Hydraulic and hydroelectric construction monitoring method and system based on digital twinning and storage medium

Abstract

The present application relates to the technical field of water conservancy and hydropower engineering construction monitoring, especially to a water conservancy and hydropower construction monitoring method and system based on digital twinning and a storage medium. Concrete structure design information and cooling system arrangement are obtained, a digital twinning temperature field model is constructed, a mapping relationship is established, and monitoring data are received; a monitoring data health state quantity is generated, temperature correction is performed, and a partition representative temperature is formed; the monitoring data health state quantity and the partition monitoring data quality index are used to adaptively adjust the control gain, the deviation of the partition representative temperature and the control target is mapped to a cooling water flow set value, and the model boundary is updated in combination with the execution feedback to generate a cooling water flow control instruction. The present application can solve or at least alleviate the technical problem that the monitoring data deviation and the deviation fluctuation do not participate in the control decision, resulting in flow regulation jitter, and provides a water conservancy and hydropower construction monitoring method and system based on digital twinning and a storage medium.

Description

Technical Field

[0001] This invention relates to the field of construction monitoring technology for water conservancy and hydropower projects, and in particular to a method, system, and storage medium for construction monitoring of water conservancy and hydropower projects based on digital twins. Background Technology

[0002] During the construction phase, large-volume concrete hydraulic structures commonly employ methods such as embedding cooling water pipes, installing internal and surface temperature sensors, and implementing compartmentalized water flow schemes for temperature monitoring and crack prevention. Existing temperature control systems often use fixed sampling periods to collect temperature data, relying on a small number of representative cross-sections or measuring points for manual or simple program-based trend analysis and flow rate adjustments. While such systems can meet basic temperature control requirements under steady-state conditions, they are prone to problems such as redundant monitoring information, excessively large impact ranges of single-point anomalies, and unstable cooling control responses when the number of sensors increases, cooling zones become more complex, and on-site operating conditions change frequently.

[0003] Industry typically improves temperature control system performance by enhancing sensor reliability, improving model accuracy, and setting alarm thresholds. For example, PT100 platinum resistance temperature sensors conforming to IEC 60751 are used to reduce measurement errors; finite element analysis or thermal-fluid coupling algorithms are employed to improve the accuracy of temperature field simulation; alarms are triggered by setting upper limits for temperature, temperature difference, and cooling rate; and field personnel adjust the cooling water flow rate based on experience. To reduce the impact of single-point anomalies on control decisions, some systems also introduce simple data smoothing methods, such as moving averages or median filtering. However, these methods still primarily rely on the absolute value of the measured temperature and its exceeding limits, lacking a systematic analysis of the deviation structure between the measured temperature and the model-predicted temperature, and also lacking a clear quantitative characterization of the relationship between monitoring data quality and control gain.

[0004] From the perspective of observable and calculable quantities, existing solutions have three fundamental problems. First, when the measured temperature is directly used as a control input or simply averaged and used as the representative temperature for a zone, it is impossible to distinguish between the actual temperature rise of concrete and sensor drift or communication interference, leading to a deviation between the measured temperature and the virtual temperature. and deviation fluctuation Firstly, because these factors are not explicitly utilized, single-point deviations may be amplified in the temperature range represented by each zone. Secondly, controllers often use fixed control gain or adjust control strength only based on temperature deviations, failing to incorporate zone monitoring data quality indicators. and execution reliability Including control gain calculations can lead to variations in the cooling water flow setpoint when monitoring data quality is low or execution deviations are large. The difference between adjacent cycles exhibits high-frequency jitter, causing frequent changes in flow boundary conditions. Thirdly, in some applications, digital twins or simulation models are only used for visualization or offline analysis, and have not formed a virtual temperature... Health status quantity as a benchmark Furthermore, a closed loop was not established between virtual temperature, measured temperature, and health status quantities in model calibration and control calculations, resulting in a lack of data quality bridging mechanism between the model and field monitoring.

[0005] To address the aforementioned problems, this invention proposes a temperature monitoring and adaptive control scheme centered on observable metrics such as deviation, deviation fluctuation, zonal monitoring data quality indicators, and execution reliability indicators. This invention obtains the deviation between the measured temperature and the virtual temperature by constructing a digital twin temperature field model. and deviation fluctuation And define the health status quantity of the monitoring data. ,use The measured temperature is weighted and corrected to form the corrected temperature. Then, the representative temperature of each zone is generated by weighting the cooling zones. Meanwhile, this invention defines regional monitoring data quality indicators. and execution reliability indicators This enables adaptive control gain. As the quality of monitoring data and execution reliability change, the system maintains a normal control response when the monitoring data quality is high and automatically reduces the control intensity when the monitoring data quality or execution reliability is low. Furthermore, by incorporating execution feedback into the boundary conditions of the digital twin model, this invention ensures that the virtual temperature field state remains consistent with the field execution results, forming a closed-loop link between monitoring, model, health status variables, control, and execution feedback. Summary of the Invention

[0006] To achieve the above-mentioned objectives, this invention provides a method, system, and storage medium for monitoring water conservancy and hydropower construction based on digital twins, aiming to solve or at least mitigate the problems in the prior art where monitoring data deviation and fluctuations are not involved in control decisions, abnormal measuring points amplify the results of temperature and flow regulation in different zones, and execution feedback is not effectively incorporated into the model and control loop.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for monitoring water conservancy and hydropower construction based on digital twins, applied to the temperature control construction of large-volume concrete in hydraulic structures, comprising:

[0008] A digital twin temperature field model is constructed based on concrete structure information, construction compartment data, and cooling system layout, and a mapping relationship is established between monitoring points, model calculation units, and cooling zones.

