alarm parameter and production scheduling parameter.

[0051] In this embodiment, after determining that the real-time data of one or more core operating indicators of the devices is abnormal, it is also necessary to determine whether the real-time data abnormality of the core operating indicators is genuine, i.e., not caused by detection problems (such as detection environment, data source, acquisition method, acquisition equipment problems, etc.). If it is determined to be non-genuine, the alarm parameter is adjusted to not alarm, and the display on the page is green; if it is determined to be a genuine abnormality, the alarm parameter is adjusted to alarm, and the display on the page is red. Through this step, false alarms caused by detection problems can be effectively eliminated, improving the accuracy and reliability of abnormality judgment.

[0052] In one embodiment, automatic production scheduling calculation includes: determining the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree; determining the calculation model, calculation model parameters, calculation duration, and recalculation interval duration based on the abnormal devices, the devices associated with the abnormal devices, and a pre-configured configuration table; inputting the abnormal real-time data and the real-time data corresponding to the calculation model into the calculation model to calculate the calculation duration, and after the recalculation interval duration following the end time, inputting the abnormal real-time data and the real-time data corresponding to the calculation model into the calculation model again to calculate the calculation duration, and generating an automatic production scheduling calculation result; and determining the automatic production scheduling calculation result as the refinery's emergency production scheduling plan.

[0053] The configuration table includes the correspondence between all units in the refinery's unit relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model; the goal of the mathematical programming model is to minimize the impact of abnormal real-time data. Users can input the calculation period and time scale in the configuration table to determine the calculation time period, and the recalculation interval determines how long to wait between calculations to perform a second calculation, avoiding frequent alarms and repeated calculations caused by fluctuations in operating indicators. Through this step, precise scheduling of abnormal units and their related units can be achieved, minimizing the impact of anomalies and improving the stability and efficiency of refinery production.

[0054] Figure 2 This is a flowchart illustrating a specific example of an emergency production scheduling method for refineries provided in this embodiment of the invention, such as... Figure 2As shown, in one embodiment, the configuration table also includes the scheme pusher and the correspondence between the scheme pusher and the devices in the refinery unit relationship tree; the refinery production emergency scheduling method may further include: step 201, finding the scheme pusher that has a corresponding relationship with the abnormal device and the device associated with the abnormal device according to the configuration table; step 202, pushing the refinery production emergency scheduling plan to the scheme pusher through in-station message push and / or instant messaging platform. Through this step, it can be ensured that key users obtain the scheduling plan as soon as possible, providing data support and plan reference for subsequent scheduling decisions.

[0055] In one embodiment, the measurement duration, remeasurement interval, and time scale can be automatically generated based on historical data. For example, if there was a previous case where the data of indicator a of device A was abnormal, and this abnormal data of indicator a caused device A to malfunction for 7 days, then when indicator a is abnormal again, the measurement duration can be directly set to 7 days, the time scale can be 24 hours, and the remeasurement interval can be 60 minutes. The determination of the time scale and the remeasurement interval is similar to that of the measurement duration, and will not be repeated here.

[0056] In one embodiment, based on the abnormal device, the relevant devices are identified in the device relationship tree; the device input and output, device load, physical tanks, material properties, crude oil processing plan, and product demand plan of the abnormal device and related devices are obtained. Based on the above data, a scheduling optimization algorithm (which can be an existing mathematical programming algorithm) is invoked to set the maximum solution time of the model, the operation status, etc., and an emergency production scheduling plan is automatically calculated.

[0057] In one embodiment, the refinery emergency production scheduling method further includes: displaying the refinery emergency production scheduling plan from multiple dimensions; these multiple dimensions include one or any combination of the following for each scheduled unit: unit input and output, unit load, material tank farm receipt and payment, physical tank inventory, crude oil processing schedule, raw material delivery schedule, product production schedule, backfilling schedule, material flow change schedule, unit feed control, material inventory control, unit scheme switching, production and sales balance, and combined units; the refinery emergency production scheduling plan includes one or any combination of the following: unit material balance sheet, tank farm material balance report, physical tank inventory trend table, unit processing scheduling schedule, crude oil processing schedule, raw material delivery schedule, product production schedule, automotive gasoline data forecast table, automotive diesel data forecast table, hydrogen balance report, and unit material report. This step enables a multi-dimensional display of the scheduling plan, providing comprehensive and intuitive data support for production decisions.

