Polar energy cabin remote mapping and control method and system based on data twinning

Abstract

This invention provides a remote mapping and control method and system for a polar energy cabin based on data twins, relating to the fields of polar scientific research power supply and digital twin microgrid control technology. The method includes: acquiring the spatial coordinates and shortest spatial distance of each power generation device within the polar energy cabin, and constructing a three-dimensional virtual digital twin map; constructing a multi-device thermal coupling matrix based on the three-dimensional virtual digital twin map, and generating a spatial thermodynamic twin state change equation for the polar energy cabin by combining a single device heating and cooling model under the extreme cold boundary of the polar region; iteratively optimizing the spatial thermodynamic twin state change equation to output a globally optimal power allocation parameter matrix; distributing the globally optimal power allocation parameter matrix to the polar energy cabin for control mapping; and adaptively recalibrating and reversing the parameters in the digital twin model base based on the prediction residuals.

Description

Technical Field

[0001] This invention relates to the field of polar scientific research power supply and digital twin microgrid control technology, and in particular to a method and system for remote mapping and control of polar energy cabins based on data twins. Background Technology

[0002] Polar scientific expeditions explore Earth's extreme environments, and a stable and reliable power supply is fundamental to ensuring the operation of polar equipment and the safety of personnel. Due to the extremely harsh natural environment of polar regions, characterized by extreme cold, strong winds, and blizzards year-round, traditional power grid infrastructure cannot be extended to these areas. Therefore, highly integrated containerized microgrid energy modules have become the primary solution for power supply in polar regions. A typical polar energy module uses a standard shipping container as its carrier, internally integrating high-power diesel generators, high-capacity energy storage battery packs, power converters, and other equipment. This energy module needs to continuously provide a stable power output of hundreds of kilowatts for extended periods with minimal or no personnel on duty.

[0003] In the extreme polar environment, the internal operating conditions of containerized energy compartments face extremely stringent constraints and complex temperature field coupling effects. On the one hand, the polar external environment is characterized by extremely low temperatures year-round, which poses a severe threat of overcooling to the equipment inside the compartment during cold starts or low-load operation. This is especially true for temperature-sensitive energy storage battery systems, where the extremely low temperatures can cause a sharp drop in usable capacity, deterioration in charge and discharge performance, and even damage. On the other hand, to meet the demand for high-load power supply, high-power equipment such as diesel generators generate enormous amounts of heat during operation. Constrained by the confined space of the container, the heat radiation and convection of these high-heat-generating devices can cause strong thermal interference to the surrounding space. Therefore, polar energy compartments often experience a contradiction between localized overcooling and localized overheating, and the temperature fields of various devices are intertwined.

[0004] Currently, conventional energy management systems for polar microgrid energy modules primarily focus on the scheduling of electrical parameters. Their thermal management strategies often rely on individual temperature sensors and preset temperature thresholds within each device for independent heating and cooling. This passive response control mechanism has significant limitations, completely severing the interaction between adjacent devices. When faced with a situation where one device is too cold while adjacent devices are overheating, traditional systems typically activate high-energy-consuming internal electric heating devices for independent temperature rise, ignoring the possibility of temperature balancing through heat radiation from adjacent high-heating devices, resulting in energy waste. Furthermore, when multiple devices are operating under high load, independent temperature control logic cannot cope with the combined temperature rise effect caused by the superposition of heat between devices, easily leading to internal thermal imbalance and ultimately forcing the entire energy system to throttle or shut down.

[0005] In recent years, digital twin technology has begun to be introduced into the field of microgrid management, but existing solutions still have significant shortcomings in dealing with the complex thermal environment of the polar regions. Most existing digital twins are based on single-unit mechanism models under ideal conditions, lacking the ability to accurately map the topological distances between multiple devices in space and the complex thermal field coupling effects, and failing to incorporate the mutual compensation of thermal radiation between devices into the system considerations. Furthermore, existing methods typically only reproduce data from the current state, lacking the ability to dynamically predict and plan based on future power loads. Under the harsh and variable climate boundary conditions of the polar regions, the thermodynamic parameters of the devices will shift over time. The lack of adaptive evolution and static models that do not consider spatial thermal coupling makes it impossible to accurately predict temperature trends in advance under complex operating conditions, making it difficult to achieve global temperature balance and long-term stable operation of all power supply equipment within the module. Summary of the Invention

[0006] This invention provides a remote mapping and control method for a polar energy cabin based on data twins, characterized by the following steps: obtaining the spatial coordinates and shortest spatial distance of each power generation device in the polar energy cabin, and integrating the collected electrical operation attributes to construct a three-dimensional spatial virtual digital twin map; Based on a three-dimensional virtual digital twin graph, a multi-device thermal coupling matrix is ​​constructed. Combined with the heat generation and heat dissipation model of a single device under the polar and frigid boundary, a space thermodynamic twin state change equation for the polar energy cabin is generated. Input the predicted future polar electricity load curve and set the safe temperature constraint of the equipment. Use the space thermodynamic twin state change equation to perform iterative optimization and output the global optimal power allocation parameter matrix. The global optimal power allocation parameter matrix is ​​sent to the polar energy cabin for control mapping. The actual observed temperature sequence of the power generation equipment is obtained, the prediction residual between the actual and theoretically predicted temperatures is calculated, and the parameters in the digital twin model base are adaptively recalibrated and reverse-corrected based on the prediction residual.

[0007] The spatial coordinates and shortest spatial distances of each power generation device within the polar energy cabin are obtained, and the collected electrical operation attributes are integrated to construct a three-dimensional virtual digital twin map. Specifically, this includes: obtaining the spatial coordinates and shortest spatial distances of the power generation devices within the polar energy cabin. The center coordinates of each power generation device Physical length and physical width Construct a 3D bounding box in the digital twin space and calculate the first... The power generation equipment and the first The shortest spatial distance between power generation devices The calculation formula is as follows: ; in, and , The total number of power generation equipment; when constructing a three-dimensional virtual digital twin, connect the first... The twin node and the first The twin edges of each twin node are assigned spatial topological association weights. Its definition is: in, This represents the thermal attenuation coefficient of air in a polar sealed cabin.

