Oil storage and transportation system temperature control system and method thereof

Abstract

The invention discloses a petroleum storage and transportation system temperature control system and method, and belongs to the technical field of petroleum storage and transportation. The method aims at solving the problems that an existing temperature control strategy lags behind and is high in energy consumption, and changes of a temperature field in fluid cannot be accurately predicted. The method provided by the invention comprises the following steps: acquiring real-time multi-source state data of fluid, wherein the real-time multi-source state data comprises environment prediction data, physical property data and internal state data; discretizing a fluid space into a three-dimensional voxel grid; constructing a thermal inertia field representing thermal response characteristics of each voxel and an external influence matrix representing future external heat exchange; in combination with the current temperature distribution, the thermal inertia field and the external influence matrix, generating a predictive thermal state grid at the tail end of the preset time range through a state propagation algorithm; key control voxels with the temperature most likely to deviate from the safety range are recognized in the prediction grid, and a temperature control instruction is determined.

Description

Technical Field

[0001] This application relates to the field of petroleum storage and transportation technology, and in particular to a temperature control system and method for petroleum storage and transportation systems. Background Technology

[0002] As the lifeblood of modern industry, the safe and efficient storage and transportation of petroleum are of paramount importance. In petroleum storage and transportation systems, temperature is a critical control parameter, whether in large storage tanks or long-distance pipelines. Precise temperature control directly affects energy consumption, oil quality, operational safety, and equipment lifespan. For example, for high-pour-point, high-wax crude oil, excessively low temperatures can cause wax crystallization, clogging pipelines, valves, and filters, and in severe cases, paralyzing the entire transportation system. Conversely, excessively high temperatures can exacerbate the evaporation loss of lighter components, causing economic losses and environmental pollution, while also potentially creating safety hazards.

[0003] In existing technologies, temperature control methods for oil storage and transportation systems are typically quite rudimentary. A common approach is feedback control based on a single point or a few temperature monitoring points. The control system monitors the temperature at one or a few preset locations, and activates heating or cooling equipment when the temperature falls below or exceeds a set threshold. The drawbacks of this method are obvious: First, it is passive, only reacting after a problem has occurred (the temperature has deviated), resulting in significant control lag. Second, the temperature field distribution within large storage tanks or long-distance pipelines is extremely uneven, influenced by a variety of complex factors such as solar radiation, diurnal variations in ambient temperature, wind speed, and natural convection of the fluid within the tank. The temperature at only a few points cannot accurately reflect the overall thermal state of the entire fluid space. This rudimentary control based on localized information often leads to unnecessary energy waste (such as overheating) or potential risks (such as undetected excessively low temperatures in certain areas).

[0004] Another type of improvement method attempts to introduce more complex control algorithms, but often relies on simplified lumped parameter models, treating the entire tank as one or a few uniform temperature zones. While these models improve control performance to some extent, they still cannot accurately depict the spatiotemporal evolution of the three-dimensional temperature field within the fluid caused by thermal stratification, boundary effects, and other factors. Especially when dealing with strong external disturbances such as drastic weather changes, the predictive power and control accuracy of these models are significantly reduced. Summary of the Invention

[0005] The purpose of this application is to provide a temperature control system and method for an oil storage and transportation system, which aims to improve the technical problems existing in the background technology, such as lagging temperature control strategies, inability to accurately predict and respond to changes in the internal temperature field of fluids under complex working conditions, resulting in high energy consumption and high operational risks.

[0006] To achieve the above objectives, this application provides the following technical solutions:

[0007] In a first aspect, this application provides a temperature control method for an oil storage and transportation system, comprising: acquiring real-time multi-source state data of fluids in the oil storage and transportation system, wherein the multi-source state data includes environmental prediction data within a preset time range, physical property data of the fluid itself, and internal state data collected by a sensor array deployed inside the oil storage and transportation system; discretizing the fluid space contained in the system into a three-dimensional voxel grid based on the spatial structure of the oil storage and transportation system, wherein the three-dimensional voxel grid is composed of multiple fluid voxels; and for each fluid voxel, constructing a characterization of the fluid voxel based on the physical property data of the fluid and the position information of the fluid voxel. The system employs a thermal inertia field to assess the response characteristics to heat changes. Based on the environmental prediction data and the structural parameters of the oil storage and transportation system, an external influence matrix characterizing the heat exchange between the boundary of the oil storage and transportation system and the outside world within a preset time range is constructed. Combining the thermal inertia field, the external influence matrix, and the current temperature distribution extracted from the internal state data, a predictive thermal state grid is generated at the end of the preset time range. Within the predictive thermal state grid, at least one key control voxel with its temperature at a preset critical state is identified. Based on the deviation between the predicted temperature and the target temperature of the key control voxel, a temperature control command for the oil storage and transportation system is determined. This scheme, by constructing a refined three-dimensional model of the fluid and combining it with the quantification of future environmental impacts and the internal thermal characteristics of the fluid, can predictively generate the global temperature distribution at future moments. This enables the identification of potential risk points and the implementation of forward-looking and precise control, solving the problems of lag and locality in traditional methods and improving energy efficiency and safety.

