Petrochemical storage tank area inventory collaborative management system based on end-cloud collaboration

Abstract

The invention provides a petrochemical engineering storage tank area inventory collaborative management system based on end-cloud collaboration, and relates to the technical field of petrochemical engineering inventory management, and the system comprises the steps that a global sensing module collects storage tank operation state data and environment data; the edge processing module carries out localization processing and feature extraction on the operation state data of the storage tank; the cloud brain analysis module carries out risk prediction; the environment analysis module performs risk assessment on the environment data; the cooperation adaptation module matches an optimal cooperation rule; and the decision module generates a collaborative management strategy and issues an instruction. According to the invention, through localized preprocessing and feature extraction, the cloud transmission load is effectively reduced and the response speed is improved; through an intelligent rule matching and dynamic updating mechanism, it is ensured that the system can automatically select an optimal coping strategy according to the real-time risk situation, a collaborative management strategy covering inventory adjustment, equipment control and early warning notification is finally generated, and the safety, stability and economical efficiency of storage tank area operation are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of petrochemical inventory management, and particularly relates to a petrochemical storage tank area inventory collaborative management system based on end-cloud collaboration. BACKGROUND

[0002] Petrochemical storage is a key link connecting upstream production and downstream sales and transportation, and its core task is to safely, stably receive, store and distribute various petrochemical raw materials, intermediates and finished products. Due to the dangerous characteristics of flammability, explosiveness, toxicity and corrosiveness of the materials, and the huge storage scale, the storage system has high professionalism and complexity in the technical aspect.

[0003] The stored materials are classified and stored according to their physical state and characteristics. Liquid materials are the main body of storage, mainly including light oil products such as crude oil, gasoline, naphtha, and heavy oil products such as diesel, wax oil, fuel oil, and liquid chemicals such as benzene, p-xylene and acrylonitrile. Gaseous materials include liquefied petroleum gas, liquefied natural gas and liquefied hydrocarbons such as propylene and ethylene, which are usually stored in liquid form under low temperature or high pressure.

[0004] Storage facilities take the storage tank as the core. According to the characteristics of the storage medium, the storage tank is significantly different in type, material and design. Common types include: floating roof tanks for storing crude oil and heavy oil; fixed roof tanks for storing volatile light oil and chemicals; and spherical tanks or double-wall metal vacuum insulated vertical tanks for storing liquefied gas. The storage tank is equipped with perfect safety accessories, such as mechanical breather valve, flame arrestor, emergency pressure relief device, foam fire extinguishing system, etc.

[0005] The tank farm is usually equipped with a complete pipe network and pumping system for material receiving and outputting. In addition, auxiliary systems are essential, including: oil gas recovery device for handling volatile organic compounds generated during loading and unloading and small breathing, reducing loss and emission; fire fighting system composed of fire water pipe network, foam station, fire alarm system; and environmental protection facilities such as cofferdam, accident pool, rainwater discharge control system for preventing leakage of materials from polluting the external environment.

[0006] The inventory management of petrochemical storage tank area refers to a systematic project of using a series of technical means to monitor, accurately calculate and dynamically track the quantity, quality and state of crude oil, semi-finished products and finished oil and other media stored in the storage tank in real time. Its core goal is to accurately grasp the static inventory and dynamic turnover data of the tank area.

[0007] The inventory management of existing petroleum and chemical storage tank farms usually relies on a combination of relatively independent systems, which has obvious limitations. First, the data of the subsystems such as inventory monitoring, environmental monitoring, equipment control and safety warning are poor in interconnectivity, and there is a lack of a unified analysis and decision-making platform, resulting in insufficient management coordination. Risk judgment is mostly dependent on a simple threshold alarm mechanism, that is, an alarm is triggered when a certain parameter exceeds a pre-set fixed threshold. This way is passive and lagging, and it is difficult to make early warning and predictive judgment on potential risks caused by the coupling of multiple parameters and gradual evolution under complex working conditions. Furthermore, the decision-making process is highly dependent on human experience, and when faced with a large amount of data and unexpected situations, management personnel have difficulty in quickly and accurately developing the optimal response strategy, and are prone to delayed response or decision bias.