[0008] A model for the heat generation and dissipation of individual devices under polar and frigid boundary conditions, specifically including: based on the current moment of data acquisition. Operating voltage and operating current Calculate the first The self-heating power of each power generation device : in, Represents the electrothermal conversion efficiency constant; defines the first... The heat dissipation power of each power generation device to the polar external environmental boundary is : in, Indicates the current time Surface temperature, Indicates the external temperature of the polar regions. Indicates the equivalent heat dissipation surface area; This represents the combined coefficient of thermal infiltration and heat dissipation. Indicates the thickness of the container's insulation layer. This indicates the thermal conductivity of polyurethane insulation materials. This represents the baseline value for the natural convection heat transfer coefficient. This represents the correction factor for the external forced convection heat transfer wind speed. This indicates the wind speed in the polar environment.

[0009] Based on a 3D virtual digital twin graph, a multi-device thermal coupling matrix is ​​constructed. Combined with a single device heating and cooling model under extreme polar conditions, a thermodynamic twin state change equation for the polar energy cabin is generated. Specifically, this includes: a multi-device thermal coupling matrix. elements in The definition of is: in, Represents the thermal radiation absorptivity constant; generates discrete-time state change equations suitable for iterative calculations in digital twin systems, and obtains the next time step. Temperature forecast : ; in, Indicates the first The equivalent heat capacity constant of a power generation device This represents the discrete time step.

[0010] Input the predicted future polar electricity load curve and set the safe temperature constraints for the equipment, specifically including: discrete time nodes within the prediction time domain. To extrapolate and predict the virtual temperature change trajectory of each power generation device in the future time domain. Calculate the polar bidirectional thermal penalty function : in, This indicates the penalty coefficient for overheating violations. This indicates the penalty coefficient for violations related to polar supercooling. and These represent the upper limit of the superheat temperature and the lower limit of the supercool temperature, respectively; the calculation covers the entire prediction time domain. fitness evaluation function : in, This represents the controlled instruction allocation vector. Weighting factors representing polar fuel consumption This indicates the target output power assigned to the diesel generator.

[0011] Iterative optimization is performed using the space thermodynamic twin state change equation, specifically including: in the... In the next iteration, let the current controlled power allocation vector of the polar microgrid be... Constructing the Jacobian heat sensitivity matrix of the polar energy cabin The elements in this matrix Indicates the first Input control power of each device Tiny perturbations on the first Sensitivity gradient of the influence of the next moment's predicted temperature on the device: Combined with fitness evaluation function Calculate the global gradient descent vector of the current assignment scheme with respect to the fitness function. : ; in, Let be a constant unit vector characterizing the fuel consumption gradient.

[0012] Iterative optimization also includes: based on the global gradient descent vector Perform projection optimization iterations to obtain the updated assignment vector. The iterative formula is: in, For dynamic iteration step size, This is an unconstrained, temporary amendment. It is an orthogonal projection mapping matrix. This is the projection bias constant vector.

[0013] Obtain the actual observed temperature series of the power generation equipment, calculate the prediction residual between the actual observed temperature and the theoretically predicted temperature, specifically including: extracting the... Actual observed temperature sequence of a power generation device within the predicted time domain Calculate the cumulative prediction residual of polar thermodynamics : Construct an inverse correction cost function with the objective of minimizing the mean square error between the measured actual temperature and the re-predicted temperature. : in, Let be the vector of electrothermal conversion efficiency constants for all devices. Indicates that the parameter and The re-predicted temperature is obtained by substituting it back into the generated state change equation.

[0014] Adaptive recalibration and reverse correction of parameters in the digital twin model base based on prediction residuals, specifically including: in the first... In each learning evolution iteration, the current electrothermal conversion efficiency constant vector is fixed. Partial derivative correction for the thermal attenuation coefficient of air in a polar sealed cabin: Fixed correction polar sealed cabin air thermal attenuation coefficient Partial derivative corrections are made to the electrothermal conversion efficiency constant: in, This represents the learning rate of polar environment parameters. This represents the learning rate of the equipment aging parameters; the cost function is iterated through alternating evolution until it is corrected in reverse. Convergence is achieved by overwriting the updated values ​​of the polar sealed cabin air heat attenuation coefficient and the electrothermal conversion efficiency constant vector into the digital twin model base.

[0015] This invention also provides a remote mapping and control system for polar energy cabins based on data twins, characterized in that the system comprises: Digital twin construction module: Obtain the spatial coordinates and shortest spatial distance of each power generation device in the polar energy cabin, and integrate the collected electrical operation attributes to construct a three-dimensional spatial virtual digital twin map; Polar Energy Module Space Thermodynamic Twin Change Module: Based on a three-dimensional virtual digital twin map, a multi-device thermal coupling matrix is ​​constructed. Combined with the heat generation and heat dissipation model of individual devices under the polar extreme cold boundary, the space thermodynamic twin state change equation of the polar energy module is generated. Optimization module: Input the predicted future polar electricity load curve and set the safe temperature constraints of the equipment, use the space thermodynamic twin state change equation to perform iterative optimization, and output the global optimal power allocation parameter matrix; Mapping and Correction Module: Sends the global optimal power allocation parameter matrix to the polar energy cabin for control mapping; The actual observed temperature sequence of the power generation equipment is obtained, the prediction residual between the actual and theoretically predicted temperatures is calculated, and the parameters in the digital twin model base are adaptively recalibrated and reverse-corrected based on the prediction residual.

[0016] This invention proposes a remote mapping and control method and system for polar energy modules based on data twins. Due to the highly integrated standard container structure of polar energy modules, the high-power power generation equipment inside is arranged extremely densely, and they are constantly exposed to extremely low temperatures and complex weather conditions. The module often faces a severe contradiction between localized overcooling caused by low loads and localized overheating caused by high-power discharges. Existing energy dispatch and thermal management systems typically rely on independent temperature control feedback from individual devices, which cannot cope with the complex heat accumulation and spatial interference between devices. This application breaks through this limitation by introducing machine vision and multimodal data fusion technology to construct a digital twin model that includes three-dimensional spatial positions and surface topological relationships. This allows the system to move beyond abstract electrical connection diagrams and accurately reconstruct the actual spatial distances and heat radiation conduction paths between high-heat-generating devices within the confined, sealed module, providing a reliable three-dimensional virtual mapping foundation for solving complex thermal coupling problems in polar regions.