[0008] In one possible implementation of the first aspect, acquiring the real-time multi-source state data of the fluid includes: acquiring predicted data of ambient temperature, solar radiation intensity, and wind speed for the next 24 hours, updated at a preset frequency, as the environmental prediction data; acquiring the density, specific heat capacity, and thermal conductivity of the fluid as the physical property data; and acquiring real-time temperature and pressure values ​​collected by multiple temperature and pressure sensors deployed at different depths within the fluid space as the internal state data.

[0009] In one possible implementation of the first aspect, constructing the thermal inertia field includes: assigning a unique spatial coordinate index to each fluid voxel in the three-dimensional voxel grid; for each fluid voxel, inputting its corresponding fluid density, specific heat capacity, volume, and the spatial coordinate index to a preset thermal inertia function, and outputting a scalar value as the thermal inertia value of the fluid voxel; and constructing the thermal inertia field from the thermal inertia values ​​of all fluid voxels.

[0010] In one possible implementation of the first aspect, constructing the external influence matrix includes: dividing the outer surface of the oil storage and transportation system into multiple boundary units; for each boundary unit, at each discrete time point within the preset time range, constructing an external heat flux function by combining the environmental prediction data and the heat transfer properties of the boundary unit to obtain the predicted heat flux value of the boundary unit at that time point; and constructing the external influence matrix by the predicted heat flux values ​​of all boundary units at all discrete time points.

[0011] In one possible implementation of the first aspect, generating the predictive thermal state grid includes: initializing a state propagation grid; performing iterative propagation at a preset time step; in each iteration step, for each fluid voxel, updating its temperature state based on its own and adjacent voxels' current temperatures, the thermal conductivity coefficients between the fluid voxels, and the thermal inertia field; wherein, for fluid voxels adjacent to the boundary of the oil storage and transportation system, its temperature update also incorporates the corresponding predicted heat flux value in the external influence matrix; after the iteration is completed, the final state propagation grid is used as the predictive thermal state grid.

[0012] In one possible implementation of the first aspect, identifying at least one key control voxel whose temperature is in a preset critical state includes: traversing the predicted temperature values ​​of all fluid voxels in the predictive thermal state grid; filtering out fluid voxels whose predicted temperature values ​​are lower than a preset low temperature threshold or higher than a preset high temperature threshold; and identifying the fluid voxel with the lowest or highest predicted temperature value as the key control voxel.

[0013] In one possible implementation of the first aspect, after determining the temperature control command, the method further includes: at a preset calibration time after the execution of the temperature control command, acquiring the actual temperature distribution of the fluid space and generating an actual thermal state grid; comparing the actual thermal state grid with a predictive thermal state grid generated for the calibration time to generate a deviation matrix; based on the deviation matrix, generating a calibration parameter field for correcting the thermal inertia field or the external influence matrix; and in subsequent temperature control method execution, applying the calibration parameter field to adjust the thermal inertia field or the external influence matrix.

[0014] Secondly, this application provides a temperature control system for an oil storage and transportation system, including: a data acquisition module, a spatial discretization module, a thermal field construction module, a state prediction module, and a decision control module. Attached Figure Description

[0015] Figure 1 A schematic flowchart of a temperature control method for an oil storage and transportation system provided for some embodiments of this application;

[0016] Figure 2 A schematic flowchart of a temperature control method for an oil storage and transportation system provided for other embodiments of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0018] In the following description, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] Furthermore, in this application, directional terms such as "upper," "lower," "left," and "right" may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and may change accordingly depending on the orientation of the components in the accompanying drawings.

[0020] In this application, unless otherwise expressly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral part; it can be a direct connection or an indirect connection through an intermediate medium. Furthermore, the term "electrical connection" can refer to the manner in which an electrical connection is used to achieve signal transmission.

[0021] As used herein, “about,” “approximately,” or “approximately” includes the stated value and a reference value within an acceptable range of deviation from the given value, characterized in that the acceptable range of deviation is determined by a person skilled in the art taking into account the measurement under discussion and the error associated with the measurement of the given quantity (i.e., the limitations of the measurement method).

[0022] To make the objectives, technical solutions, and advantages of this application clearer and more explicit, the following detailed description of this application is provided with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0023] This application provides a temperature control system and method for an oil storage and transportation system, aiming to solve the problems of outdated temperature control strategies, high energy consumption, and inability to accurately predict and cope with changes in the internal temperature field of fluids under complex operating conditions (such as drastic changes in ambient temperature and the influence of solar radiation) in existing technologies. The technical solution of this application constructs a refined three-dimensional model of the fluid space and combines it with multi-source data to generate a predictive thermal state distribution, thereby achieving proactive temperature control, improving energy utilization efficiency and system operational safety.