[0058] In this embodiment, an emergency scheduling and production plan analysis report is also generated, including the production status of the equipment.

[0059] This invention also provides a refinery production emergency scheduling system, which offers the following functions:

[0060] 1. Production metric monitoring:

[0061] It enables the configuration of monitoring core operating indicators of the device, and automatically extracts real-time data of the corresponding indicators to provide alarm reminders for the monitoring indicators.

[0062] The scheduling and monitoring indicator configuration page is mainly used to configure the core operating indicators of the monitoring device. It allows you to set the scheduling and monitoring indicator name, the name of the device to which it belongs, its status, calculation formula, whether it is scheduled for production, whether it is alarmed, the upper and lower limits of the indicator parameters, the person who pushes the plan, and the push method, etc., to monitor the operation of the core indicators of the device and perform production scheduling calculations as soon as an alarm is triggered.

[0063] On this page, add new core operating indicators for the device to be monitored. By configuring corresponding formulas for the acquired real-time data, the accuracy of the core operating indicator data is ensured. Whether or not monitoring is performed is determined by whether the function is enabled. For newly added core operating indicators, the "production scheduling" function determines whether automatic production scheduling calculation is activated when the core operating indicator parameter exceeds a preset threshold. By associating the scheduling indicator with the device, the "production scheduling" configuration determines whether emergency calculation is required for that indicator. The associated device information is used to call the emergency scheduling device and bind its calculation model.

[0064] The dispatch monitoring data display page shows the core operating indicators of the device added to the dispatch monitoring indicator configuration page. Users can view the real-time values ​​of the configured indicators, the current alarm status, and the upper and lower limits of the set indicators.

[0065] When the alarm status of the device with the automatic production scheduling function is turned from green to red, the automatic production scheduling calculation will be triggered. When the alarm status is red, the calculation will only be performed once to avoid multiple calculations for a single alarm. When the alarm status turns from red to green and the indicator returns to normal, the alarm will be triggered again. The production scheduling calculation will only be performed again when the alarm status turns from green to red.

[0066] 2. Emergency production scheduling:

[0067] By configuring basic information for emergency scheduling, emergency scheduling calculations are automatically triggered when monitoring indicators trigger alarms.

[0068] The emergency dispatch and production scheduling configuration page mainly configures some key parameters for production scheduling calculation. This ensures that when the parameters exceed the preset threshold, the system can retrieve the preset configuration parameters and calculation models of the relevant devices to perform production scheduling calculations.

[0069] On the new configuration page, select the base model to be used during the calculation, enter the calculation period and time scale to determine the calculation time period. The interval between calculations is determined by the recalculation time, which avoids frequent alarms and repeated calculations caused by fluctuations in operating indicators.

[0070] Select the associated device, the data to be extracted, and the sender for this configuration. After determining which device's operating indicator alarms, call this configuration parameter. At the same time, determine which device's data to extract and the sender of the calculation results.

[0071] Based on the production indicator alarm information monitored, the system automatically establishes an emergency production scheduling plan, automatically extracts production data, and automatically continues to calculate the emergency production scheduling plan.

[0072] The specific steps are as follows:

[0073] a. When a core production indicator triggers an alarm, retrieve the device associated with the corresponding alarm indicator.

[0074] b. Based on the basic model corresponding to the device associated with the alarm indicator, automatically carry out emergency dispatch and production scheduling calculations.

[0075] c. Automatically establish calculation schemes based on alarm indicator information.

[0076] d. Automatically extract the latest production data.

[0077] e. Call the scheduling optimization algorithm (existing mathematical programming algorithm) to automatically calculate the emergency production scheduling plan.

[0078] 3. Results push notification:

[0079] The results of emergency production scheduling plans are displayed and pushed out, and the calculation results of production scheduling are presented from multiple dimensions. The calculation results are shared with others through plan sharing.

[0080] The scheduling and production calculation results support multi-dimensional presentation, enabling data visualization through unit material balance reports, tank area material balance reports, and material tank inventory trend reports. Dynamic graphical analysis is performed by combining tank inventory trend charts and Gantt charts, and the system simultaneously supports the generation of analysis reports that combine data and graphics. Key indicators and graphical displays provide a basis for production decisions.