[0008] In order to solve the defects in the prior art, the technical scheme provides a petroleum and chemical storage tank farm inventory collaborative management system based on end-cloud collaboration. SUMMARY

[0009] The present application provides a petroleum and chemical storage tank farm inventory collaborative management system based on end-cloud collaboration to solve the defects in the prior art.

[0010] In one aspect, the present application provides a petroleum and chemical storage tank farm inventory collaborative management system based on end-cloud collaboration, comprising: A global perception module for collecting storage tank operating state data and environmental data; An edge processing module for local processing and feature extraction of the storage tank operating state data, and outputting operating feature data; A cloud brain analysis module for risk prediction based on digital twinning of the operating feature data, and outputting an initial risk prediction result; An environmental analysis module for risk assessment of the environmental data, and outputting an environmental assessment result; A collaborative adaptation module for matching the most suitable collaborative rule based on the risk prediction result and the environmental assessment result, in combination with a pre-set collaborative rule; A decision-making module for generating a collaborative management strategy based on the most suitable collaborative rule and issuing instructions.

[0011] According to the petroleum and chemical storage tank farm inventory collaborative management system based on end-cloud collaboration provided by the present application, the edge processing module comprises: A data preprocessing unit for preprocessing the storage tank operating state data and outputting standardized data; A feature extraction unit for extracting trend change feature data from the standardized data; An edge storage unit for local storage and integration of the standardized data and the trend change feature data, and obtaining operating feature data.

[0012] The application provides a petroleum chemical storage tank area inventory collaborative management system based on end-cloud cooperation, and the cloud brain analysis module comprises: A digital twin modeling unit is configured to construct a storage tank area geometric model, a physical simulation model and an operation behavior model, and form a full-element digital twin; A twin data synchronization unit is configured to synchronize trend change characteristic data of standardized data to the full-element digital twin; A risk prediction model unit is configured to analyze the synchronized data in the full-element digital twin and output an initial risk prediction result; A prediction result correction unit is configured to correct the initial risk prediction result based on historical risk event data.

[0013] The application provides a petroleum chemical storage tank area inventory collaborative management system based on end-cloud cooperation, and the step of outputting the initial risk prediction result by the risk prediction model unit comprises: The trend change characteristic data is input into a mechanism model, and based on a petroleum chemical storage tank thermodynamic equation, a fluid mechanics equation and a device loss mechanism, a theoretical risk threshold value and a risk development rate are calculated to form mechanism model risk benchmark data; Engineering time sequence characteristic extraction and analysis are performed on the standardized time sequence data, time sequence prediction risk values and confidence intervals are output, and time sequence risk benchmark data are obtained; A weighted fusion algorithm is used to fuse the mechanism model risk benchmark data and the time sequence risk benchmark data, and a fusion prediction result is output; The fusion prediction result is processed, and an initial risk prediction result containing a risk level, a risk occurrence probability and a predicted occurrence time window is output.

[0014] The application provides a petroleum chemical storage tank area inventory collaborative management system based on end-cloud cooperation, and in the risk prediction model unit, the step of obtaining the time sequence risk benchmark data comprises: The standardized time sequence data is arranged according to a fixed time window to form an input sample set; Trend characteristic extraction is performed on the input sample set, and the change slope, fluctuation amplitude and abnormal point distribution of parameters are mainly captured, the extracted characteristics are associated and matched with a storage tank historical risk event database, key characteristics having a significant influence on risk prediction are screened out to form a characteristic subset; Attention weights are calculated based on the risk influence degrees of parameters in the characteristic subset; Based on the fusion prediction result, a final time sequence prediction risk is calculated.

[0015] The application provides a petroleum chemical storage tank area inventory collaborative management system based on end-cloud cooperation, and the environment analysis module comprises: Environmental parameter threshold configuration unit, used to preset risk thresholds for different environmental parameters; The coupled risk assessment unit is used to analyze the coupling relationship between environmental data and tank operation risks based on risk thresholds, and to calculate the environmental coupled risk value. The environmental assessment result classification unit is used to classify the environmental coupling risk values ​​into levels and label the environmental impact factors corresponding to each level.