[0017] This invention deeply integrates the extreme cold boundary conditions of polar regions with the laws of thermodynamics within the storage chamber, proposing an adaptive directional optimization control mechanism based on spatial thermal coupling gradients. Traditional control methods typically activate high-energy-consuming independent electric heaters directly when equipment is detected to be overcooled. However, this method can pre-determine the evolution trend of the internal temperature field under future electrical loads within a virtual twin space. When it is predicted that a certain energy storage device will face the threat of extreme cold and overcooling, the system does not engage in blind trial and error, but instead uses a constructed multi-device thermal coupling matrix to reverse-engineer the optimal radiative heat source adjustment strategy. This mechanism can accurately increase the output power command of nearby high-heat-generating devices while maintaining global power balance, utilizing the excess waste heat generated by these devices to provide directional thermal radiation compensation to the overcooled devices across space. This strategy not only completely avoids the energy waste caused by activating additional electric heaters, but also effectively prevents the risk of overheating and shutdown due to localized overload, achieving high timeliness and high reliability of global temperature dynamic balance under limited polar computing resources.

[0018] Furthermore, the harsh polar climate causes a decline in the thermal insulation performance of the cabin, and equipment ages during long-term service in extreme cold, leading to deviations in system thermal conditions from their factory specifications. This invention establishes a closed-loop system for adaptive recalibration of twin evolution parameters based on measured infrared thermal field data. After issuing control commands, the system automatically corrects core parameters such as the air thermal attenuation coefficient and equipment heating efficiency in the twin model's base by continuously comparing the residuals between the virtual twin pre-simulation temperature and the measured polar temperature. This closed-loop correction endows the digital twin model with the self-learning ability to evolve synchronously with changes in the polar environment and equipment aging. This enables the invention to maintain extremely high state simulation accuracy and control command effectiveness throughout the entire lifecycle of the unmanned polar energy cabin, fundamentally improving the long-term stable operation and extreme cold climate adaptability of the polar scientific research power supply system. Attached Figure Description

[0019] Figure 1 This is a flowchart of the remote mapping and control of the polar energy cabin based on data twins according to the present invention; Figure 2 This is a schematic diagram of the overall layout of the polar energy module of the present invention; Figure 3 for Figure 2 Distance distribution diagram between various devices in the central energy compartment; Figure 4 This is a contour map showing the steady-state thermodynamic temperature field distribution inside the cabin under an alternative layout. Figure 5 This is a comparison diagram of the temperature evolution of thermal coupling compensation for multiple devices under polar conditions. Detailed Implementation

[0020] This embodiment provides a hardware system and architecture for a polar energy module, serving as the system foundation for realizing a remote mapping and control method for polar energy modules based on operational data twins.

[0021] This system is applied to the extremely cold and complex climate environment of the polar inland region, where the external ambient temperature is as low as -30 degrees Celsius or below. The polar energy cabin is entirely housed inside a standard 20-foot shipping container, with the preferred external dimensions of the container being 6.058 meters in length, 2.438 meters in width, and 2.591 meters in height. The container is fixedly mounted on a polar sled to accommodate movement and deployment in icy terrain. Due to the limited internal space of the container, the high-power heating equipment inside exhibits a high-density, compact spatial layout, with the equipment positioned very close together, resulting in significant spatial thermal radiation and convection coupling characteristics.

[0022] To construct a digital twin model with three-dimensional spatial topological relationships and thermodynamic coupling effects, a dual-spectrum visual acquisition device is fixedly installed at the top center or unobstructed edge area inside the polar energy cabin. Preferably, the dual-spectrum visual acquisition device includes a wide-angle visible light camera and an infrared thermal imaging camera. The wide-angle visible light camera is arranged at a top-down angle, and its field of view covers all the main power generation equipment inside the cabin, used to acquire the external outline, spatial coordinates, and relative distances between each device; the infrared thermal imaging camera is arranged coaxially or side by side with the wide-angle visible light camera, used to non-contactly acquire real-time temperature field distribution data on the surface of each device inside the cabin.

[0023] The polar energy cabin integrates a diesel generator power supply system, a photovoltaic system, and an energy storage inverter system. Specifically, the diesel generator power supply system includes a 300kW rated diesel generator. The diesel generator is preferably a four-stroke, inline six-cylinder, direct-injection fuel supply type with a turbocharged intercooled structure and a brushless self-excited generator, equipped with an AVR automatic voltage regulator. The diesel generator directly outputs three-phase AC power to the polar hot water drilling rig and auxiliary power. The photovoltaic system includes a photovoltaic controller and a retractable photovoltaic panel array mounted on the top of the container. The retractable photovoltaic panel array uses a three-layer sliding rail stacking design and preferably consists of 15 monocrystalline single-glass photovoltaic panels. The upper photovoltaic panels are fixed to the top of the cabin, while the middle and lower layers are mounted on a fixed frame using guide rails. When the energy cabin is moving, the photovoltaic panel array is in a retracted state; after reaching the polar expedition deployment point and the cabin stops moving, the lower and middle photovoltaic panels are extended along the sliding rails on both sides of the cabin top. The photovoltaic system is connected to the DC side of the energy storage inverter system via a DC-DC photovoltaic controller with a rated power of 10kW.

[0024] The energy storage inverter system is integrated into a separate air-cooled energy storage cabinet. The air-cooled energy storage cabinet includes a 100kWh lithium iron phosphate battery pack, a 100kW inverter (PCS), a battery management system (BMS), and a variable frequency thermal management air conditioning unit. The variable frequency thermal management air conditioning unit has cooling, heating, and ambient humidity control functions. The BMS is used to collect real-time data on the internal operating temperature and charge / discharge data of the battery cells and battery pack. The air-cooled energy storage cabinet supports a wide DC voltage input range of 424V to 676V and outputs 380V AC power to meet the power needs of the polar field living cabin.

[0025] The polar energy cabin is also equipped with a power distribution and topology conversion system, serving as the underlying execution mechanism for digital twin control commands. This system includes a three-phase rectifier bridge, a three-phase to single-phase transformer, and multiple controlled circuit breakers. The three-phase rectifier bridge consists of multiple rectifier modules with a rated current of 1000A, connected to a diesel generator at the front end and to the DC side of the air-cooled energy storage cabinet at the rear end. It converts the 380V three-phase AC power output from the diesel generator into 513V DC power, providing a charging channel for the air-cooled energy storage cabinet. The three-phase rectifier bridge is equipped with a dedicated heat dissipation module. The three-phase to single-phase transformer is preferably an 80kVA transformer using high-temperature resistant insulating paper as insulation material. Its input is connected to 380V three-phase AC power, and its output provides 220V single-phase AC power for auxiliary equipment.