[0009] Acquire temperature monitoring data, cooling water operation data, and environmental boundary data, and drive the digital twin temperature field model to output the virtual temperature of the monitoring points;

[0010] The health status of the monitoring data is generated based on the deviation and fluctuation of the measured temperature and the virtual temperature; the measured temperature is corrected based on the health status of the monitoring data, and representative temperatures for each zone are formed.

[0011] The cooling water flow rate setpoint is generated based on the zone representative temperature, control target, and zone monitoring data quality indicators.

[0012] The set value is sent to the execution device, and the execution feedback is used as the input for subsequent model calculations and control decisions.

[0013] To further realize the present invention, the following technical solutions may be preferred:

[0014] Preferably, the step of constructing a digital twin temperature field model and establishing a mapping relationship specifically includes: establishing a set of concrete structure calculation units based on the geometric boundaries, construction joint locations, and pouring sequence of the hydraulic structure; configuring thermal property parameters and age-related heat source parameters for the calculation units; mapping the spatial location, flow direction, inlet and outlet water locations, and independent electromagnetic regulating valves of the parallel branch cooling pipes of each corresponding cooling zone to the corresponding calculation units; and writing the installation coordinates, burial depth, and compartment to which the temperature sensor belongs into the monitoring point index table, so that the monitoring points, model calculation units, and cooling zones form a callable spatial association relationship.

[0015] Preferably, the step of driving the digital twin temperature field model to output the virtual temperature of the monitoring point specifically includes: receiving the internal temperature of the concrete, the surface temperature of the concrete, the ambient temperature, the cooling water flow rate, the cooling water inlet temperature, and the cooling water outlet temperature under the same time reference; performing time stamp verification, missing data identification, and out-of-bounds identification on the received data; using the verified data as the model boundary input for the current calculation cycle; and extracting the virtual temperature of the corresponding monitoring point from the digital twin temperature field model according to the established mapping relationship.

[0016] Preferably, the step of generating the health status quantity of monitoring data specifically includes: pairing the measured temperature and virtual temperature of the same monitoring point under the same time reference to form paired data; generating a continuous deviation sequence based on the paired data; obtaining a deviation fluctuation index to characterize the stability of the monitoring point from the continuous deviation sequence; and inputting the deviation magnitude and deviation fluctuation index into a preset monitoring data health status quantity calculation rule so that when the deviation magnitude or deviation fluctuation increases, the monitoring data health status quantity of the corresponding monitoring point is reduced.

[0017] Preferably, the step of correcting the measured temperature and forming a representative temperature for the zone specifically includes: using the health status of the monitoring data as the fusion weight between the measured temperature and the virtual temperature to generate the corrected measured temperature of the corresponding monitoring point; calling the set of monitoring points within the same cooling zone according to the cooling zone and the established mapping relationship; and weighting and summarizing the corrected measured temperatures within the set of monitoring points according to the health status of the monitoring data to obtain the representative temperature of the cooling zone.

[0018] When the health status data of a single monitoring point decreases in isolation, reduce the impact of that monitoring point on the representative temperature of the zone.

[0019] When multiple spatially adjacent monitoring points within the same cooling zone simultaneously show an increase in the same direction of deviation, and the heating rate exceeds a preset safety threshold, a spatial consistency check is triggered. The calculation of the health status quantity of the monitoring data of the multiple spatially adjacent monitoring points is suspended, the fusion weight of the measured temperature of the multiple spatially adjacent monitoring points is maintained or increased, and a real high temperature warning is triggered.

[0020] Preferably, after the step of forming the representative temperature of the zone, the method further includes: inputting the calibrated measured temperature and the representative temperature of the zone as observation constraints into the digital twin temperature field model; correcting the boundary heat transfer state, cooling water action state, and volume heat source evolution state in the digital twin temperature field model according to the observation constraints; and ensuring that the temperature field output by the digital twin temperature field model and the calibrated observed temperature are kept within the preset model calibration tolerance while maintaining consistency in the construction stage, concrete age, and cooling system operation state.

[0021] Preferably, the step of generating the cooling water flow rate setpoint specifically includes: determining a temperature control error signal based on the relationship between the representative temperature of the zone and the control target; forming a zone monitoring data quality index based on the health status of monitoring data from each monitoring point within the same cooling zone; adjusting the basic control gain according to the zone monitoring data quality index to obtain an adaptive control gain corresponding to the monitoring data quality; generating a cooling water flow rate adjustment amount based on the temperature control error signal and the adaptive control gain, and obtaining the cooling water flow rate setpoint within the cooling system capacity constraints.