[0081] (1) Display of production scheduling results:

[0082] a. Display the calculation results of the scheduling and production plan from multiple dimensions.

[0083] b. Automatically generate emergency scheduling and production plan result reports.

[0084] c. Automatically generate corresponding emergency scheduling and production plan analysis reports.

[0085] (2) Production scheduling results push:

[0086] The scheduling and production calculation results can be automatically pushed to relevant personnel via in-station messages and instant messaging.

[0087] This invention also proposes an emergency production scheduling device for refineries, the principle of which is similar to the emergency production scheduling method for refineries, and will not be described in detail here.

[0088] Figure 3 This is a schematic diagram of a refinery emergency production scheduling device provided in an embodiment of the present invention, such as... Figure 3 As shown, the emergency production scheduling device for a refinery may include:

[0089] The acquisition module 301 is used to acquire real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; the refinery unit relationship tree is established using knowledge graph technology based on the type of units in the refinery and the production relationship between units; the core operating indicators are operating indicators whose impact on refinery production is higher than the preset impact level.

[0090] The execution module 302 is used to execute the automatic production scheduling calculation steps when the real-time data is determined to be abnormal according to the preset abnormality judgment algorithm and the automatic production scheduling calculation is determined to be enabled according to the pre-configured production scheduling parameters corresponding to the core operating indicators. The preset abnormality judgment algorithm is determined according to the data type of the real-time data. The production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is determined to be abnormal.

[0091] The automatic production scheduling calculation steps include:

[0092] Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree;

[0093] Based on the abnormal device, the devices associated with the abnormal device, and the pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data.

[0094] After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated.

[0095] The results of the automatic production scheduling calculations were used as the emergency production scheduling plan for the refinery.

[0096] In one embodiment, the core operating indicators are obtained by inputting the test set into a pre-trained core operating indicator recognition model; the core operating indicator recognition model is obtained by training a classification model using a training set; the test set consists of the data of all operating indicators of all units in a pre-established refinery unit relationship tree for the first historical time period and the refinery production failure information for the first historical time period; the training set consists of the data of all operating indicators of all units in a pre-established refinery unit relationship tree for the second historical time period and the refinery production failure information for the second historical time period.

[0097] In one embodiment, if the data type of the real-time data is numerical, the preset anomaly detection algorithm is a dynamic threshold algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is Boolean, the preset anomaly detection algorithm is a classification algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is an image, the preset anomaly detection algorithm is one or any combination of an edge detection algorithm, a regression-based deep learning algorithm, and a classification-based deep learning algorithm.

[0098] Execution module 302 is specifically used for:

[0099] The system compares real-time data with a dynamic threshold determined by a dynamic thresholding algorithm, and determines that the real-time data is abnormal based on the comparison results; it inputs real-time data into a recognition model pre-trained using an artificial intelligence algorithm to output recognition results, and determines that the real-time data is abnormal based on the recognition results; it classifies real-time data using a classification algorithm, and determines that the real-time data is abnormal based on the classification results; it analyzes the statistical or texture features of real-time data using an edge detection algorithm, and determines that the real-time data is abnormal based on the analysis results; it encodes normal images using a regression-based deep learning network to learn the features of normal images, and determines that the real-time data is abnormal based on the learned regression-based deep learning network; and / or, it learns the features of normal and abnormal images using a classification-based deep learning network, and determines that the real-time data is abnormal based on the learned classification-based deep learning network.

[0100] Based on the pre-configured production scheduling parameters corresponding to the core operating indicators, determine whether to enable automatic production scheduling calculation and execute the automatic production scheduling calculation steps.

[0101] In one embodiment, the execution module 302 is specifically used for:

[0102] After determining that the real-time data is abnormal according to the preset anomaly judgment algorithm, determine one or any combination of the abnormal real-time data collection method, collection equipment, data source and detection environment;

[0103] Determine whether one or any combination of the methods, devices, sources and detection environments for collecting abnormal real-time data is abnormal, and generate an anomaly determination result;

[0104] If the anomaly judgment result is abnormal, it means that the real-time data anomaly is caused by a detection problem and is not real. Adjust the alarm parameter to not alarm.