[0026] A first controlled circuit breaker is connected in series between the diesel generator and the front end of the three-phase rectifier bridge; a second controlled circuit breaker is connected in series between the diesel generator and the power supply link to external scientific research equipment. These controlled circuit breakers are controlled by remote commands from the digital twin system, and by cutting off or closing electrical connections, they divide the energy compartment into an off-grid independent operation mode where it is entirely powered by photovoltaics and energy storage, or a grid-connected collaborative operation mode where the diesel generator simultaneously supplies power to the load and charges the energy storage.

[0027] The polar energy module is equipped with a field IoT gateway and a wired data transmission communication system. The IoT gateway interacts with the diesel generator control board, photovoltaic controller, BMS of the air-cooled energy storage cabinet, and sensors of the power distribution system via the Modbus TCP communication protocol, collecting and uploading real-time data on voltage, current, output frequency, generator water temperature, generator oil pressure, and cumulative equipment operating time. The collected underlying operational data, along with spatial and thermal field image data acquired by dual-spectrum visual acquisition equipment, are transmitted in real-time to a remote server via a polar satellite communication link or a long-distance wired fiber optic link, serving as the basic data input source for the construction and iterative correction of the digital twin model. A 35mm² copper cable is preferably laid between the module and the external living quarters for long-distance power transmission.

[0028] This embodiment provides a method for remote mapping and control of polar energy modules based on data twins. This embodiment specifically describes the steps for constructing a digital twin model, namely, the initial mapping of the digital twin based on machine vision and operational status.

[0029] Traditional digital twin models typically rely solely on electrical connection diagrams for logical mapping, neglecting the spatial arrangement of equipment. In the specific scenario of a standard container in a polar energy module, the internal clearance is only about 2.591 meters, resulting in an extremely compact arrangement of high-power equipment with very close proximity between their outer surfaces. This spatial constraint makes thermal radiation and convection interference between equipment a core factor determining the system's safe operation. Therefore, step S1 constructs a digital twin model to overcome traditional limitations, establishing a virtual digital twin that includes three-dimensional spatial topological relationships through visual computing and data fusion.

[0030] Step S1, constructing the digital twin model, specifically includes the following sub-steps: Step S1.1: Calculation of 3D spatial coordinates based on monocular inverse perspective mapping: A wide-angle visible light camera in a dual-spectrum visual acquisition system installed at the center of the polar energy module's top was used to capture a global top-down image of the module's interior. Object detection algorithms were then employed to identify objects within the images. The first core power generation equipment, and the first A power generation device is defined as ,in .extract The pixel coordinates of the top center point in the image pixel coordinate system .

[0031] Due to the limited internal height of the container, the top camera is very close to the power generation equipment, resulting in severe edge perspective distortion in the top-view image. To accurately determine the location of the power generation equipment within the container, this application establishes a world coordinate system with one corner of the container's internal bottom surface as the origin. During the design and assembly of the polar energy module, data was acquired and recorded in advance for each power generation device. Its own fixed prior height .

[0032] Construct an inverse perspective mapping model that incorporates the constraints of the polar module, and convert the two-dimensional pixel coordinates Convert to ground projection coordinates in a 3D world coordinate system The mathematical transformation relationship is expressed by the following formula: in, Indicates the first The depth scaling factor for each power generation device; This indicates that the wide-angle visible light camera data was obtained using the Zhang Zhengyou calibration method. Intrinsic parameter matrix; Indicates the camera's position relative to the world coordinate system. Extrinsic rotation matrix; express The extrinsic translation vector; due to The top height has a constant constraint in the world coordinate system, that is, its The axis coordinates are fixed as Therefore, the above matrix equations can be solved in reverse to calculate the equipment. The true coordinates of the geometric center on the bottom surface of the container To eliminate visual distortion caused by confined spaces.

[0033] Step S1.2: Calculation of the shortest spatial distance based on the 3D bounding box: Obtain the center coordinates of each power generation device and height Then, combined with the length of the equipment as it was manufactured. and width In the digital twin space for each power generation device Construct a 3D bounding box with parallel axes.

[0034] Considering that the heat radiation between the diesel generator and the energy storage tank inside the polar energy chamber does not originate from the geometric center of the equipment, but rather exchanges heat directly through the surface of the heated metal outer shell, the simple central Euclidean distance cannot be used to calculate the relative relationship between the equipment. It is necessary to calculate the shortest spatial distance between the surfaces of the enclosing box. (Definition of the first...) The power generation equipment and the first The shortest three-dimensional spatial distance between the power generation devices is The calculation formula is as follows: in, and This distance parameter It directly determines the boundary conditions for thermal radiation attenuation between two devices in a closed polar space, providing an absolute basis for calculating the thermodynamic coupling effect in space.

[0035] Step S1.3: Run data fusion and construct a 3D spatial virtual digital twin: While acquiring the aforementioned spatial geometric input parameters, the first data is collected in real time via the IoT gateway inside the polar energy module. One power generation unit At the present moment Electrical operating attributes, including at least the operating voltage. and operating current .

[0036] The spatial attributes extracted by machine vision are packaged with the electrical operation attributes collected by the IoT gateway, and a three-dimensional spatial virtual digital twin is instantiated and constructed in the virtual space of a remote server. .definition ,in: Represents a set of nodes, where the first node in the set is... Twin nodes Includes equipment The full set of multimodal data has the following data structure definition: in, Represents a set of edges, connecting... and twin sides Set as spatial topological association weight In the enclosed environment of a container in polar regions, air acts as a heat transfer medium, and its thermal radiation and convection effects decrease exponentially with distance. Therefore, this weight not only represents electrical connections but also directly reflects the degree of thermal interaction. The definition of spatial topological association weight is: in, This represents the air thermal attenuation coefficient of a sealed polar chamber, calibrated using atmospheric pressure and relative humidity conditions at the polar site. Considering the low air pressure due to high altitudes in the polar interior, and the changes in air moisture content under temperature control within the sealed chamber, which alter the air's resistance to heat convection and its absorption rate of infrared thermal radiation, the coefficient... It needs to be obtained through dynamic calibration of atmospheric pressure and relative humidity inside the cabin at the polar site: in, This represents the coefficient of thermal decay under standard atmospheric pressure and absolutely dry conditions; This indicates the real-time atmospheric pressure at the polar location, collected by external weather stations or barometric sensors. Represents the standard atmospheric pressure constant; This indicates the relative humidity of the air inside the polar energy cabin, as collected by the cabin temperature and humidity sensor. This represents the pressure regulation coefficient, which characterizes the effect of air density on heat transfer efficiency. This represents the humidity regulation coefficient, which characterizes the absorption characteristics of water vapor in the infrared band of thermal radiation.