[0022] Preferably, the step of issuing the setpoint and processing the execution feedback specifically includes: converting the cooling water flow rate setpoint into a control command for the pump station speed, pump group operation status, or the opening of an independent electromagnetic regulating valve on the parallel branch of the corresponding cooling zone; issuing the control command to the field controller through an industrial communication interface; collecting the actual cooling water flow rate, cooling water inlet and outlet temperatures, and the operating status of the execution equipment to form execution feedback data; when the quality indicators of the zone monitoring data meet the anomaly judgment conditions, applying a safety constraint to the cooling water flow rate setpoint and generating a review prompt; and when the anomaly judgment conditions are lifted, using the execution feedback data for model boundary updates and control decisions in the next calculation cycle.

[0023] A water conservancy and hydropower construction monitoring system based on digital twins, used to execute the above-described method, characterized in that it includes:

[0024] The monitoring and acquisition module is used to acquire temperature monitoring data, cooling water operation data, and environmental boundary data;

[0025] The digital twin model module is used to construct a digital twin temperature field model and output the virtual temperature of the monitoring points;

[0026] The data health calculation module is used to generate monitoring data health status quantities based on the deviation and fluctuation of the measured temperature and the virtual temperature.

[0027] The temperature correction module is used to generate correction temperatures and zone representative temperatures based on the health status data monitored.

[0028] The adaptive control module is used to generate cooling water flow setpoints based on the zone representative temperature, control objectives, and zone monitoring data quality indicators.

[0029] The execution feedback module is used to issue control commands, collect execution feedback data, and perform safety constraint processing.

[0030] A storage medium storing a computer program, which, when executed by a processor, implements the above-described method, or causes a computing device to perform the following processes: constructing a digital twin temperature field model; establishing a mapping relationship between monitoring points, model calculation units, and cooling zones; acquiring temperature monitoring data, cooling water operation data, and environmental boundary data; outputting virtual temperatures of monitoring points; generating health status quantities of monitoring data; generating correction temperatures and representative temperatures for zones; generating cooling water flow setpoints; issuing control commands and using execution feedback as input for subsequent model calculations and control decisions.

[0031] The beneficial effects of this invention are:

[0032] This invention, by introducing and applying health status data, explicitly quantifies and utilizes the deviation and fluctuation between measured and virtual temperatures, transforming the quality of monitoring data from an implicit variable to a controllable variable. Compared to schemes based solely on measured temperature or simple averaging, this invention reduces the weight of abnormal measuring points on the zone representative temperature and cooling water flow rate setpoints when sensor drift, local disturbances, or communication anomalies exist, thereby reducing the magnitude of zone representative temperature deviation and flow rate regulation fluctuations.

[0033] This invention utilizes zone-based monitoring data quality indicators and execution reliability indicators in adaptive control gain calculation, achieving dual constraints on control gain, data quality, and execution capability. When monitoring data quality is high and execution reliability is good, the system maintains a basic control gain to ensure cooling response speed. When monitoring data quality or execution reliability declines, the system automatically reduces the control gain and applies safety constraints, ensuring that the cooling water flow rate setpoint varies within a controllable range, thereby reducing the risk of abrupt changes in temperature field boundary conditions due to monitoring anomalies or execution uncertainties. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the mapping between the digital twin temperature field and the monitoring point of the present invention.

[0035] Figure 2 This is a diagram of the system functional modules and data flow topology of the present invention.

[0036] Figure 3 This is a flowchart of the calculation process for the health status of monitoring data in this invention.

[0037] Figure 4 This is a schematic diagram of the generation of the correction temperature and the representative temperature of the zone in this invention. Detailed Implementation

[0038] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] Example 1

[0041] This embodiment provides a water conservancy and hydropower construction monitoring system based on digital twins. Combined with… Figure 1 and Figure 2 As shown, the system provided in this embodiment includes a field observation layer, a data access and preprocessing layer, a digital twin temperature field calculation layer, a monitoring data health status calculation layer, a temperature correction and zone representative temperature calculation layer, an adaptive temperature control decision layer, a control execution and operation feedback layer, and a storage medium.

[0042] The cooling system layout described in this invention can be implemented in various ways. In one exemplary embodiment (as shown in the attached diagram)... Figure 1 As shown in the figure, the cooling system layout adopts a cooling pipe structure including a main manifold and multiple parallel independent branches. However, those skilled in the art will understand that the cooling system layout is not limited to this, and can also be applied to series cooling pipes, serpentine open-loop cooling pipes, or other conventional large-volume concrete cooling water pipe networks. The digital twin mapping and adaptive control method described in this invention is also applicable.

[0043] The on-site observation layer includes internal concrete temperature sensors, concrete surface temperature sensors, ambient temperature sensors, cooling water flow meters, cooling water inlet thermometers, cooling water outlet thermometers, valve opening feedback units, and pump station operation status acquisition units. The internal concrete temperature sensor can be an industrial-grade PT100 platinum resistance temperature sensor; the "Pt" in PT100 represents platinum, and the number indicates that the resistance of the sensing element at 0 degrees Celsius is 100 ohms, and the temperature can be calculated from the resistance change. The platinum resistance temperature sensor can use a sensing element conforming to IEC 60751 and employs a stainless steel sheath, shielded cable, and sealed lead-out structure to adapt to concrete pouring, vibration, curing, and humid construction environments.