[0016] According to the present invention, a collaborative management system for petrochemical tank farm inventory based on edge-cloud collaboration includes the following steps for calculating the environmental coupling risk value using a coupled risk assessment unit: The preset threshold is called through the environmental parameter threshold configuration unit, and the actual value of each environmental parameter is compared with the corresponding risk threshold to calculate the degree of exceedance of a single parameter. Based on a historical risk event database, the coupling coefficient between various environmental parameters and the operational risks of storage tanks is determined; The environmental coupling risk value is calculated using the coupling risk synthesis formula; Output the environmental coupling risk value and the risk contribution ratio of each environmental parameter.

[0017] According to the present invention, a collaborative management system for petrochemical tank farm inventory based on edge-cloud collaboration is provided, the collaborative adaptation module including: A collaborative rule base is used to store preset collaborative rules; The rule matching algorithm unit uses a weighted priority matching algorithm, which sets weight coefficients based on risk thresholds to select the most suitable collaborative rules. The rule dynamic update unit is used to iteratively update the preset collaboration rules based on historical collaboration effect evaluation data and newly added risk event data.

[0018] According to the present invention, a collaborative management system for petrochemical tank farm inventory based on edge-cloud collaboration includes a decision module comprising: The strategy generation unit is used to generate collaborative management strategies based on the optimal collaborative rules, including inventory adjustment strategies, equipment control strategies, and early warning notification strategies. Inventory adjustment strategies include feed rate adjustment, discharge rate adjustment, and inventory transfer path planning. Equipment control strategies include valve opening adjustment, pump start-stop control, and safety protection equipment start-up control. Early warning notification strategies include sending early warning information to designated personnel according to the assessment level. The instruction issuing unit is used to convert collaborative management strategies into standardized control instructions and issue them to the corresponding execution terminals through an encrypted communication channel. The strategy execution feedback unit is used to receive strategy execution result feedback data from the execution terminal, monitor the execution effect of the collaborative management strategy in real time, and output feedback data. A strategy optimization unit is configured to perform real-time optimization adjustment on the collaborative management strategy based on the feedback data and the initial risk prediction result.

[0019] According to the petroleum chemical storage tank area inventory collaborative management system based on end-cloud collaboration provided by the application, the steps of the strategy optimization unit for real-time optimization adjustment on the collaborative management strategy include: The feedback data transmitted by the strategy execution feedback unit is collected, including real-time inventory data, equipment operation state data, environment data update values and risk level change data; A strategy execution effect evaluation index system is constructed, including risk control efficiency, execution cost and inventory stability; The deviation rate of each evaluation index from the preset target value is calculated, and if the absolute value of the deviation rate exceeds the preset deviation range, the strategy optimization process is triggered; Based on the deviation reason analysis result, the collaborative management strategy parameters are adjusted, including: if the risk control efficiency is insufficient, the rate parameter of the inventory adjustment strategy or the response threshold of the equipment control strategy is adjusted; if the execution cost is too high, the inventory transfer path planning or the equipment start-stop timing is optimized; if the inventory stability is not up to standard, the inventory safety threshold interval and the early warning trigger condition are adjusted; The optimized strategy parameters are input into the digital twin for simulation verification, and if the simulation result meets the evaluation index requirement, the formal optimization strategy is generated; if not, the collaborative management strategy parameters are adjusted again; The formal optimization strategy is sent to the instruction issuing unit, and the collaborative management strategy parameters and execution effect data before and after optimization are recorded.

[0020] The petroleum chemical storage tank area inventory collaborative management system based on end-cloud collaboration provided by the application realizes omnibearing intelligent management and control of the petroleum chemical storage tank area inventory through the end-cloud collaborative architecture. The global perception module collects the storage tank operation state and environment data in real time, and through local preprocessing and feature extraction, the cloud transmission load is effectively reduced and the response speed is improved. The cloud brain analysis module constructs a high-precision simulation model based on digital twin technology, combines mechanism analysis and time series prediction fusion algorithm, realizes accurate prediction and dynamic correction of equipment operation risk. At the same time, the coupling risk of external environmental factors and storage tank operation state is evaluated, which provides an important basis for comprehensive decision-making. Through intelligent rule matching and dynamic updating mechanism, the system can automatically select the optimal response strategy according to the real-time risk situation. Finally, the collaborative management strategy covering inventory adjustment, equipment control and early warning notification is generated, and the strategy execution effect is continuously optimized through the closed-loop feedback mechanism, thereby significantly improving the safety, stability and economy of the storage tank area operation. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings needed to be used in the following embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort.