[0037] In the extreme polar climate, the polar energy module not only faces external frigid temperatures of -30 degrees Celsius or even lower, but also experiences intense thermal interference due to the high-density concentration of high-power equipment inside. The temperature change of a single device no longer depends solely on its own power loss, but is highly correlated with the polar background environment and the distance and power of surrounding heat-generating equipment. Therefore, step S2 utilizes the spatial relative relationships and electrical operation data obtained in step S1 to construct a multi-device thermal relationship matrix, generating a space thermodynamic twin model that matches the polar module.

[0038] Step S2, establishing a twin thermodynamic model that incorporates the spatial thermal coupling effect, specifically includes the following sub-steps: Step S2.1: Construct a single device heat generation and heat dissipation model: In a digital twin system, the first step is to establish a basic thermodynamic baseline for individual devices in a polar environment. For the first... One power generation unit The current time based on data collected by the IoT gateway Operating voltage and operating current Calculate its self-heating power : in, Indicates the first The electrothermal conversion efficiency constant of a power generation device (i.e., the proportion of electrical energy converted into effective output, with the remainder converted into heat energy dissipation).

[0039] Meanwhile, the extreme cold of the polar environment causes continuous cold air infiltration and heat dissipation into the internal equipment through the container's outer shell. (Definition of the first...) The heat dissipation power of each power generation device to the polar external environmental boundary is : in, This represents the output of the previous iteration of the data collected by an infrared thermal imaging camera or a twin model. At the current moment, the power generation equipment Surface temperature; This indicates the real-time temperature of the polar external environment, collected from external weather stations. This indicates the first dimension determined based on the dimensions obtained in step S1. The equivalent heat dissipation surface area of ​​each power generation device; This represents the cold air infiltration and heat dissipation coefficient at the boundary of the polar container. It is strongly correlated with the thermal resistance of the insulation layer of the polar energy cabin and the forced convection effect caused by external polar wind speeds: in, This indicates the thickness of the container's insulation layer. In this embodiment, it is preferably selected based on the polar energy module standard. ; The thermal conductivity of the polyurethane insulation material in the cabin is represented by a value that is preferably taken as [value missing]. ; This represents the reference value for the natural convection heat transfer coefficient, preferably... ; The external forced convection heat transfer wind speed correction coefficient is preferably... ; This represents the real-time polar environmental wind speed collected by external weather stations (unit: This application extracts the offline calibration empirical value table shown in Table 1 below. During real-time model iteration, the table can be directly accessed and interpolated based on the wind speed data collected from the underlying layers. Table 1: Empirical Calibration Table (Based on 150mm Polyurethane Chamber) Step S2.2: Construction of a multi-device thermal coupling matrix based on spatial topology association weights: Within the confined space of a polar environment, the heat emitted by adjacent high-power equipment such as diesel generators can significantly raise the temperature of surrounding energy storage devices, including batteries. To quantify this spatial interference, this application incorporates the heat output of surrounding equipment and their spatial distance from the device as interference variables, introducing them into the spatial topology association weights calculated in step S1. .

[0040] Definition of the first The power generation equipment receives power from surrounding areas. The total spatial thermal coupling interference power of the devices is : in, Indicates the surrounding area The self-heating power of each power generation device; Represents a multi-device thermal coupling matrix The elements in the matrix. for The off-diagonal matrix has all elements on its main diagonal as 0, and all elements on its off-diagonal as 0. The definition of is: in, The thermal radiation absorptivity constant between devices is given by the first device receiving thermal radiation. One power generation unit The surface emissivity of the outer shell material and the The power generation equipment to the first The geometric angle coefficient of a power generation device is also known as the spatial diagonal factor. Joint decision: This embodiment is based on the surface thermal radiation parameters of common materials (emissivity of cold-rolled steel sheet for energy storage cabinet outer shell). Emissivity of diesel generator metal casing The empirical model of the angle coefficient for parallel / vertical projection was used to pre-determine the thermal radiation absorptivity constant between the core equipment inside the cabin. The specific calibration values ​​are shown in Table 2 below.

[0041] Table 2: Empirical calibration matrix By constructing this multi-device thermal coupling matrix Digital twin models will reduce spatial distance The exponential decay effect and the real-time heating status of surrounding high-power equipment Integration.

[0042] Step S2.3: Generation of the thermodynamic twin state change equations for the polar energy capsule: Considering the self-heating of individual devices, the dissipation at the polar and frigid boundaries, and the spatial thermal coupling interference of multiple devices, based on the law of conservation of energy, the first... The temperature evolution differential equation of a power generation device: in, Indicates the first The equivalent heat capacity constant of each power generation device. The above differential equation is transformed into a discrete-time state change equation suitable for iterative calculations in a digital twin system, yielding the next time step. Temperature forecast : ; in, This represents the discrete time step of the twin system.

[0043] When a device (such as an energy storage battery) is in a low-load standby state, its Approaching zero, destructive polar cold heat dissipation items This would cause a sharp drop in temperature, posing a serious threat of supercooling; however, this twin equation introduces a multi-device thermal coupling compensation term. It recreates the actual process of thermal radiation compensation for adjacent high-heat-generating equipment (such as diesel generators).

[0044] After establishing the space thermodynamic twin state change equation through step S2 in the digital twin system, the system has the ability to predict future temperature fields. In polar scenarios, traditional microgrid energy dispatch algorithms only consider power balance and battery state of charge. However, in the extremely cold, enclosed chambers of the polar regions at -30 degrees Celsius, it is crucial not only to prevent localized overheating during high-power discharges but also to prevent equipment overcooling under low-load conditions. If traditional algorithms blindly activate independent electric heaters upon detecting overcooling, it will result in extremely serious energy waste in the polar regions.

[0045] Therefore, step S3 combines the predicted polar electricity demand and initiates an adaptive directional optimization based on spatial thermal coupling gradient in the twin space to find the optimal power allocation ratio between the diesel generator and the energy storage system. Its core logic is to use the waste heat generated by the diesel generator and other equipment to perform thermal radiation compensation for the undercooled equipment (such as energy storage batteries) under the premise of meeting the power supply requirements, thereby finding a set of dynamic power control parameters in space that both meet the load and ensure that the global equipment does not overheat or overcool.