[0105] If the anomaly judgment result is non-anomaly, it means that the real-time data anomaly is caused by an undetected problem and is real. Adjust the alarm parameter to alarm.

[0106] Automatic production scheduling calculation is activated based on pre-configured production scheduling parameters corresponding to core operating indicators.

[0107] Based on the alarm parameters and production scheduling parameters, execute the automatic production scheduling calculation steps.

[0108] In one embodiment, the configuration table also includes the scheme pusher and the correspondence between the scheme pusher and the units in the refinery unit relationship tree.

[0109] Figure 4 This is a schematic diagram of a specific example of a refinery emergency production scheduling device provided in an embodiment of the present invention, such as... Figure 4 As shown, the refinery's emergency production scheduling device may also include:

[0110] The lookup module 401 is used to: find, based on the configuration table, the solution pusher that has a corresponding relationship with the abnormal device and the device associated with the abnormal device;

[0111] The push module 402 is used to push the emergency production scheduling plan of the refinery to the plan pusher through the station message push method and / or instant messaging platform.

[0112] In one embodiment, the refinery production emergency scheduling device may further include: a refinery unit relationship tree establishment module, used for:

[0113] Obtain the types of units in the refinery and the production relationships between them; the production relationships between units include the impact on other units after a core operating indicator of a unit becomes abnormal in real time.

[0114] Knowledge graph technology is used to establish a refinery unit relationship tree based on the types of units in the refinery and the production relationships between the units.

[0115] In one embodiment, the refinery production emergency scheduling device may further include: a display module, used for:

[0116] The emergency production scheduling plan for refineries is presented from multiple dimensions. These dimensions include the input and output of each scheduled unit, unit load, oil receipt and payment in the material tank area, physical tank inventory, crude oil processing schedule, feedstock arrival schedule, product production schedule, backflow schedule, material flow change schedule, unit feed control, material inventory control, unit scheme switching, production and sales balance, and one or any combination of combined units. The emergency production scheduling plan for refineries includes one or any combination of the following: unit material balance sheet, tank area material balance report, physical tank inventory trend table, unit processing scheduling schedule, crude oil processing schedule, feedstock arrival schedule, product production schedule, automotive gasoline data forecast table, automotive diesel data forecast table, hydrogen balance report, and unit material report.

[0117] Compared with existing refinery production emergency scheduling technology, this invention obtains real-time data of core operating indicators of all units in a pre-established refinery unit relationship tree. The refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between them. Core operating indicators are those with a higher-than-preset impact on refinery production. These core operating indicators are obtained by inputting a test set into a pre-trained core operating indicator recognition model. The core operating indicator recognition model is obtained by training a classification model using a training set. The test set consists of data from the first historical time period of all operating indicators of all units in the pre-established refinery unit relationship tree, along with refinery production fault information from that time period. The training set consists of data from the second historical time period of all operating indicators of all units in the pre-established refinery unit relationship tree, along with refinery production fault information from that time period. When a pre-set anomaly judgment algorithm determines that real-time data is abnormal, and the pre-configured scheduling parameters corresponding to the core operating indicators determine whether automatic scheduling calculation is enabled, the automatic scheduling calculation steps are executed. The pre-set anomaly judgment algorithm is determined based on the data type of the real-time data. The scheduling parameters are used to determine... When real-time data of core operating indicators becomes abnormal, the system determines whether to activate automatic production scheduling calculation. The automatic production scheduling calculation steps include: identifying the associated units of the abnormal unit corresponding to the abnormal real-time data based on the refinery unit relationship tree; determining the calculation model, calculation model parameters, calculation duration, and recalculation interval based on the abnormal unit, the associated units, and a pre-configured configuration table; the configuration table includes the correspondence between all units in the refinery unit relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval; the calculation model is a pre-established mathematical programming model; the objective of the mathematical programming model is... To minimize the impact of abnormal real-time data, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model for a calculation period. After the calculation period ends, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again for a calculation period. After this process, an automatic production scheduling calculation result is generated. The automatic production scheduling calculation result is determined as the emergency production scheduling plan for the refinery. This can quickly generate a reasonable emergency production scheduling plan for the entire plant, improve the efficiency, initiative, and timeliness of emergency production scheduling, and reduce the risk of production delays.