[0044] The data access and preprocessing layer is deployed on field acquisition terminals, industrial gateways, or edge computing nodes. This layer converts raw data such as temperature, resistance, flow rate, valve opening, and pump station status into unified engineering quantities, and writes the measurement point number, channel number, acquisition time, communication status, data source, and validity marker to each data record. For missing, out-of-bounds, duplicate, out-of-order, and abrupt data, the system does not directly delete the original record, but adds an anomaly marker to the data record, enabling subsequent health status quantity calculations to distinguish between actual temperature changes, sensor drift, and communication anomalies.

[0045] The digital twin temperature field calculation layer is used to construct, store, and update the digital twin temperature field model of the target hydraulic concrete structure. This digital twin temperature field model is a mathematical model of the temperature field established based on the concrete structure's geometric model, construction compartment model, pouring sequence, material thermal parameters, age-related hydration heat parameters, cooling pipe topology, environmental heat transfer boundaries, and monitoring point index relationships. The model uses concrete structure calculation units as the basic calculation objects, incorporating heat conduction between adjacent calculation units, concrete hydration heat release, surface heat transfer, and cooling water pipe heat transfer as temperature field calculation items. The model's boundary conditions are updated based on ambient temperature, actual cooling water flow rate, inlet water temperature, outlet water temperature, and construction stage information collected on-site. After the model calculation is completed, the virtual temperature at the corresponding monitoring point location is extracted from the monitoring point index table, serving as reference data for subsequent deviation calculations, health status quantity generation, temperature correction, and control decisions.

[0046] The monitoring data health status calculation layer receives the measured temperature and the virtual temperature output from the digital twin model. This layer uses the measured temperature and virtual temperature at the same monitoring point and time reference to form paired data, calculates the deviation, deviation fluctuation, and missing status between the two, and generates the monitoring data health status quantity for a single monitoring point. Combined with... Figure 3 As shown, the health status of the monitoring data is not an alarm level, nor is it a human experience score, but rather a calculation variable that is used in subsequent temperature fusion, model correction, and control gain adjustment.

[0047] The temperature correction and zone-representative temperature calculation layer uses the health status of monitored data as weights, merges the measured temperature and virtual temperature into a corrected temperature, and aggregates them according to the cooling zones to form the zone-representative temperature. Combined with... Figure 4 As shown, if the health status of a certain measuring point is high, the calibration temperature retains more of the measured temperature; if the health status of a certain measuring point decreases, the calibration temperature reverts to the virtual temperature of the model. The zone representative temperature is formed by weighting multiple calibration temperatures within the zone, ensuring that abnormal measuring points do not dominate cooling control with equal weight.

[0048] The adaptive temperature control decision layer receives the representative temperature of each zone, the control target, the zone monitoring data quality index, and the execution reliability index. This layer determines the direction of cooling water flow adjustment based on the deviation between the zone's representative temperature and the control target, and adjusts the control gain based on the zone monitoring data quality index. The execution reliability index is determined by the difference between the set flow rate and the actual flow rate, valve opening feedback, pump station operating status, and the cooling water inlet and outlet temperature response; its function is to prevent actuator malfunctions from being mistakenly attributed to temperature sensor malfunctions.

[0049] The control execution and operation feedback layer converts the cooling water flow setpoint into control commands for pump station speed, pump unit operating status, or the opening of independent solenoid regulating valves on the corresponding parallel branches of the cooling zone, which are then executed by the field controller. After execution, the system collects actual cooling water flow, cooling water inlet temperature, cooling water outlet temperature, valve feedback, and pump station status, and writes these data into the model boundary conditions and control constraints for the next calculation cycle. This design allows the digital twin model to use actual field execution results, rather than just theoretical setpoints.

[0050] The storage medium is used to store computer programs, status index tables, historical monitoring data, model parameters, control command records, and anomaly verification records. When the processor executes the computer program, it completes model building, data acquisition, data preprocessing, virtual temperature extraction, health status quantity generation, temperature correction, partition representative temperature generation, adaptive cooling control, safety constraint control, and execution feedback updates.

[0051] Example 2

[0052] Based on the monitoring system provided in Embodiment 1, this embodiment provides a digital twin-based method for monitoring water conservancy and hydropower construction. This method constructs a digital twin temperature field model based on concrete structure information, construction compartment data, and cooling system layout, and establishes a mapping relationship between monitoring points, model calculation units, and cooling zones. As an example, the cooling system layout can adopt a main manifold and multiple parallel independent branch cooling pipe structures.