[0046] Step S3, adaptive orientation optimization based on spatial thermally coupled gradients, specifically includes the following sub-steps: Step S3.1: Set the power balance and boundary constraints for multi-source power supply in polar regions: At the current moment of the twin system's operation Input the future time domain prediction of polar research stations Internal power load curve And the predicted output of photovoltaic systems ,in This represents the discrete time nodes in the prediction time domain. This indicates the total number of prediction steps.

[0047] Define the polar energy module at a specific time point. The controlled instruction allocation vector is ,in This indicates the target output power assigned to the diesel generator. This represents the target charging and discharging power supplied to the air-cooled energy storage cabinet (positive values ​​indicate discharging, negative values ​​indicate charging). According to Kirchhoff's law of conservation of power, the system must satisfy a strict power balance equation at any given time: In response to the coexistence of extreme cold and localized high heat in the polar region, for the first time in the cabin... One power generation unit ( Set a dual-boundary constraint for safe temperature, namely, an upper limit for overheating temperature. and the lower limit of supercooled temperature For example, air-cooled energy storage cabinets Preferred setting is This is to prevent the cell's internal resistance from surging or freezing due to extreme cold in the polar regions.

[0048] Step S3.2: Construct the polar heat penalty fitness function: To simultaneously consider polar fuel consumption and cabin thermodynamic balance, this application constructs a fitness evaluation function that includes a fuel consumption cost term and a polar dual-temperature penalty term. .

[0049] For candidate power allocation vector The corresponding device's future heat generation power is derived. Then, by substituting this into the space thermodynamic twin state change equation generated in step S2, the virtual temperature change trajectory of each power generation device in the future prediction time domain is calculated. .

[0050] Calculate the first Each power generation device at the time node Polar bidirectional thermal penalty function : in, This indicates the penalty coefficient for overheating violations. This represents the penalty coefficient for violations related to polar supercooling. When the predicted temperature... Within a safe range Internally, the penalty function is 0; once overheating or overcooling occurs, a quadratic exponential penalty will be applied.

[0051] Combining the equivalent fuel loss generated by the operation of the diesel generator, a result covering the entire prediction time domain is obtained. fitness evaluation function : in, The weighting factors represent polar fuel consumption. The goal of optimization is to find a set of sequences. This makes the fitness evaluation function It reaches a minimum value.

[0052] Step S3.3: The allocation parameters obtained through dynamic optimization will serve as the optimal control. Traditional heuristic population optimization algorithms (such as genetic algorithms) often rely on randomly generated candidate populations for blind trial-and-error evolution when dealing with penalty terms involving complex thermodynamic coupling. In real-world scenarios like polar energy cabins where computational resources are limited and dynamic operating conditions change rapidly, traditional trial-and-error methods suffer from long convergence times and are prone to getting trapped in local optima, making it impossible to achieve highly timely real-time control of microgrids.

[0053] The multi-device thermal coupling matrix constructed using step S2 In the twin space, an analytical mapping relationship is established between the control input (power allocation of each device) and the system state output (temperature of each device), and its partial derivatives are obtained. This allows us to directly determine how to precisely fine-tune the power of a certain device to most effectively compensate for or reduce the temperature of a remote device in the other space.

[0054] In the set prediction time domain Within, for any discrete time node The specific iterative optimization steps are as follows: (1) Constructing the analytical Jacobian heat sensitivity matrix: in the first... At the start of the next iteration, let the current controlled power allocation vector of the polar microgrid be: .

[0055] To achieve precise guidance of the direction of temperature field changes, this application constructs the Jacobian heat sensitivity matrix of the polar energy module by taking partial derivatives with respect to the control input variables based on the twin state change equation. The elements in this matrix Indicates the first Input control power of each device Tiny perturbations affect the first Predict the temperature of the device at the next moment. The influence sensitivity gradient. Combining the thermodynamic derivation in step S2, its partial derivative analytical expression is clearly characterized as: This analytical expression represents the control feedback mechanism within a confined polar space, when... At that time, the temperature sensitivity depends entirely on the device's own heating efficiency. and heat capacity ; and when At that time, the temperature sensitivity is determined by the spatial correlation elements in the multi-device thermal coupling matrix. Strictly controlled.

[0056] (2) The direction of the decreasing thermodynamic gradient: Using the polar bidirectional thermal penalty fitness function defined in step S3.2 Calculate the current allocation scheme The global gradient descent vector relative to the fitness function : ; in, This gradient vector is a constant unit direction characterizing the fuel consumption gradient. This indicates the optimal analytical direction for adjusting the power allocation command under the current thermal distribution conditions in order to quickly eliminate polar supercooling and superheating violations and reduce fuel consumption.

[0057] (3) Projection optimization iteration on the constrained power supply balance manifold surface: To ensure that adjusting the power according to the above gradient direction does not violate the strict equality constraint of polar power supply balance defined in step S3.1 (i.e.) In this embodiment, a hyperplane is defined in the multidimensional solution space, namely the constrained power supply balance manifold surface.

[0058] In updating the next generation power allocation vector At this time, the system first performs unconstrained forward exploration based on the global gradient descent vector, and then uses the projection matrix. The exploration results are then projected vertically back onto the aforementioned constrained power supply equilibrium manifold. The complete projection iteration formula is expressed as: in, For dynamic iteration step size; This is an unconstrained, temporary amendment. To ensure that the update volume is always parallel The orthogonal projection mapping matrix of the plane; To compensate for the current load The projection bias constant vector.

[0059] When the iteration loop causes the difference in fitness function between two adjacent iterations to be less than a preset accuracy threshold, or reaches the maximum allowable computation time limit of the polar gateway controller, the algorithm stops iterating and directly outputs the globally optimal power allocation parameter matrix that has now fully converged. .

[0060] This invention addresses the interplay between extreme polar cold and localized high-heat generation within the storage compartment directly at the partial derivative analytical level, without requiring random mutations. When the twin system predicts that polar supercooling will occur at a certain energy storage tank, the Jacobian sensitivity matrix immediately infers the need to increase the power output of nearby diesel generators. This provides the most effective far-end thermal radiation supplementation. The gradient vector precisely guides the diesel generator power to increase, while the projection matrix forces the energy storage cabinet to synchronously adjust its charging and discharging power to maintain a constant total power. The final output allocation parameters... This will serve as the best control basis and will be directly issued to the polar execution-level power distribution system.