[0118] Compared with existing technologies, the emergency scheduling function of this invention monitors the core operating indicators of the device in real time. When a parameter exceeds a preset threshold, the system automatically triggers an emergency scheduling plan, retrieves the preset configuration parameters and calculation models of the relevant devices, and calculates a multi-dimensional emergency scheduling plan based on the model algorithm. The final optimized scheduling plan will be pushed to relevant responsible persons through in-system messages and instant messaging, ensuring that key users receive it as soon as possible, providing data support and plan reference for subsequent scheduling decisions.

[0119] This invention also provides a computer device. Figure 5 This is a schematic diagram of a computer device in an embodiment of the present invention. The computer device 500 includes a memory 510, a processor 520, and a computer program 530 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 530, it implements the above-mentioned emergency production scheduling method for refineries.

[0120] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described emergency production scheduling method for refineries.

[0121] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned emergency production scheduling method for refineries.

[0122] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention 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.

[0123] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. 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 illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0124] 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 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0125] 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 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0126] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for emergency production scheduling in a refinery, characterized in that, include: Obtain real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; The refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between units; the core operating indicators are operating indicators whose impact on refinery production is higher than a preset impact level. When the real-time data is determined to be abnormal according to the preset anomaly judgment algorithm, and the automatic production scheduling calculation is activated according to the pre-configured production scheduling parameters corresponding to the core operating indicators, the automatic production scheduling calculation steps are executed; the preset anomaly judgment algorithm is determined according to the data type of the real-time data; The production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is found to be abnormal. The automatic production scheduling calculation includes: Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree; Based on the abnormal device, the devices associated with the abnormal device, and a pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data. After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated. The results of the automatic production scheduling calculations are used as the emergency production scheduling plan for the refinery.

2. The method as described in claim 1, characterized in that, The core operating indicators are obtained by inputting the test set into a pre-trained core operating indicator recognition model; the core operating indicator recognition model is obtained by training a classification model using a training set; the test set consists of the data of all operating indicators of all units in the pre-established refinery unit relationship tree for the first historical time period and the refinery production failure information for the first historical time period; the training set consists of the data of all operating indicators of all units in the pre-established refinery unit relationship tree for the second historical time period and the refinery production failure information for the second historical time period.

3. The method as described in claim 1, characterized in that, If the data type of the real-time data is numerical, the preset anomaly detection algorithm is a dynamic threshold algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is Boolean, the preset anomaly detection algorithm is a classification algorithm and / or an artificial intelligence algorithm; if the data type of the real-time data is an image, the preset anomaly detection algorithm is one or any combination of an edge detection algorithm, a regression-based deep learning algorithm, and a classification-based deep learning algorithm. When the real-time data is determined to be abnormal according to a preset anomaly detection algorithm, and automatic production scheduling is activated based on pre-configured production scheduling parameters corresponding to core operating indicators, the automatic production scheduling calculation steps are executed, including: The real-time data is compared with a dynamic threshold determined by a dynamic threshold algorithm, and the real-time data is judged to be abnormal based on the comparison result; the real-time data is input into a recognition model pre-trained using an artificial intelligence algorithm to output a recognition result, and the real-time data is judged to be abnormal based on the recognition result; the real-time data is classified using a classification algorithm, and the real-time data is judged to be abnormal based on the classification result; the statistical or texture features of the real-time data are analyzed using an edge detection algorithm, and the real-time data is judged to be abnormal based on the analysis result; normal images are encoded using a regression-based deep learning network to learn the features of normal images, and the learned regression-based deep learning network is judged to be abnormal based on the real-time data; and / or, the features of normal and abnormal images are learned using a classification-based deep learning network, and the learned classification-based deep learning network is judged to be abnormal based on the real-time data. Based on the pre-configured production scheduling parameters corresponding to the core operating indicators, determine whether to enable automatic production scheduling calculation and execute the automatic production scheduling calculation steps.