[0024] The following will describe in detail the specific implementation process of the temperature control method for an oil storage and transportation system provided in this application. In this embodiment, a large open-air crude oil storage tank is used as a specific application scenario for the oil storage and transportation system.

[0025] The temperature control method can be implemented by a controller installed locally in the oil storage and transportation system, or by a server deployed in the cloud. It completes the entire control process by communicating with sensors, actuators (such as heaters and coolers) within the storage and transportation system, and external data sources. Figure 1 As shown, the method includes:

[0026] S100: Obtain real-time multi-source status data of the fluid in the oil storage and transportation system.

[0027] The multi-source state data is a composite dataset, which can be logically divided into three main categories: environmental prediction data, fluid property data, and internal state data.

[0028] Environmental forecast data refers to a quantitative description of the changing trends of the external environment surrounding a storage tank over a future period. Since the heat loss and absorption of storage tanks, especially large open-air tanks, are closely related to the external environment, obtaining environmental change trends in advance is crucial for predictive control. In this embodiment, the preset time range can be set to the next 24 hours. The system obtains the ambient temperature, measured in hours, for the next 24 hours by connecting to a commercial meteorological service API (Application Programming Interface) or a self-built meteorological monitoring station. Total solar radiation intensity and average wind speed The predicted sequence.

[0029] Fluid physical property data refers to the inherent physicochemical properties of crude oil stored in tanks. These properties directly determine the response characteristics of crude oil under thermal stimulation. This data is typically measured when crude oil is received into storage, or retrieved from a crude oil database based on batch number. It mainly includes: the average density of the crude oil. (unit: ), specific heat capacity (unit: ), and thermal conductivity (unit: For certain special crude oils, such as high-wax crude oils, information such as wax precipitation temperature profiles may also be included. These data are typically treated as constants within a single control cycle.

[0030] Internal state data refers to direct measurements reflecting the current thermodynamic state of crude oil, collected in real time by sensors deployed inside the storage tank. To capture the non-uniformity of the temperature field within the tank (e.g., higher temperatures at the top due to solar radiation, and lower temperatures at the bottom and near the walls), this solution employs a sensor array for measurement. For example, in a cylindrical storage tank, measurements can be taken at multiple radial locations along the central axis and near the tank wall at different depths. A series of temperature and pressure sensors are arranged on top. These sensors collect temperature values ​​at each measuring point at a high frequency (e.g., once per minute). and pressure value This forms a discrete snapshot of the internal state.

[0031] S200. Based on the spatial structure of the oil storage and transportation system, the fluid space contained in the system is discretized into a three-dimensional voxel grid.

[0032] The system acquires the geometric information of the storage tank; for example, for a cylindrical storage tank, it needs to know its radius. and current liquid level height This is determined by the radius. and height A defined cylindrical fluid space is embedded in a regular Cartesian coordinate system and divided into a three-dimensional grid. This grid consists of a large number of closely packed, regularly shaped tiny cubic units, each of which is called a fluid voxel, denoted as . ,in It is the integer coordinate index of the voxel in the three-dimensional raster.

[0033] voxel size , , This determines the spatial resolution. Higher resolution yields more accurate results, but also increases computational complexity. Choosing an appropriate voxel size requires a trade-off between accuracy and computational efficiency. For example, for a voxel with a diameter of... Liquid level height is The storage tank can be divided into a The grid, with each voxel having a size of .

[0034] For irregularly shaped regions (such as the bottom or top of a tank), some voxels may only be filled by the fluid portion. During discretization, labeling is necessary. A simple approach is to treat a voxel as a valid fluid voxel if its center point lies within the fluid region; otherwise, it is considered a boundary or empty voxel. Thus, the entire fluid space is mapped to a voxel containing... A collection of n elements, each element Each element carries a Boolean flag indicating whether it is a valid fluid voxel. This three-dimensional voxel grid... It is the foundational data structure for all subsequent space-related calculations.

[0035] S300. For each of the fluid voxels, construct a thermal inertia field characterizing the thermal change response characteristics of the fluid voxel.

[0036] The thermal inertia field is not directly measured, but rather a derived data field constructed based on physical property data and spatial information. Its physical significance lies in quantifying the ease with which the temperature of each fluid voxel changes after absorbing or releasing a unit of heat. The greater the thermal inertia, the less likely the temperature of that voxel is to change, and the better its thermal stability.

[0037] The execution process for this step is as follows:

[0038] for Each effective fluid voxel in to give it A unique identifier in the coordinate system. The system constructs a proprietary thermal inertia function. Understandably, the thermal inertia of a voxel depends not only on its own mass and specific heat capacity, but may also be affected by its position throughout the fluid space.