[0061] In the extreme and harsh environment of the polar regions, where it is difficult for personnel to stay for extended periods, the significance of the energy module digital twin system lies not only in virtual prediction, but also in its execution and adaptive learning to resist aging and environmental degradation. Therefore, step S4 is responsible for mapping the predicted optimal allocation parameters, while step S5 uses feedback data from on-site measurements in the polar regions to self-correct the parameters of the twin base.

[0062] Step S4: Remote mapping and execution of the optimal control strategy: Through dynamic optimization in step S3, the server obtains the globally optimal power allocation parameter matrix in the digital twin space to avoid extreme cooling in polar regions and overheating from high-power discharge. .

[0063] The server encapsulates the optimal control parameters into a data packet, which is then transmitted to the on-site IoT gateway inside the polar energy cabin via a low-Earth orbit satellite communication link or a long-distance anti-freeze fiber optic link.

[0064] After receiving and verifying the command, the on-site IoT gateway maps and breaks it down into specific control signals for the underlying actuators: On one hand, the gateway transmits the command via the Modbus TCP protocol. This is converted into a voltage / frequency regulation signal and sent to the AVR automatic voltage regulator and electronic speed governor on the diesel generator control board. This controls the diesel generator to output the target power to meet grid load and stabilize waste heat radiation to surrounding subcooled equipment according to a predicted trajectory. On the other hand, the gateway will issue instructions... The signal is converted into a converter control signal and sent to the inverter (PCS) in the air-cooled energy storage cabinet, forcing the energy storage system to operate according to the optimal charging and discharging power. At the same time, based on the power distribution topology state during the virtual calculation process, the gateway controls the first and second controlled circuit breakers in the polar energy compartment to perform corresponding cutting or closing actions, ensuring that the energy flow network in polar space completely and strictly replicates the optimal topology of the virtual space.

[0065] Step S5: Feedback and reverse correction of the polar twin model based on measured temperature rise data: During months of operation in polar regions, extreme cold and blizzards can cause the outer shell insulation layer to shrink or freeze, altering thermal resistance. Simultaneously, aging of the power generation equipment can cause the electrothermal conversion efficiency constant to drift. If the digital twin model remains unchanged, its thermal projection trajectory will deviate significantly from the actual situation. Therefore, step S5 establishes a reverse correction mechanism based on measured infrared thermal imaging data.

[0066] Step S5.1: Obtain the deviation matrix between the actual polar thermal field and the twin simulation: Control commands in step S4 Execution through a prediction time domain Then, using an infrared thermal imaging camera fixed to the top of the polar energy module, the actual surface temperature field distribution data of each power generation device inside the module was collected. The data was then extracted. Each power generation device at various historical discrete time points Actual observed temperature series .

[0067] Definition of the first The cumulative polar thermodynamic prediction residual of each power generation device over the entire prediction time domain for: in, This refers to the theoretically predicted temperature derived from the initial twin model in step S3. When any device... Greater than the preset polar model distortion tolerance threshold When this happens, the subsequent reverse correction algorithm is triggered.

[0068] Step S5.2: Adaptive recalibration of twin evolution parameters based on polar measured data: The core reason for the distortion of the polar twin model lies in two degradation parameters: one is the thermal attenuation coefficient of the air in the polar sealed cabin affected by polar ice and snow. Secondly, the electrothermal conversion efficiency constant is affected by equipment aging. .

[0069] To correct the aforementioned bias, this application constructs a reverse correction cost function with the objective of minimizing the mean square error between the measured temperature and the recalculated predicted temperature. : in, Let be the vector of electrothermal conversion efficiency constants for all devices; This indicates that the unknown modified parameters will be included. and The re-predicted temperature is obtained by substituting it back into the state change equation generated in step S2.

[0070] Due to the air thermal attenuation coefficient What's affected is the attenuation of spatial heat radiation between equipment rooms, while the heating efficiency... The effect is on the self-generated heat of the unit, and the two are decoupled. In this embodiment, an alternating gradient descent algorithm is used to perform self-learning evolution on these two parameters respectively.

[0071] In the In each learning evolution iteration, the current aging heat generation efficiency vector is first fixed. The partial derivative of the air thermal attenuation coefficient affected by polar ice and snow cover is corrected: Subsequently, the corrected thermal decay coefficient was fixed. The partial derivative of the heating efficiency constant, which represents equipment aging, is corrected: in, and These represent the learning rate of polar environment parameters and the learning rate of equipment aging parameters, respectively.

[0072] Step S5.3: Global update of the model base: The above alternating evolution iteration continues until the cost function is reached. Convergence. Updated value of the thermal attenuation coefficient of the air in the polar sealed cabin obtained after convergence. Update value of the vector of electrothermal conversion efficiency constant It will be directly overwritten into the digital twin model base.

[0073] Through step S5, the polar digital twin model of this application is endowed with self-learning and self-growing capabilities. It can capture, through infrared thermal imaging, the deterioration of internal thermal convection caused by icing on the outer wall of the container (i.e.,... (Smaller) and increased heat generation rate due to long-term extremely cold operation of diesel generators (i.e., (Lowering), and automatically correcting the base. This ensures that, throughout the polar energy module's years of service, regardless of how severe the external blizzard environment or how much the internal equipment deteriorates, the accuracy of its twin simulation remains true to the real world.

[0074] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0075] In this specification, the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the descriptions of the embodiments described later are relatively simple, and relevant parts can be referred to the descriptions of the foregoing embodiments.

[0076] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art 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 remote mapping and control method for polar energy modules based on data twins, characterized in that, Includes the following steps: The spatial coordinates and shortest spatial distance of each power generation device in the polar energy cabin are obtained, and the collected electrical operation attributes are integrated to construct a three-dimensional spatial virtual digital twin map. Based on a three-dimensional virtual digital twin graph, a multi-device thermal coupling matrix is ​​constructed. Combined with the heat generation and heat dissipation model of a single device under the polar and frigid boundary, a space thermodynamic twin state change equation for the polar energy cabin is generated. Input the predicted future polar electricity load curve and set the safe temperature constraint of the equipment. Use the space thermodynamic twin state change equation to perform iterative optimization and output the global optimal power allocation parameter matrix. The global optimal power allocation parameter matrix is ​​sent to the polar energy cabin for control mapping. The actual observed temperature sequence of the power generation equipment is obtained, the prediction residual between the actual and theoretically predicted temperatures is calculated, and the parameters in the digital twin model base are adaptively recalibrated and reverse-corrected based on the prediction residual.