[0022] Figure 1 is a structural schematic diagram of an oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration provided by an embodiment of the present application. Figure 2 is a flowchart of obtaining time sequence risk benchmark data by a risk prediction model unit provided by an embodiment of the present application. Figure 3 is a flowchart of real-time optimization and adjustment of collaborative management strategies by a strategy optimization unit provided by an embodiment of the present application. Figure 4 is a working time sequence diagram of an oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration provided by an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in combination with the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the protection scope of the present application.

[0024] Embodiment one: The present application will be described below in combination with Figures 1-4 a kind of oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration.

[0025] As shown in Figures 1-4 , a kind of oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration provided by an embodiment of the present application, comprising: Global perception module is used to collect storage tank operation state data and environmental data. Through the deployment of various sensors in the storage tank area, real-time data collection is carried out, including tank medium temperature, pressure, liquid level, medium density, VOCs (volatile organic compounds) concentration, environmental wind speed, etc. The data format includes analog signal and digital signal, and the sampling frequency is dynamically adjusted according to the importance of the parameters, for example, the sampling frequency of liquid level and pressure is not less than 1Hz, and the sampling frequency of environmental parameters is 0.1-1Hz.

[0026] Edge processing module is used for local processing and feature extraction of storage tank operation state data, and output operation feature data.

[0027] The edge processing module includes a data preprocessing unit, a feature extraction unit, and an edge storage unit. The data preprocessing unit is used to preprocess the storage tank operating state data and output standardized data. The preprocessing steps include data cleaning, outlier removal, missing value interpolation, and normalization processing. Specifically, the data cleaning uses a sliding window method based on the storage tank process to identify and remove outliers, for example, a sudden change in liquid level exceeding the safe height change rate is considered abnormal. The missing values are filled using linear interpolation or regression method based on historical data. The normalization processing uses the minimum-maximum scaling to convert the data to the [0, 1] interval, the formula is: where x is the original value, is the standardized value, x min and x max are the minimum and maximum values of the historical data.

[0028] The feature extraction unit is used to extract the trend change feature data in the standardized data. The feature extraction method includes time domain features and frequency domain features, focusing on extracting the liquid level change rate, pressure change gradient, temperature change trend, and key features of calculating the inventory volume change. For liquid level data, the sliding window technique is used to calculate the linear regression slope of the liquid level within the window as the change trend feature, represented as:

[0029] where k H represents the change trend slope or change rate of the liquid level data within the specified time window, representing the overall trend of the liquid level over time. When k H > 0, it indicates that the liquid level is rising in this time period. When k H < 0, it indicates that the liquid level is declining. When k H ≈ 0, it indicates that the liquid level remains stable. H i is the liquid level value, t i is the time point, and are the mean values. At the same time, the range ΔH of the liquid level within the window is calculated as the fluctuation amplitude feature. Combined with the tank volume table, the inventory volume change ΔV ≈ A(H)*ΔH can be preliminarily estimated, where A(H) is the cross-sectional area of the tank body corresponding to the current liquid level H, obtained by interpolation of the tank volume table.

[0030] The edge storage unit is used to locally store and integrate the standardized data and trend change feature data to obtain the operating feature data. The storage format uses a time series database, and the data label includes the storage tank number and the material type. The storage period is set according to the importance of the data, and the key inventory data is stored for more than 30 days, and the ordinary data is stored for 7 days. The rotating door compression algorithm is used to reduce the storage space, and the compression error is controlled within 1% of the measurement accuracy.

[0031] The cloud brain analysis module is configured to perform risk prediction on the operation characteristic data based on the digital twin, and output an initial risk prediction result.

[0032] The cloud brain analysis module includes a digital twin modeling unit, a twin data synchronization unit, and a risk prediction model unit. The digital twin modeling unit is configured to construct a geometric model, a physical simulation model, and an operation behavior model of the tank farm, to form a full-factor digital twin. The geometric model is configured to construct an accurate three-dimensional model based on tank CAD drawings and point cloud data, and includes a tank body, pipelines, valves, and the like. The physical simulation model includes a static inventory calculation model, a thermodynamic model, a pressure model, and an operation behavior model. The static inventory calculation model is expressed as:

[0033]

[0034] wherein V is the volume of the material in the tank, H is the liquid level of the material in the tank, ρ is the density of the material in the tank at the current temperature T, α is the volumetric expansion coefficient, T is the temperature of the material in the tank, and f() is determined by a tank capacity table and a temperature-volume correction formula, i.e., a tank capacity table indicating the corresponding relationship between the liquid level H and the volume V of the tank at a standard temperature, which is generated by a professional measurement institution. ρ T represents the density of the material at the actual temperature T, ρ 20 represents the density of the material at the standard temperature of 20°C.