[0039] In one implementation, the thermal inertia function can be defined as the heat capacity of a voxel. That is:

[0040]

[0041] in, It is the density of crude oil. It is specific heat capacity. It is the volume of a single voxel. Under this definition, all voxels have the same thermal inertia.

[0042] However, to characterize the system more precisely, this application proposes a location-dependent thermal inertia function. The underlying logic is that the equivalent thermal response characteristics of fluid at different locations may differ due to factors such as fluidity and boundary effects. For example, the fluid closer to the heating coil experiences more active heat transfer, and its equivalent thermal inertia may appear smaller in the macroscopic response. Therefore, It is constructed as a function that includes a position correction term:

[0043]

[0044] in, It is a dimensionless position weighting factor, which is a factor relating to coordinates. The function. For example, for voxels near the tank wall, A smaller value can be chosen to reflect that its heat is more easily lost through the tank wall, exhibiting lower effective thermal stability. For the voxel at the center of the tank, A relatively large value can be chosen. The introduction of this weighting factor allows the thermal inertia field to adapt more flexibly to complex internal thermal environments.

[0045] For example, a heating coil can be constructed in a cylindrical storage tank with a heating coil in the central region of the bottom. A function is used to simulate the central thermal plume effect. First, the voxels are indexed using Cartesian coordinates. Convert to normalized cylindrical coordinates :

[0046] Normalized radial distance ,in It is an index for the storage tank center. It is the index corresponding to the maximum radius. The range of values ​​is .

[0047] Normalized height ,in It is the index corresponding to the highest liquid level. The range of values ​​is .

[0048] Based on this, a weight function of the following form can be constructed:

[0049] In this function:

[0050] It is a positive, dimensionless effect intensity coefficient used to control the overall magnitude of the correction. The negative sign indicates that the effect mainly reduces the equivalent thermal inertia to simulate accelerated heat transfer.

[0051] It is a Gaussian function of radial distance, which makes the correction effect occur at the center of the tank ( The intensity is strongest and gradually weakens towards the tank wall. It is a parameter that controls the radial influence range of the thermal plume.

[0052] It is an exponentially decaying function with respect to height, which makes the correction effect occur at the bottom of the tank ( The intensity is strongest and decreases with increasing altitude, used to simulate the dissipation of energy in a thermal plume. It is a parameter that controls the rate of vertical decay of the effect.

[0053] By setting such a set of initial parameters (for example, , , This weighting function can create a "rapid thermal response channel" in the thermal inertia field that conforms to physical expectations, with a strong center, strong bottom, and gradually weakening outwards and upwards. More importantly, these parameters... , , It can be systematically adjusted in subsequent calibration feedback steps (S700) so that the model can continuously approximate the actual thermodynamic behavior of a specific tank.

[0054] Each of the calculations Values ​​filled back corresponding Location, thus forming a relationship with A three-dimensional scalar field with the same dimensions—a thermal inertia field .

[0055] S400. Based on the environmental prediction data and the structural parameters of the oil storage and transportation system, construct an external influence matrix characterizing the amount of heat exchange between the boundary of the oil storage and transportation system and the outside world within the preset time range. This step is used to quantify the driving effect of the external environment on the future thermal state of the storage tank system.

[0056] Corresponding to the internal thermal inertia field, the external influence matrix This describes the heat "input" and "output" at the system boundary. The construction process for this step is as follows:

[0057] The outer surface of the storage tank is discretized. The outer surface of the storage tank, including the top cover, side walls, and bottom plate, is divided into a series of small surface units, denoted as . ,in It is the index of the surface unit. Each All are associated with one or more outermost fluid voxels Adjacent. Simultaneously, it is necessary to obtain the structural parameters of the storage tank, such as the thermal conductivity of the tank wall material. ,thickness and the solar radiation absorption rate of the outer surface. and thermal emissivity .

[0058] The system constructs a proprietary external heat flux function. This function is used to calculate the time at each discrete point in the future. (in ,and (hours), each boundary unit Net heat flux (unit: ).

[0059] For example, for a surface unit on the top cover of a storage tank Its time Net heat flux It can be built in the following ways:

[0060]

[0061] in, It is the absorbed solar radiation heat flux. It is determined by the total solar radiation intensity from environmental prediction data. and surface absorption rate Decide: For ease of calculation, the side walls and roof are calculated using the same method. For the floor slab and nighttime, this item is zero.

[0062] It is the heat flux of convective heat transfer. It is mainly determined by the temperature of the outer wall of the tank. With ambient temperature The temperature difference determines the value, and its expression is: The convective heat transfer coefficient here It is itself related to wind speed Related empirical functions, for example, ,in , , It is an empirical constant. Due to the outer wall temperature... Since it is unknown, in practical calculations, it is usually correlated with the temperature of the outermost fluid inside the tank through a thermal resistance network for iterative solution. Alternatively, temperature sensors can be deployed on the outer wall of the tank to directly measure the outer wall temperature in real time.