[0053] In this embodiment, the digital twin temperature field model is obtained in the following ways: First, the design geometric boundary, construction joint location, compartment boundary, pouring sequence, and pouring time of the concrete structure of the hydraulic structure are acquired to establish a structural geometric model corresponding to the construction stage; Second, the concrete structure is divided into calculation units according to the structural geometric model, and spatial coordinates, unit volume, compartment, pouring time, concrete age, thermal conductivity, specific heat, density, and age-related hydration heat parameters are written for each calculation unit; Third, a cooling pipe topology table is established based on the cooling water pipe design data and on-site installation data. The cooling pipe topology table records the spatial path, inlet end, outlet end, flow direction, cooling zone, corresponding valve, corresponding pump station, actual flow rate, inlet temperature, and outlet temperature of the cooling pipe; Then, environmental boundaries and construction state boundaries are established based on ambient temperature, surface insulation status, curing status, and construction stage information; Finally, the installation coordinates, burial depth, compartment, and communication channel of the monitoring points are written into the monitoring point index table, so that a callable spatial mapping relationship is formed between the monitoring points, calculation units, and cooling zones. The resulting digital twin temperature field model can calculate the temperature of each computing unit based on the field boundary data in each calculation cycle, and output the virtual temperature of the corresponding monitoring point according to the monitoring point index table. .

[0054] In hydraulic structures, large-volume concrete is typically poured in stages, by compartment, layer, or block. Cooling water pipes are laid out along spatial paths, and temperature sensors are distributed within the structure, on its surface, and in the environment. If the monitoring point log, cooling pipe log, and model calculation unit are maintained separately, mismatches can easily occur between monitoring points and cooling zones after construction changes, additional monitoring points, and adjustments to cooling pipes. These mismatches can lead to the system adjusting flow rates for incorrect cooling zones, resulting in insufficient cooling of truly high-temperature areas or unnecessary cooling of adjacent areas.

[0055] like Figure 1 As shown, the system first establishes a set of concrete structure calculation units based on structural design data, construction joint locations, compartment boundaries, and pouring sequence. Each calculation unit records spatial coordinates, unit volume, pouring time, concrete age, thermal conductivity, specific heat, density, and volumetric heat source status. To achieve independent flow regulation for each cooling zone, the on-site cooling water system adopts a multi-parallel branch pipe structure with the main inlet pipe branching out. The cooling pipes are represented by a topology table, which records the pipe segment number, start and end positions, inlet end, outlet end, independent electromagnetic regulating valve at the inlet of the corresponding parallel branch, corresponding pump station, and the cooling zone to which it belongs. The monitoring point index table records the sensor number, installation coordinates, burial depth, compartment to which it belongs, adjacent calculation unit sets, cooling zone to which it belongs, communication channel, and valid version.

[0056] The above mapping relationships are managed uniformly using a status index table. Each row of the status index table corresponds to a monitoring point, and each column corresponds to a callable field, including the model calculation unit number, spatial interpolation weight, cooling pipe number, valve number, pump station number, and data quality status. If a monitoring point is located at the boundary of multiple calculation units, the system does not forcibly assign it to a single calculation unit, but instead saves the set of neighboring calculation units and their spatial weights. If the location of the cooling pipes or the boundary of the compartments is adjusted during construction, the system generates a new index version. Historical data is still interpreted according to the version corresponding to the time of collection, and current control is executed according to the latest approved version.

[0057] In the aforementioned digital twin temperature field model, the temperature propagation process within concrete is based on heat conduction calculations. The temperature gradient corresponds to the temperature difference between adjacent computational units, the thermal conductivity corresponds to the material thermal properties of the computational units, the heat flux corresponds to the heat exchange terms between adjacent computational units, the volumetric heat source corresponds to the hydration heat release terms related to the age of concrete, and the cooling water effect corresponds to the heat transfer boundary terms between the cooling pipes and adjacent computational units. Therefore, the digital twin temperature field model is not merely a three-dimensional graphic for display, but a computational model capable of generating a virtual temperature field based on on-site boundary data and construction status data.

[0058] The system acquires temperature monitoring data, cooling water operation data, and environmental boundary data, and drives the digital twin temperature field model to output the virtual temperature of the monitoring points.

[0059] Virtual temperature can only be used as a reference for measured temperature when the boundary conditions are close to the actual field conditions. If the model uses the design flow rate instead of the actual flow rate, or uses a fixed ambient temperature instead of the construction site ambient temperature, the deviation between the measured temperature and the virtual temperature will be mixed into the model boundary error, and the physical meaning of the health state quantity will be weakened.

[0060] The system receives data on concrete internal temperature, concrete surface temperature, ambient temperature, cooling water flow rate, cooling water inlet temperature, cooling water outlet temperature, valve opening feedback, and pump station operating status, all based on a unified time reference. After entering the backend, the data undergoes format parsing, followed by time stamp verification, channel number verification, unit unification, missing data identification, boundary violation identification, and abrupt change marking. For data with short-term communication interruptions, the system records the missing status; for data clearly exceeding the physically reasonable range, the system retains the original value and marks it as out of bounds; for out-of-order timestamp data, the system rewrites the corresponding period according to the actual acquisition time and recalculates the affected window statistics.

[0061] The sampling period is determined using the Nyquist sampling theorem as a constraint. The basic requirement of this theorem is that the sampling rate should be no less than twice the highest frequency component of the signal of interest; in engineering practice, a margin is usually allowed to reduce the risk of aliasing. Sampling rate The highest effective frequency is the frequency at which temperature and cooling water operating data are collected. The sampling period is estimated by combining concrete temperature changes, cooling water regulation response, and environmental disturbances. and satisfy The system can estimate based on historical temperature variation spectra and cooling system response records. and make the sampling rate meet .