[0035] The thermodynamic model mainly includes a tank internal temperature field model considering environmental heat exchange and material import and export.

[0036] The pressure model is configured to calculate the pressure change of the gas phase space based on a gas state equation, and the formula is expressed as:

[0037] wherein P represents the actual pressure of the gas phase space of the tank, V a represents the volume of the gas phase space of the tank, n is the number of gas molecules in the gas phase space, R is a physical constant applicable to all ideal gases, and is a fixed value, R≈8.314kJ / (kmol·K) when the pressure P is in kPa and the volume V is in m 3 T s is the absolute temperature of the gas phase space, and Z is a correction factor for measuring the deviation between the actual gas and the ideal gas behavior.

[0038] The operation behavior model simulates dynamic processes such as material import and export operations and breathing valve actions.

[0039] The digital twin data synchronization unit is used to synchronize the trend change characteristics of standardized data to the full-factor digital twin. The risk prediction model unit is used to analyze the synchronized data in the full-factor digital twin and output initial risk prediction results. The steps include: By inputting trend change characteristic data into the mechanistic model, and based on the thermodynamic equations, fluid dynamics equations, and equipment loss mechanisms of petrochemical storage tanks, the theoretical risk threshold and risk development rate are calculated, forming the risk benchmark data for the mechanistic model. For the risk of tank overflow, the formula for calculating the maximum safe liquid level in the mechanistic model is expressed as:

[0040] Among them, H max H represents the maximum safe liquid level. design To design the highest liquid level, k s The safety margin factor is typically taken as 1.1-1.2. The risk development rate can be calculated based on the current feed flow rate and tank capacity, and is expressed as:

[0041] Among them, t risk H represents the time frame for predicting risk. current F represents the current real-time measured liquid level. in This indicates the current feed flow rate of the feed pipeline.

[0042] To address the risk of overpressure, the time required to reach the set pressure is calculated based on the set pressure of the breathing valve and the current rate of pressure rise.

[0043] Standardized time-series data undergoes engineered time-series feature extraction and analysis to output predicted time-series risk values ​​and confidence intervals, thus obtaining benchmark data for time-series risk. The time-series analysis employs seasonal decomposition and LSTM neural networks to predict key parameters such as liquid level and pressure over a future period. The risk value is calculated as the degree to which the predicted value deviates from the safety threshold; for liquid level, the formula is as follows:

[0044] Among them, R t(H) H represents the time-series predicted risk value of the liquid level height. predicted H represents the liquid level value at a future moment predicted by a time series model. safe For early warning of liquid level, max(0,...) is the function that takes the maximum value. The confidence interval is determined by the historical distribution of the model's prediction error.

[0045] A weighted fusion algorithm is used to fuse the risk baseline data of the mechanistic model and the time series risk baseline data, and the fused prediction result is output, expressed by the formula:

[0046] Among them, R fused To fuse the prediction results, R m As the risk benchmark data, R t As the benchmark data for time series risk, w m and w t As the weights of the risk benchmark data and the time-series risk benchmark data, w m +w t =1, which can be dynamically adjusted based on the model's confidence level under the current operating conditions. For example, under stable operating conditions, the mechanistic model has a higher weight; when operating conditions change drastically, the weight of the time series model is appropriately increased. The weight adjustment can be based on the reciprocal of the model's recent prediction error.

[0047] The fusion prediction results are processed to output initial risk prediction results, including risk level, probability of occurrence, and predicted occurrence time window. The risk level classification threshold can be dynamically adjusted according to the hazard of the materials in the storage tank.

[0048] In the risk prediction model unit, the steps for obtaining time-series risk baseline data include: Standardized time-series data are organized into fixed time windows to form an input sample set. The sample set contains equally spaced sampling sequences of parameters such as liquid level, pressure, and temperature.