[0063] It is the heat flux transferred through radiation between the sky and the surrounding environment. It follows the Stefan-Boltzmann law: ,in It is the Stefan-Boltzmann constant. It is the effective sky temperature, which can be estimated based on ambient temperature and cloud cover, and can be determined by a person skilled in the art using appropriate methods.

[0064] By establishing a corresponding heat flux function for each boundary (top cover, side wall, bottom plate) and substituting it with the environmental prediction data sequence for the next 24 hours, the system can calculate the heat flux of each boundary unit. At each future point in time Predicted heat flux value .

[0065] Finally, all these calculation results are organized. External Influence Matrix It is a two-dimensional matrix whose row indices correspond to boundary elements. The column index corresponds to a point in time. Each element in the matrix The value is This matrix provides a complete spatiotemporal depiction of the thermal effects of the external environment on the storage tank over the next 24 hours.

[0066] S500: By combining the thermal inertia field, the external influence matrix, and the current temperature distribution extracted from the internal state data, a predictive thermal state grid is generated at the end of the preset time range.

[0067] The purpose of this step is to deduce the temperature of each fluid voxel in the tank at a future time (24 hours later in this embodiment) based on the current state (initial conditions) and future internal and external influences (boundary conditions and physical properties) using a deterministic state propagation algorithm.

[0068] The execution flow for this step is as follows:

[0069] The system creates a connection with A three-dimensional grid of the same size is called a state propagation grid. The initial value of this grid, i.e. Temperature distribution over time The temperature is determined by internal state data collected in S100. Since the sensors are discretely distributed, interpolation algorithms (such as Kriging interpolation or inverse distance weighted interpolation) are needed to convert the discrete measurement point temperatures into values. Extend to the whole For each voxel Assign an initial temperature.

[0070] This application constructs a discretized heat conduction state propagation model. It can be understood that at each tiny time step... Inside, a voxel The temperature change depends on the amount of heat exchanged between it and its neighboring voxels and with the external environment.

[0071] For a voxel located inside the fluid and not in contact with the tank wall , its in Temperature of Time Determined by the following logic:

[0072]

[0073] in, From the thermal inertia field The value of thermal inertia of this voxel was found in the database. Is The net heat that flows into the voxel during the time period. It is determined by the temperature difference between the voxel and its six adjacent voxels (in the i, j, k directions).

[0074]

[0075] here, It is a voxel The set of adjacent voxels, It is the thermal conductivity of crude oil. It is the contact area between voxels (e.g.) ), It is the distance between the centers of adjacent voxels (e.g.) This formula is, understandably, essentially a discretized form of Fourier's law of heat conduction.

[0076] For voxels adjacent to the tank boundary (such as the tank wall or top cover), the net heat calculation... An additional factor needs to be added: heat exchange from the external environment. This factor is directly derived from the external influence matrix. Obtained from the top cover. For example, a... Adjacent voxels of boundary units Its net calorie The calculation needs to add one item:

[0077]

[0078] in It is the time point corresponding to the current propagation step.

[0079] The system from Start with a pre-set, sufficiently small time step. (For example, 10 seconds) Proceed forward. At each time step, the system processes all fluid voxels in parallel. Applying the state propagation rules described above, we calculate their temperatures at the next time step, thereby updating the entire... Grid. This process is repeated continuously. The temperature field in the middle is like a wave, starting from the initial state and gradually evolving under the combined influence of the thermal inertia field and the external influence matrix.

[0080] The iterative process continues until the preset time range is reached, i.e., after 24 hours. At this point... The grid is the predictive thermal state grid that is the final output of this step. . Each element in That is, the contents of the storage tank after 24 hours. A deterministic prediction of the temperature of crude oil at location.

[0081] S600: In the predictive thermal state grid, identify at least one key control voxel whose temperature is in a preset critical state, and determine a temperature control command for the oil storage and transportation system based on the deviation between the predicted temperature and the target temperature of the key control voxel.

[0082] This step will transform complex, high-dimensional prediction results ( This is transformed into one or more specific, executable control actions. The core idea is to find the "weak link" most likely to cause problems from a global predictive perspective and then intervene in a targeted manner.

[0083] The execution flow for this step is as follows:

[0084] For crude oil storage and transportation, critical conditions are typically related to safety and quality. For example, a low-temperature threshold can be defined. and a high temperature threshold . This is usually set at a safety margin above the wax precipitation point of crude oil to prevent wax crystallization from clogging pipelines or affecting pumping. These thresholds are used to control crude oil evaporation losses and to meet the upper temperature limits required by certain processes. These two thresholds constitute the temperature window for safe operation. .

[0085] The system will traverse the predictive thermal state grid. Examine the predicted temperature of all fluid voxels. If there are one or more voxels with predicted temperatures The system will then consider a low-temperature risk to exist. Among these voxels, the one with the lowest predicted temperature value will be identified as the "low-temperature critical control voxel." .