[0062] The digital twin temperature field model receives three types of inputs in each calculation cycle: environmental boundary, cooling boundary, and construction state boundary. The environmental boundary includes ambient temperature, surface insulation status, and curing status; the cooling boundary includes actual cooling water flow rate, inlet water temperature, outlet water temperature, valve opening, and pipeline operating status; the construction state boundary includes the pouring section number, concrete age, and current construction stage. Based on these boundary inputs, the model updates the heat transfer conditions and heat source status of each calculation unit and calculates the temperature field for the current cycle. After completing the calculation, the model extracts the virtual temperature at the corresponding monitoring point location according to the state index table; for measuring points located within a single calculation unit, the temperature of that calculation unit is directly extracted; for measuring points located at the intersection of multiple calculation units, the weighted value of the temperatures of adjacent calculation units is extracted according to spatial interpolation weights.

[0063] The monitoring data health status quantity is generated based on the deviation and fluctuation of the measured temperature and the virtual temperature.

[0064] Sensor drift, local contact changes, acquisition channel noise, and communication interruptions can all cause measured temperatures to deviate from the true temperature field. Traditional solutions that directly use measured temperatures for control will amplify measurement anomalies into flow regulation actions; while relying directly on model temperatures will weaken the ability of on-site monitoring to correct for changes in actual operating conditions. Therefore, this method introduces a health status variable from the monitoring data as a reliability bridging variable between measured and virtual temperatures.

[0065] Combination Figure 3 As shown, let the first... Each monitoring point at time The measured temperature is The virtual temperature output by the digital twin model is The deviation is:

[0066]

[0067] in, This represents the deviation of the measured temperature from the virtual temperature. A deviation at a single moment only reflects the current difference and cannot distinguish between persistent model bias and short-term measurement disturbances. To characterize the stability of the deviation, the system... (The sentence is incomplete and requires further context to be translated accurately.) Internal calculation deviation fluctuation:

[0068]

[0069] in, Indicates the first Deviation fluctuation index at each measuring point This represents the mean deviation within the window. This indicates the number of valid samples within the window. The window length is not set to a fixed constant but is determined by the rate of temperature change, data completeness, and the status of cooling control actions. When the cooling water flow rate has just been adjusted, the window is shortened to avoid the old state affecting the current judgment; when the temperature change is gradual and the data is complete, the window is appropriately lengthened to suppress random noise.

[0070] The health status of the monitored data can be calculated using the following formula:

[0071]

[0072] in, Indicates the first Health status data of each monitoring point This represents the deviation penalty coefficient. This represents the fluctuation penalty coefficient. Indicates the absence of a penalty coefficient. This indicates a missing or abnormal communication status. The larger the deviation, the greater the fluctuation in deviation, or the more obvious the missing status, the lower the health status value.

[0073] In the rules for calculating health status , and These are the deviation penalty coefficient, the deviation fluctuation penalty coefficient, and the missing penalty coefficient, respectively. Acting on the deviation term This is used to characterize the degree of reduction in health status quantities when the measured temperature deviates from the virtual temperature. Acting on the deviation fluctuation term This is used to characterize the degree to which the deviation reduces the health status quantity when the deviation is unstable within the time window; Acting on missing or communication anomaly states These coefficients are used to characterize the degree of reduction in health status when data loss, out-of-bounds errors, or persistent communication anomalies occur. The system can determine these coefficients during the offline calibration phase, or after system initialization, the corresponding coefficient combinations can be determined according to the monitoring point type, burial depth, cooling zone, or sensor accuracy level.

[0074] Specifically, the system first selects historical stable operating condition data as calibration samples. The historical stable operating condition data includes reference temperatures obtained through manual verification or consistency screening from multiple measurement points. Corresponding measured temperature Virtual temperature Missing status and deviation fluctuations Then, select candidate coefficient combinations within the preset parameter range. And calculate the health status quantity based on the combination of candidate coefficients. and correction temperature The system determines the coefficient combination with the objective of minimizing the error between the correction temperature and the reference temperature. The objective function is:

[0075]

[0076] in, This represents the coefficient calibration error. The system can find candidate coefficient combinations that minimize the coefficient calibration error within a preset parameter range using grid search, piecewise search, or gradient descent. When using gradient descent, the iterative formula is:

[0077]

[0078] in, , Indicates the number of iterations. This indicates the step size. After calibration, the system writes the corresponding coefficient combinations into the model parameter table; during online monitoring, the system updates the parameters in real time. , and The health status of each monitoring point is dynamically calculated based on the determined coefficient combinations.

[0079] The fault-tolerance strategies include three categories. First, during short-term communication interruptions, the system does not directly set the health status quantity to zero. Instead, it gradually decays the health status quantity based on the duration of the interruption and the state of communication recovery, avoiding drastic control changes triggered by single-cycle communication jitter. Second, when a single point of change occurs and neighboring measurement points are stable, the system prioritizes reducing the health status quantity of that single measurement point and reducing its impact on the representative temperature of the partition and model calibration, without immediately modifying the overall model parameters. Third, when multiple spatially adjacent measurement points simultaneously show an increase in the same direction of deviation, the system does not directly identify it as a sensor anomaly. Instead, it triggers spatial consistency verification and model boundary verification to determine whether the phenomenon originates from an anomaly in the actual heat of hydration, local cooling failure, changes in environmental boundaries, or changes in cooling boundaries.