[0049] Trend features are extracted from the input sample set, focusing on capturing the slope of parameter changes, fluctuation amplitude, and distribution of outliers. The extracted features are then matched with the historical risk event database of storage tanks to screen out key features that have a significant impact on risk prediction, forming a feature subset. Features such as the slope of continuous and rapid rise in liquid level and increased pressure fluctuations are highly correlated with the risks of tank overflow and overpressure.

[0050] Based on the risk impact of each parameter in the feature subset, the attention weight is calculated, expressed by the following formula:

[0051] Where, α i is the attention weight, representing the importance or contribution of the i-th feature to the current risk prediction. i Let represent the original importance score of the i-th feature, and j be the total number of features. This is the sum of the original importance scores for all features.

[0052] Based on the fusion prediction results, the final time series prediction risk is calculated, represented as a weighted combination of key feature values, and mapped to risk probability.

[0053] The prediction result correction unit is used to correct the initial risk prediction results based on historical risk event data. It uses the initial prediction probability as the prior probability and calculates the posterior probability by combining it with the frequency of similar recent risk events, making the prediction results closer to actual operational experience.

[0054] The environmental analysis module is used to conduct risk assessments on environmental data and output environmental assessment results.

[0055] The environmental analysis module includes an environmental parameter threshold configuration unit, a coupled risk assessment unit, and an environmental assessment result grading unit. The environmental parameter threshold configuration unit is used to preset risk thresholds for different environmental parameters. The coupled risk assessment unit is used to analyze the coupling relationship between environmental data and tank operation risks based on the risk thresholds, and calculate the environmental coupled risk value. The steps include: The environmental parameter threshold configuration unit calls preset thresholds, compares the actual value of each environmental parameter with the corresponding risk threshold, and calculates the degree β of each parameter exceeding the threshold. i When the actual value > the threshold, β i =(actual value - threshold) / threshold. When the actual value ≤ the threshold, β i =0, where i is the environmental parameter index. For parameters with upper and lower limits, β exceeding the upper limit and below the lower limit need to be calculated separately. i .

[0056] Based on a historical risk event database, the coupling coefficients between various environmental parameters and the operational risks of storage tanks were determined. The coupling coefficients ranged from 0 to 1, with the coupling coefficients for toxic and harmful gas concentrations and temperature not less than 0.7, and the coupling coefficients for humidity and wind speed ranging from 0.3 to 0.5.

[0057] The environmental coupling risk value R is calculated using the coupling risk synthesis formula, which is expressed as follows:

[0058] Where, γ i ω is the coupling coefficient. i Let Σω represent the importance weights of each environmental parameter. i =1, and the importance weight is determined by the analytic hierarchy process.

[0059] The output environmental coupling risk value and the risk contribution ratio of each environmental parameter are expressed as follows:

[0060] S represents the risk contribution ratio, used to identify the main sources of environmental risk.

[0061] The environmental assessment result classification unit is used to classify the environmental coupling risk value into levels and label the environmental impact factors corresponding to each level.

[0062] The collaborative adaptation module is used to match the most suitable collaborative rules based on the risk prediction results and environmental assessment results, combined with preset collaborative rules.

[0063] The collaborative adaptation module includes a collaborative rule base, a rule matching algorithm unit, and a rule dynamic update unit. The collaborative rule base stores preset collaborative rules. The rule matching algorithm unit uses a weighted priority matching algorithm, setting weight coefficients based on risk thresholds to select the most suitable collaborative rules. The rule dynamic update unit iteratively updates the preset collaborative rules based on historical collaborative effect evaluation data and new risk event data.

[0064] The decision-making module is used to generate collaborative management strategies based on the optimal collaborative rules and to issue instructions. The decision-making module includes a strategy generation unit, an instruction issuance unit, a strategy execution feedback unit, and a strategy optimization unit.

[0065] The strategy generation unit generates collaborative management strategies based on optimal collaborative rules, including inventory adjustment strategies, equipment control strategies, and early warning notification strategies. Inventory adjustment strategies include feed rate adjustment, discharge rate adjustment, and inventory transfer path planning. Equipment control strategies include valve opening adjustment, pump start / stop control, and safety protection equipment start control. Early warning notification strategies include sending early warning information to designated personnel according to assessment levels, with notification methods including SMS, app push notifications, and voice calls.