[0086] Similarly, if there are one or more voxels with predicted temperatures The system will consider the presence of a high-temperature risk and identify the voxel with the highest predicted temperature value as a "high-temperature critical control voxel". .

[0087] In some operating conditions, there may be areas with both low and high temperature risks, or no risks at all. The output of this step is the coordinates of at least one key control voxel and its predicted temperature.

[0088] Once the critical control voxel is identified, the system needs to calculate how much control input needs to be applied to correct the predicted deviation.

[0089] For example, suppose that a cryogenic key control voxel has been identified. Its predicted temperature is The control objective is to ensure that, within the next 24 hours, the final temperature of the heating system is not lower than the target temperature. (generally It can be set to or slightly higher Total heat required. It is not only related to the temperature difference, but also to the thermal inertia of the key voxel and the heat transfer characteristics of the entire system.

[0090] A refined control instruction generation logic is as follows:

[0091] Temperature difference is .

[0092] To make Temperature rise The amount of heat that needs to be injected into it At least for ,in This is the thermal inertia of the key voxel. However, because heat diffuses to the surroundings, the actual total heat required by the heating system is... Much larger .

[0093] In some embodiments, As input, a simplified state propagation model is run in reverse to estimate the total heating power required. Alternatively, a pre-calibrated control function can be used:

[0094]

[0095] This function The output is the heater's power setpoint (unit: kW). This function can be a polynomial function or a lookup table, which implicitly represents the heat transfer efficiency from the heater to the key control voxel and is calibrated using offline simulation or historical data. For example, the function is:

[0096]

[0097] It is a basic proportional gain coefficient (unit: It can be tuned through system identification or offline simulation. It is a dimensionless heat transfer efficiency function, the value of which is in Between. This function reflects the distance from the heating source to the key voxel location. The effectiveness of heat transfer. Generally, the farther and more remote the location is from the heat source, the less effective the heat transfer. The lower the value, the lower the efficiency function. It can be pre-modeled as a lookup table of spatial coordinates or a simple distance decay function.

[0098] The final generated temperature control command can be "in the next..." Within hours, with "Turn on the heating system at the specified power," or for scenarios with a cooling system, "Turn on the heating system at the specified power." "Power up the cooling system." This command is sent to the tank's actuator (PLC controller) to execute.

[0099] In this way, the control system no longer passively responds to current low or high temperature events, but anticipates risks before events occur and calculates just the right amount of energy input to bring the future temperature curve back to a safe range, thereby achieving dual optimization of energy consumption and safety.

[0100] like Figure 2 As shown, to further improve the long-term prediction accuracy of the model, this application also provides a calibration feedback mechanism. The specific steps are as follows:

[0101] S710. At a preset calibration time after the temperature control command is executed, the actual temperature distribution of the fluid space is collected, and an actual thermal state grid is generated.

[0102] After the control commands have been executed for a period of time (e.g., at the midpoint of the 24-hour prediction cycle, i.e., after 12 hours), the system triggers the internal sensor array again to collect a new, dense set of actual temperature measurements. Using the same interpolation method as the S500 initialization steps, a real thermal state grid representing the actual situation 12 hours later is generated. .

[0103] S720. Compare the actual thermal state grid with the predictive thermal state grid generated for the calibration time to generate a deviation matrix.

[0104] At the same time, when the system initially performs the S500 state propagation calculation, it not only saves the final result for the 24th hour. It also saves the intermediate results for the 12th hour, namely the predictive thermal state grid after 12 hours, denoted as .

[0105] Then, the system performs a voxel-by-voxel subtraction operation on the two grids that are identical in dimensions to generate a deviation matrix (or deviation grid). :

[0106]

[0107] Deviation matrix Each element in the model quantifies its performance. Prediction error at location. Positive values ​​indicate that the model predicts a colder location, while negative values ​​indicate that the model predicts a hotter location.

[0108] S730. Based on the deviation matrix, generate a calibration parameter field for correcting the thermal inertia field or the external influence matrix.

[0109] Systematic and persistent deviations usually mean that some parameters in the model do not match reality. For example, if the actual temperature in a certain area of ​​a storage tank is always lower than the predicted value, it may mean that the equivalent thermal inertia of that area is... It has been underestimated, or the corresponding external heat loss in this area is underestimated. It's been underestimated.

[0110] The purpose of this step is to use the deviation matrix... The system intelligently adjusts model parameters. A calibration function is constructed. Its input is the deviation matrix. The output is a calibration parameter field. .