[0080] Spatial consistency verification is used to distinguish between single-point anomalies and real high-temperature events. The system determines the set of adjacent measuring points based on the spatial distance between monitoring points within the same cooling zone, their respective compartments, relationships with adjacent computing units, and the influence range of cooling pipes. When at least a preset number of monitoring points in the adjacent measuring point set simultaneously exhibit positive deviations in the same direction within a continuous calculation period, and the heating rate of each measuring point exceeds a preset safety threshold, the system determines that this phenomenon is not due to isolated sensor drift and enters the real high-temperature verification state. In the real high-temperature verification state, the system suspends the calculation of the health status attenuation of the multiple spatially adjacent monitoring points, maintains or increases the fusion weight of the measured temperatures of the multiple spatially adjacent monitoring points in the correction temperature, and triggers real high-temperature warnings, cooling capacity verification, and manual verification prompts. Within the cooling system capacity constraints, the system increases the cooling control intensity of the corresponding cooling zone to avoid real high-temperature events being misjudged as monitoring anomalies. When the deviation of adjacent measuring points falls back, the heating rate is lower than the preset safety threshold, and the execution feedback is normal, the system releases the real high-temperature verification state and resumes the calculation of normal health status quantities.

[0081] The measured temperature is corrected based on the health status data from monitoring data, and representative temperatures for each zone are generated. Cooling control requires the overall thermal state of the zone, not unfiltered readings from individual measuring points. Arithmetic averaging would cause abnormal and stable measuring points to have the same impact, the highest temperature method would amplify single-point spikes, and a simple model temperature cannot fully absorb the actual changes in the field. Therefore, this method uses health status data for both single-point correction and zone aggregation.

[0082] like Figure 4 As shown, the single-point calibration temperature is:

[0083]

[0084] in, Indicates the first The calibration temperature of each monitoring point. If The temperature is relatively high, and the calibration temperature is close to the measured temperature; if The temperature is relatively low, and the correction temperature reverts to the virtual temperature. This continuous fusion method avoids data gaps caused by simply removing abnormal measurement points, and also avoids abnormal measured values ​​directly dominating the control.

[0085] Let the first The set of effective monitoring points for each cooling zone is as follows: The temperature represented by the zone is:

[0086]

[0087] in, Indicates the first Each cooling zone represents a specific temperature. If the health status of a measuring point decreases, its contribution to the temperature represented by that zone automatically decreases. If the denominator is too small, it indicates that the overall monitoring data quality for that zone is insufficient, and the system marks that zone as an object requiring safety constraint control.

[0088] The calibration temperature and the representative temperature of each zone are also used as observational constraints to reinject the digital twin temperature field model. The reinjection method is not simply replacing the model temperature with the measured temperature, but rather correcting the boundary heat transfer state, cooling water action state, and volumetric heat source evolution state while maintaining consistency in construction stage, age, and cooling operation status. If multiple high-healthy-state measurement points within the same zone continuously deviate from the model, the system increases the model correction weight; if only a single low-healthy-state measurement point deviates, the system reduces the impact of that point on the model correction.

[0089] The cooling water flow rate setpoint is generated based on the zone representative temperature, control targets, and zone monitoring data quality indicators. Adjusting the cooling water flow rate alters the internal heat transfer boundary of the concrete. If the control gain is fixed, and the input temperature includes the influence of abnormal measuring points, abnormal temperature readings will be amplified as fluctuations in the flow rate setpoint. Excessive flow rate adjustments may cause localized rapid cooling, while insufficient adjustments may fail to suppress internal temperature rise. Engineering data indicates that concrete temperature control requires controlling temperature changes and temperature gradients to prevent temperature cracks; measures include internally embedded cooling water pipes for water cooling, surface protection, and curing.

[0090] The quality indicators for zoned monitoring data are:

[0091]

[0092] in, Indicates the first The overall quality of monitoring data within each cooling zone. The control error signal can be defined as:

[0093]

[0094] in, Indicates the first The target temperature for each cooling zone is calculated. If target range control is used, a dead zone is set within the target range, and flow adjustment is not triggered within the dead zone to reduce frequent actions caused by small noise.

[0095] The adaptive control gain is:

[0096]

[0097] in, Indicates the base control gain. This represents the gain adjustment function. This indicates the performance reliability index. The performance reliability index is formed by the difference between the set flow rate and the actual flow rate, valve feedback, pump station status, and inlet and outlet water temperature response. When the monitoring data quality is high and the performance reliability is high, Approaching the baseline control gain; when monitoring data quality is low or execution reliability is low. It is suppressed.

[0098] The cooling water flow rate adjustment is as follows:

[0099]

[0100] The cooling water flow rate setting for the next cycle is:

[0101]

[0102] in, This means the results will be limited to the cooling system capacity, minimum stable flow rate, maximum allowable flow rate, valve resolution, and single-cycle variation range. If the quality indicators of the zone monitoring data meet the anomaly judgment conditions, the system enters a safety constraint control state, reduces gain, limits flow rate changes, and generates verification prompts for sensors, lines, acquisition channels, or cooling circuits.