[0066] The instruction issuance unit converts collaborative management strategies into standardized control instructions and sends them to the corresponding execution terminals via an encrypted communication channel. The strategy execution feedback unit receives policy execution result feedback data from the execution terminals, monitors the execution effect of the collaborative management strategy in real time, and outputs feedback data. Feedback data includes instruction execution status, deviations between actual execution parameter values ​​and target values, device response time, and trends in key parameter changes after execution. The strategy optimization unit optimizes and adjusts the collaborative management strategy in real time based on the feedback data and initial risk prediction results.

[0067] The steps by which the strategy optimization unit performs real-time optimization and adjustment of the collaborative management strategy include: The feedback data transmitted by the strategy execution feedback unit includes real-time inventory data, equipment operating status data, environmental data update values, and risk level change data.

[0068] Construct a strategy execution effectiveness evaluation index system, including risk control efficiency, execution cost, and inventory stability. The risk control efficiency is calculated as follows:

[0069] Where, η risk R represents risk control efficiency. before R represents the risk value before risk control.after t represents the risk value after risk control. elapsed For execution time, C cost This is a cost factor.

[0070] Calculate the deviation rate between each evaluation indicator and the preset target value. If the absolute value of the deviation rate exceeds the preset deviation range, the strategy optimization process is triggered.

[0071] Based on the analysis of deviation causes, the parameters of the collaborative management strategy are adjusted accordingly, including: if risk control efficiency is insufficient, adjusting the rate parameters of the inventory adjustment strategy or the response threshold of the equipment control strategy; if execution costs are too high, optimizing inventory transfer path planning or equipment start-up and shutdown sequence; if inventory stability is not up to standard, adjusting the inventory safety threshold range and early warning triggering conditions.

[0072] The optimized strategy parameters are input into the digital twin for simulation verification. If the simulation results meet the evaluation criteria, a formal optimized strategy is generated. If not, the collaborative management strategy parameters are readjusted.

[0073] The formal optimization strategy is sent to the instruction issuing unit, and the collaborative management strategy parameters and execution effect data before and after optimization are recorded.

[0074] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration, characterized in that, The method comprises the following steps: A global perception module is used to collect tank operating state data and environmental data; An edge processing module is used to perform localized processing and feature extraction on the tank operating state data, and output operating feature data; A cloud brain analysis module is used to perform risk prediction on the operating feature data based on digital twinning, and output an initial risk prediction result; An environmental analysis module is used to perform risk assessment on the environmental data, and output an environmental assessment result; A collaborative adaptation module is used to match the most suitable collaborative rule based on the risk prediction result and the environmental assessment result, combined with a preset collaborative rule; A decision module is used to generate a collaborative management strategy based on the most suitable collaborative rule and issue instructions.

2. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 1, characterized in that, The edge processing module comprises: A data preprocessing unit is used to preprocess the tank operating state data and output standardized data; A feature extraction unit is used to extract trend change feature data from the standardized data; An edge storage unit is used to locally store and integrate the standardized data and the trend change feature data, and obtain the operating feature data.

3. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 2, characterized in that, The cloud brain analysis module comprises: A digital twinning modeling unit is used to construct a tank farm geometric model, a physical simulation model and an operating behavior model, and form a full-factor digital twinning body; A twinning data synchronization unit is used to synchronize the trend change feature data of the standardized data to the full-factor digital twinning body; A risk prediction model unit is used to analyze the synchronized data in the full-factor digital twinning body and output the initial risk prediction result; A prediction result correction unit is used to correct the initial risk prediction result based on historical risk event data.

4. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 3, characterized in that, The step of outputting the initial risk prediction result by the risk prediction model unit comprises: Trend change feature data is input into a mechanism model, and based on the equations of petroleum and chemical tank thermodynamics, fluid mechanics and equipment loss mechanism, a theoretical risk threshold and a risk development rate are calculated to form mechanism model risk benchmark data; Engineering time series feature extraction and analysis are performed on standardized time series data to output time series prediction risk values and confidence intervals, and time series risk benchmark data is obtained; A weighted fusion algorithm is used to fuse the mechanism model risk benchmark data and the time series risk benchmark data to output a fusion prediction result; The fusion prediction result is processed to output an initial risk prediction result containing a risk level, a risk occurrence probability and a predicted occurrence time window.

5. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 4, characterized in that, In the risk prediction model unit, the step of obtaining time series risk benchmark data comprises: The standardized time series data is arranged according to a fixed time window to form an input sample set; Trend feature extraction is performed on the input sample set, focusing on capturing the change slope, fluctuation amplitude and abnormal point distribution of parameters, and the extracted features are associated and matched with a tank historical risk event database to filter out key features that have a significant impact on risk prediction, forming a feature subset; Attention weights are calculated based on the risk influence degree of each parameter in the feature subset; Based on the fusion prediction result, a final time series prediction risk is calculated.

6. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 1, characterized in that, The environmental analysis module comprises: An environmental parameter threshold configuration unit is configured to preset risk thresholds for different environmental parameters. A coupling risk assessment unit is configured to analyze the coupling relationship between the environmental data and the operation risk of the storage tank based on the risk thresholds, and calculate an environmental coupling risk value. An environmental assessment result grading unit is configured to grade the environmental coupling risk value and label the corresponding environmental impact factors of each grade.

7. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 6, characterized in that, The coupling risk assessment unit calculates the environmental coupling risk value by: Calling the preset threshold value through the environmental parameter threshold configuration unit, comparing each actual environmental parameter value with the corresponding risk threshold value one by one, and calculating the single-parameter threshold-exceeding degree. Determining the coupling coefficients of each environmental parameter and the operation risk of the storage tank based on the historical risk event database. Calculating the environmental coupling risk value using a coupling risk synthesis formula. Outputting the environmental coupling risk value and the risk contribution proportion of each environmental parameter.

8. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 7, characterized in that, The collaborative adaptation module includes: A collaborative rule library for storing preset collaborative rules. A rule matching algorithm unit that uses a weighted priority matching algorithm to set weight coefficients based on the risk thresholds, and filters the most suitable collaborative rules. A rule dynamic updating unit that iteratively updates the preset collaborative rules based on historical collaborative effect evaluation data and new risk event data.

9. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 1, characterized in that, The decision-making module includes: A strategy generation unit that generates collaborative management strategies based on the most suitable collaborative rules, including inventory adjustment strategies, device control strategies, and early warning notification strategies; the inventory adjustment strategies include feed rate adjustment, discharge rate adjustment, and inventory transfer path planning; the device control strategies include valve opening degree adjustment, pump group start-stop control, and safety protection device start control; the early warning notification strategies include sending early warning information to designated personnel according to the evaluation grade. An instruction issuing unit that converts the collaborative management strategies into standardized control instructions and issues them to the corresponding execution terminals through an encrypted communication channel. A strategy execution feedback unit that receives strategy execution result feedback data from the execution terminals, monitors the execution effect of the collaborative management strategies in real time, and outputs the feedback data. A strategy optimization unit that optimizes and adjusts the collaborative management strategies in real time based on the feedback data and the initial risk prediction results.

10. The oil and chemical storage tank area inventory collaborative management system based on end-cloud collaboration according to claim 9, characterized in that, The strategy optimization unit optimizes and adjusts the collaborative management strategies in real time by: Collecting feedback data transmitted by the strategy execution feedback unit, including real-time inventory data, device operation state data, environmental data updates, and risk level change data. Building a strategy execution effect evaluation index system, including risk control efficiency, execution cost, and inventory stability. Calculating the deviation rate of each evaluation index from the preset target value. If the absolute value of the deviation rate exceeds the preset deviation range, the strategy optimization process is triggered. Based on the deviation reason analysis results, the collaborative management strategy parameters are adjusted accordingly, including adjusting the rate parameters of the inventory adjustment strategy or the response threshold of the device control strategy if the risk control efficiency is insufficient; optimizing the inventory transfer path planning or the device start-stop timing if the execution cost is too high; adjusting the inventory safety threshold interval and the early warning trigger condition if the inventory stability is not up to standard. The optimized strategy parameter is input into the digital twin for simulation verification, and if the simulation result meets the evaluation index requirement, a formal optimized strategy is generated; if not, the collaborative management strategy parameter is adjusted again. The formal optimized strategy is sent to the instruction issuing unit, and the collaborative management strategy parameters before and after optimization and the execution effect data are recorded.

Images