[0111]

[0112] For example, It can be implemented as a simple, error-sign-and-magnitude-based iterative learning rule to update the multiplicative correction coefficients of the thermal inertia field. For each voxel Its calibration parameters It can be generated through the following logic:

[0113]

[0114] in, It is the prediction error of this voxel. . It is a small positive number, called the learning rate (e.g., It controls the adjustment step size for each calibration to prevent over-adjustment. It is a characteristic temperature difference (e.g., ), used to convert errors Normalization ensures that its impact is within a reasonable range. The function will The value is limited to Within the range, ensure that the correction term is not too large.

[0115] If the forecast is cold ( (The actual temperature is higher), indicating that the model's equivalent thermal inertia has been overestimated. It will become a number less than 1, and the thermal inertia will decrease in the next iteration.

[0116] If the forecast is overheated ( (The actual temperature is even lower), indicating that the model's equivalent thermal inertia has been underestimated. It will become a number greater than 1, increasing the thermal inertia in the next iteration. If the prediction is accurate ( ),but

[0117] Basically, no adjustments will be made.

[0118] S740. In the subsequent execution of the temperature control method, the calibration parameter field is applied to adjust the thermal inertia field or the external influence matrix.

[0119] This calibration parameter field will be applied at the start of the next 24-hour prediction cycle when the system executes S300 to construct the thermal inertia field or S400 to construct the external influence matrix.

[0120] For example, the updated thermal inertia field It can be calculated in the following ways:

[0121]

[0122] In this way, the model can learn from its own prediction errors and continuously self-correct, making its depiction of specific storage tanks and specific oil products increasingly accurate. This adaptive calibration feedback loop ensures that this technical solution can maintain a high level of prediction accuracy and control performance during long-term operation.

[0123] This application also provides a temperature control system for an oil storage and transportation system, designed to execute the aforementioned temperature control method. This system can be an integrated hardware device or a software system deployed on a server. Its internal functional modules correspond to the steps of the method.

[0124] Specifically, the system includes:

[0125] A data acquisition module is responsible for communicating with various sensors and data sources, executing the data acquisition task described in S100. A spatial discretization module, containing the geometric model of the storage tank, executes the voxel grid generation task described in S200. A thermal field construction module, the core of which implements... and Functions are used to execute S300 and S400, constructing the thermal inertia field and external influence matrix. A state prediction module, whose core implements the state propagation algorithm described in S500, serves as the system's computational engine. A decision control module implements the logic described in S600, responsible for extracting control commands from the prediction results.

[0126] Optionally, the system may also include a calibration feedback module specifically designed to perform the adaptive calibration process from S710 to S740 and feed back the generated calibration parameter field to the thermal field construction module or the state prediction module to achieve closed-loop optimization of the model.

[0127] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0128] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0129] The units described as separate components may or may not be physically separate. A component shown as a unit can be one physical unit or multiple physical units; that is, it can be located in one place or distributed in multiple different places. Depending on actual needs, some or all of the units can be selected to achieve the purpose of this embodiment.

[0130] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware.

[0131] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A temperature control method for an oil storage and transportation system, characterized in that, include: The system acquires real-time multi-source state data of the fluid in the oil storage and transportation system. The multi-source state data includes environmental prediction data within a preset time range, the fluid's own physical property data, and internal state data collected by a sensor array deployed inside the oil storage and transportation system. Based on the spatial structure of the oil storage and transportation system, the fluid space contained in the system is discretized into a three-dimensional voxel grid, which is composed of multiple fluid voxels. For each fluid voxel, a thermal inertia field characterizing the thermal change response characteristics of the fluid voxel is constructed based on the fluid's physical property data and the voxel's position information. Based on the environmental prediction data and the structural parameters of the oil storage and transportation system, an external influence matrix is ​​constructed to characterize the heat exchange between the boundary of the oil storage and transportation system and the outside world within the preset time range. By combining the thermal inertia field, the external influence matrix, and the current temperature distribution extracted from the internal state data, a predictive thermal state grid is generated at the end of the preset time range. In the predictive thermal state grid, at least one key control voxel whose temperature is in a preset critical state is identified, and based on the deviation between the predicted temperature and the target temperature of the key control voxel, a temperature control command for the oil storage and transportation system is determined.

2. The method according to claim 1, characterized in that, The acquisition of the real-time multi-source state data of the fluid includes: The predicted environmental temperature, solar radiation intensity, and wind speed for the next 24 hours, updated at a preset frequency, are obtained as the environmental prediction data. The density, specific heat capacity, and thermal conductivity of the fluid are obtained as the physical property data. The real-time temperature and pressure values ​​collected by multiple temperature and pressure sensors deployed at different depths within the fluid space are used as the internal state data.

3. The method according to claim 1, characterized in that, The construction of the thermal inertia field includes: Assign a unique spatial coordinate index to each fluid voxel in the three-dimensional voxel grid; For each fluid voxel, its corresponding fluid density, specific heat capacity, volume, and spatial coordinate index are input into a preset thermal inertia function, and a scalar value is output as the thermal inertia value of the fluid voxel. The thermal inertia field is composed of the thermal inertia values ​​of all fluid voxels.