2. The method according to claim 1, characterized in that, The spatial coordinates and shortest spatial distances of each power generation device within the polar energy cabin are obtained, and the collected electrical operation attributes are integrated to construct a three-dimensional virtual digital twin map. Specifically, this includes: obtaining the spatial coordinates and shortest spatial distances of the power generation devices within the polar energy cabin. The center coordinates of each power generation device Physical length and physical width Construct a 3D bounding box in the digital twin space and calculate the first... The power generation equipment and the first The shortest spatial distance between power generation devices The calculation formula is as follows: ； in, and , The total number of power generation equipment; when constructing a three-dimensional virtual digital twin, connect the first... The twin node and the first The twin edges of each twin node are assigned spatial topological association weights. Its definition is: ; in, This represents the thermal attenuation coefficient of air in a polar sealed cabin.

3. The method according to claim 2, characterized in that, A model for the heat generation and dissipation of individual devices in polar and frigid environments, specifically including: based on the current moment of data acquisition. Operating voltage and operating current Calculate the first The self-heating power of each power generation device : ; in, Represents the electrothermal conversion efficiency constant; defines the first... The heat dissipation power of each power generation device to the polar external environmental boundary is : ; ; in, Indicates the current time Surface temperature, Indicates the external temperature of the polar regions. Indicates the equivalent heat dissipation surface area; This represents the combined coefficient of thermal infiltration and heat dissipation. Indicates the thickness of the container's insulation layer. This indicates the thermal conductivity of polyurethane insulation materials. This represents the baseline value for the natural convection heat transfer coefficient. This represents the correction factor for the external forced convection heat transfer wind speed. This indicates the wind speed in the polar environment.

4. The method according to claim 3, characterized in that, Based on a 3D virtual digital twin graph, a multi-device thermal coupling matrix is ​​constructed. Combined with a single device heating and cooling model under extreme polar conditions, a thermodynamic twin state change equation for the polar energy cabin is generated. Specifically, this includes: a multi-device thermal coupling matrix. elements in The definition of is: ; in, Represents the thermal radiation absorptivity constant; generates discrete-time state change equations suitable for iterative calculations in digital twin systems, and obtains the next time step. Temperature forecast : ； in, Indicates the first The equivalent heat capacity constant of a power generation device This represents the discrete time step.

5. The method according to claim 4, characterized in that, Input the predicted future polar electricity load curve and set the safe temperature constraints for the equipment, specifically including: discrete time nodes within the prediction time domain. To extrapolate and predict the virtual temperature change trajectory of each power generation device in the future time domain. Calculate the polar bidirectional thermal penalty function : ; in, This indicates the penalty coefficient for overheating violations. This indicates the penalty coefficient for violations related to polar supercooling. and These represent the upper limit of the superheat temperature and the lower limit of the supercool temperature, respectively; the calculation covers the entire prediction time domain. fitness evaluation function : ; in, This represents the controlled instruction allocation vector. Weighting factors representing polar fuel consumption This indicates the target output power assigned to the diesel generator.

6. The method according to claim 5, characterized in that, Iterative optimization is performed using the space thermodynamic twin state change equation, specifically including: in the... In the next iteration, let the current controlled power allocation vector of the polar microgrid be... Constructing the Jacobian heat sensitivity matrix of the polar energy cabin The elements in this matrix Indicates the first Input control power of each device Tiny perturbations on the first Sensitivity gradient of the influence of the next moment's predicted temperature on the device: ; Combined with fitness evaluation function Calculate the global gradient descent vector of the current assignment scheme with respect to the fitness function. : ； in, Let be a constant unit vector characterizing the fuel consumption gradient.

7. The method according to claim 6, characterized in that, Iterative optimization also includes: based on the global gradient descent vector Perform projection optimization iterations to obtain the updated assignment vector. The iterative formula is: ; ; in, For dynamic iteration step size, This is an unconstrained, temporary amendment. It is an orthogonal projection mapping matrix. This is the projection bias constant vector.

8. The method according to claim 4, characterized in that, Obtain the actual observed temperature series of the power generation equipment, calculate the prediction residual between the actual observed temperature and the theoretically predicted temperature, specifically including: extracting the... Actual observed temperature sequence of a power generation device within the predicted time domain Calculate the cumulative prediction residual of polar thermodynamics : ; Construct an inverse correction cost function with the objective of minimizing the mean square error between the measured actual temperature and the re-predicted temperature. : ; in, Let be the vector of the electrothermal conversion efficiency constants of all devices. Indicates that the parameter and The re-predicted temperature is obtained by substituting it back into the generated state change equation.

9. The method according to claim 8, characterized in that, Adaptive recalibration and reverse correction of parameters in the digital twin model base based on prediction residuals, specifically including: in the first... In each learning evolution iteration, the current electrothermal conversion efficiency constant vector is fixed. Partial derivative correction for the thermal attenuation coefficient of air in a polar sealed cabin: ; Fixed correction polar sealed cabin air thermal attenuation coefficient Partial derivative corrections are made to the electrothermal conversion efficiency constant: ; in, This represents the learning rate of polar environment parameters. This represents the learning rate of the equipment aging parameters; the cost function is iterated through alternating evolution until it is corrected in reverse. Convergence is achieved by overwriting the updated values ​​of the polar sealed cabin air heat attenuation coefficient and the electrothermal conversion efficiency constant vector into the digital twin model base.

10. A remote mapping and control system for polar energy cabins based on data twins, characterized in that, The system includes: Digital twin construction module: Obtain the spatial coordinates and shortest spatial distance of each power generation device in the polar energy cabin, and integrate the collected electrical operation attributes to construct a three-dimensional spatial virtual digital twin map; Polar Energy Module Space Thermodynamic Twin Change Module: Based on a three-dimensional virtual digital twin map, a multi-device thermal coupling matrix is ​​constructed. Combined with the heat generation and heat dissipation model of individual devices under the polar extreme cold boundary, the space thermodynamic twin state change equation of the polar energy module is generated. Optimization module: Input the predicted future polar electricity load curve and set the safe temperature constraints of the equipment, use the space thermodynamic twin state change equation to perform iterative optimization, and output the global optimal power allocation parameter matrix; Mapping and Correction Module: Sends the global optimal power allocation parameter matrix to the polar energy cabin for control mapping; The actual observed temperature sequence of the power generation equipment is obtained, the prediction residual between the actual and theoretically predicted temperatures is calculated, and the parameters in the digital twin model base are adaptively recalibrated and reverse-corrected based on the prediction residual.

Images