4. The method as described in claim 1, characterized in that, When the real-time data is determined to be abnormal according to a preset anomaly detection algorithm, and automatic production scheduling is activated based on pre-configured production scheduling parameters corresponding to core operating indicators, the automatic production scheduling calculation steps are executed, including: After determining that the real-time data is abnormal according to the preset anomaly judgment algorithm, one or any combination of the abnormal real-time data collection method, collection equipment, data source and detection environment is determined. Determine whether one or any combination of the methods, devices, sources and detection environments for collecting abnormal real-time data is abnormal, and generate an anomaly determination result; If the anomaly judgment result is abnormal, it means that the real-time data anomaly is caused by a detection problem and is not real. Adjust the alarm parameter to not alarm. If the anomaly judgment result is non-anomaly, it means that the real-time data anomaly is caused by an undetected problem and is real. Adjust the alarm parameter to alarm. Automatic production scheduling calculation is activated based on pre-configured production scheduling parameters corresponding to core operating indicators. Based on the alarm parameters and production scheduling parameters, execute the automatic production scheduling calculation steps.

5. The method as described in claim 1, characterized in that, The configuration table also includes the scheme pusher and the correspondence between the scheme pusher and the units in the refinery unit relationship tree; The method further includes: According to the configuration table, find the solution pusher that corresponds to the abnormal device and the device associated with the abnormal device; The emergency production scheduling plan for the refinery will be pushed to the sender via in-station messaging and / or instant messaging platform.

6. The method as described in claim 1, characterized in that, Before acquiring real-time data on the core operating metrics of all units in the pre-established refinery unit relationship tree, the process also includes: Obtain the types of units in the refinery and the production relationships between them; the production relationships between units include the impact on other units after a core operating indicator of a unit becomes abnormal in real time. Knowledge graph technology is used to establish a refinery unit relationship tree based on the types of units in the refinery and the production relationships between the units.

7. The method as described in claim 1, characterized in that, Also includes: The emergency production scheduling plan for refineries is presented from multiple dimensions. These dimensions include the input and output of each scheduled unit, unit load, oil receipt and payment in the material tank area, physical tank inventory, crude oil processing schedule, raw material delivery schedule, product production schedule, backflow schedule, material flow change schedule, unit feed control, material inventory control, unit scheme switching, production and sales balance, and one or any combination of combined units. The emergency production scheduling plan for refineries includes one or any combination of the following: unit material balance sheet, tank area material balance report, physical tank inventory trend table, unit processing scheduling schedule, crude oil processing schedule, raw material delivery schedule, product production schedule, automotive gasoline data forecast table, automotive diesel data forecast table, hydrogen balance report, and unit material report.

8. A refinery production emergency scheduling device, characterized in that, include: The acquisition module is used to acquire real-time data of the core operating indicators of all units in the pre-established refinery unit relationship tree; The refinery unit relationship tree is established using knowledge graph technology based on the types of units in the refinery and the production relationships between units; the core operating indicators are operating indicators whose impact on refinery production is higher than a preset impact level. The execution module is used to execute the automatic production scheduling calculation steps when the real-time data is determined to be abnormal according to a preset anomaly judgment algorithm and the automatic production scheduling calculation is activated according to the pre-configured production scheduling parameters corresponding to the core operating indicators; the preset anomaly judgment algorithm is determined according to the data type of the real-time data. The production scheduling parameters are used to determine whether to enable automatic production scheduling calculation when the real-time data of the core operating indicators is found to be abnormal. The automatic production scheduling calculation includes: Identify the devices associated with the abnormal devices corresponding to the abnormal real-time data based on the refinery unit relationship tree; Based on the abnormal device, the devices associated with the abnormal device, and a pre-configured configuration table, the calculation model, calculation model parameters, calculation duration, and recalculation interval are determined. The configuration table includes the correspondence between all devices in the refinery device relationship tree and the calculation model, calculation model parameters, calculation duration, and recalculation interval. The calculation model is a pre-established mathematical programming model. The objective of the mathematical programming model is to minimize the impact of abnormal real-time data. After the calculation of the calculation time is completed by inputting the abnormal real-time data and the real-time data corresponding to the calculation model parameters into the calculation model, the interval is calculated again after the end time. Then, the abnormal real-time data and the real-time data corresponding to the calculation model parameters are input into the calculation model again to calculate the calculation time, and the automatic production scheduling calculation result is generated. The results of the automatic production scheduling calculations are used as the emergency production scheduling plan for the refinery.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images