4. The method according to claim 1, characterized in that, The construction of the external influence matrix includes: The outer surface of the oil storage and transportation system is divided into multiple boundary units; For each boundary cell, at each discrete time point within the preset time range, an external heat flux function is constructed by combining the environmental prediction data with the heat transfer properties of the boundary cell to obtain the predicted heat flux value of the boundary cell at that time point. The external influence matrix is ​​composed of the predicted heat flux values ​​of all boundary elements at all discrete time points.

5. The method according to claim 1, characterized in that, The generation of the predictive thermal state grid includes: Initialize a state propagation grid, the initial temperature value of which is assigned by the real-time temperature value in the internal state data; Iterative propagation is performed within the preset time range at a preset time step; In each iteration step, for each fluid voxel, its temperature state is updated based on its own and adjacent voxels' current temperature, the thermal conductivity coefficient between the fluid voxels, and the thermal inertia field; wherein, for the fluid voxel adjacent to the boundary of the oil storage and transportation system, its temperature update also incorporates the corresponding predicted heat flux value in the external influence matrix. After the iteration is completed, the final state propagation grid is used as the predictive thermal state grid.

6. The method according to claim 1, characterized in that, The identification of at least one key control voxel whose temperature is in a preset critical state includes: In the predictive thermal state grid, the predicted temperature values ​​of all fluid voxels are traversed. Filter out fluid voxels whose predicted temperature values ​​are lower than a preset low temperature threshold or higher than a preset high temperature threshold; The fluid voxel with the lowest or highest predicted temperature value is identified as the key control voxel.

7. The method according to any one of claims 1-6, characterized in that, After determining the temperature control command, the process also includes: At a preset calibration time after the temperature control command is executed, the actual temperature distribution of the fluid space is collected to generate an actual thermal state grid. The actual thermal state grid is compared with the predicted thermal state grid generated for the calibration time to generate a deviation matrix; Based on the deviation matrix, a calibration parameter field is generated to correct the thermal inertia field or the external influence matrix; In the subsequent temperature control process, the calibration parameter field is used to adjust the thermal inertia field or the external influence matrix.

8. A temperature control system for an oil storage and transportation system, characterized in that, include: The data acquisition module is used to acquire real-time multi-source state data of the fluid in the oil storage and transportation system. The multi-source state data includes environmental prediction data within a preset time range, the physical property data of the fluid itself, and internal state data collected by a sensor array deployed inside the oil storage and transportation system. A spatial discretization module is used to discretize the fluid space contained in the oil storage and transportation system into a three-dimensional voxel grid based on the spatial structure of the oil storage and transportation system. The three-dimensional voxel grid is composed of multiple fluid voxels. The thermal field construction module is used to construct a thermal inertia field characterizing the thermal change response characteristics of each fluid voxel based on the physical property data of the fluid and the position information of the fluid voxel, and to construct an external influence matrix characterizing the heat exchange between the boundary of the oil storage and transportation system and the outside world within a preset time range based on the environmental prediction data and the structural parameters of the oil storage and transportation system. The state prediction module is used to combine the thermal inertia field, the external influence matrix, and the current temperature distribution extracted from the internal state data to generate a predictive thermal state grid at the end of the preset time range. The decision control module is used to identify at least one key control voxel in the predictive thermal state grid where the temperature is in a preset critical state, and to determine the temperature control command for the oil storage and transportation system based on the deviation between the predicted temperature and the target temperature of the key control voxel.

9. The system according to claim 8, characterized in that, The thermal field construction module is specifically used for: Assign a unique spatial coordinate index to each fluid voxel in the three-dimensional voxel grid; For each fluid voxel, its corresponding fluid density, specific heat capacity, volume, and spatial coordinate index are input into a preset thermal inertia function, and a scalar value is output as the thermal inertia value of the fluid voxel. The thermal inertia field is composed of the thermal inertia values ​​of all fluid voxels. The outer surface of the oil storage and transportation system is divided into multiple boundary units; For each boundary cell, at each discrete time point within the preset time range, an external heat flux function is constructed by combining the environmental prediction data and the heat transfer properties of the boundary cell to obtain the predicted heat flux value of the boundary cell at that time point. The external influence matrix is ​​then formed by the predicted heat flux values ​​of all boundary cells at all discrete time points.

10. The system according to claim 8 or 9, characterized in that, Also includes: The calibration feedback module is used to collect the actual temperature distribution of the fluid space at a preset calibration time after the temperature control command is executed, and generate an actual thermal state grid. The actual thermal state grid is compared with the predictive thermal state grid generated for the calibration time to generate a deviation matrix; based on the deviation matrix, a calibration parameter field is generated to correct the thermal inertia field or the external influence matrix. The thermal field construction module or the state prediction module is also used to adjust the calibration parameter field in the subsequent temperature control command determination.

Images