[0103] The setpoint is sent to the execution device, and the execution feedback is used as input for subsequent model calculations and control decisions. The system converts the flow setpoint into valve opening, pump station speed, or pump unit operating status based on the status index table. After execution by the field controller, the actual flow rate, cooling water inlet and outlet temperatures, valve opening feedback, and pump station status are returned to the backend. If the setpoint matches the actual feedback, the model uses the actual flow rate as the cooling boundary in the next cycle; if the setpoint does not match the actual feedback, the system records an execution-side anomaly and lowers the execution reliability index, rather than directly lowering the temperature sensor health status.

[0104] When the monitored data health status decreases, the quality index of the zonal monitoring data also decreases, and the adaptive control gain is suppressed. If the actual flow rate fails to reach the set flow rate, the execution reliability index decreases, and the control gain in the next cycle is further limited. This feedback mechanism eliminates data gaps caused by settings being issued but not executed on-site. If a valve is stuck, the model will not misuse the theoretical set flow rate; if the pump station capacity is limited, the controller will no longer generate flow set values ​​that exceed the actual execution capacity in subsequent cycles.

[0105] Example 3 Verification Test

[0106] The verification employed a hardware-in-the-loop simulation approach. The monitoring equipment included an embedded PT100 platinum resistance temperature sensor, a concrete surface temperature sensor, an ambient temperature sensor, an electromagnetic flowmeter, cooling water inlet and outlet resistance thermometers, an industrial data acquisition terminal, a programmable logic controller (PLC), and an edge server running a digital twin computing program. Software monitoring tools included a data frame parser, a time-series database, a model calculation service, a control command recording service, and an anomaly log service.

[0107] To facilitate the explanation of the verification process, this embodiment uses a combination of constant coefficients obtained after calibration under historical stable operating conditions. , and The system performs calculations; during actual online operation, it can also adopt corresponding methods based on different monitoring points. , and .

[0108] The data preprocessing process is as follows: The acquisition terminal first converts the temperature and flow rates into engineering quantities. The background then parses the data frames and performs time alignment, unit unification, and measurement point number verification. Missing data is flagged to prevent interpolated values ​​from being used as actual acquired values. Out-of-bounds data retains its original record and is flagged as out-of-bounds. Short-term spike data is flagged using median filtering, but the original values ​​remain traceable. Cooling water flow rate, inlet and outlet water temperatures, and valve feedback are mapped to the same time axis as the true boundary conditions for the next cycle of the model.

[0109] The core algorithm formula is as follows. The deviation is:

[0110]

[0111] Deviation fluctuation is:

[0112]

[0113] The health status is measured as follows:

[0114]

[0115] The corrected temperature is:

[0116]

[0117] The zone represents the temperature:

[0118]

[0119] The quality indicators for zoned monitoring data are:

[0120]

[0121] The adaptive control gain and flow adjustment are:

[0122]

[0123]

[0124] The physical or logical meaning of the above variables is as follows: For actual measured temperature, For virtual temperature, For the measurement point deviation, For deviation fluctuations, For monitoring the health status of measurement points, To correct the temperature, The zones represent temperatures. For the monitoring of data quality indicators in different zones, To achieve reliability metrics, For adaptive control of gain, This is for adjusting the cooling water flow rate.

[0125] Select a cooling zone There are three monitoring points. , , The virtual temperature output by the digital twin model for the current period is: , , The actual measured temperature collected on-site was: , , .in The reading is too high, but the adjacent P1 and P3 readings are stable. The system initially determines that P2 is a single-point jump caused by sensor drift or electromagnetic interference.

[0126] Calculate the current deviation:

[0127]

[0128]

[0129]

[0130] Based on the deviation sequence of the most recent effective window, the deviation fluctuation is obtained: , , Currently, there is no communication gap at any of the three measurement points, therefore... The sample coefficients obtained from historical stable operating conditions are used. , , Calculate health status parameters:

[0131]

[0132]

[0133]

[0134] visible The deviation and fluctuation are significantly higher than and Therefore, its health status is automatically downgraded by the system; but The measurement point information is retained by the system and is not directly set to zero.

[0135] Calculate the correction temperature:

[0136]

[0137]

[0138]

[0139] If the average measured temperature is used directly, then:

[0140]

[0141] The representative temperature of each zone is obtained by weighting the values ​​of health status:

[0142]

[0143] This indicates that the temperature is below the direct average, proving... The abnormally high interference readings have been effectively downweighted and absorbed by the system.

[0144] Calculate the quality indicators of the monitoring data for each zone:

[0145]

[0146] Let the current control target temperature be... The control error signal is defined as:

[0147]

[0148] Set the base control gain Execution reliability indicators Substituting the gain adjustment function, we get:

[0149]

[0150] Adaptive control gain:

[0151] Output cooling water flow rate adjustment:

[0152] If direct averaging temperature and fixed gain are used, the control error is:

[0153]

[0154]

[0155] Comparative verification: If the existing technology of direct temperature averaging and fixed gain is used, the control error is... The corresponding flow adjustment will reach .

[0156] Both solutions aim to increase cooling, but when the "false high-temperature spike" is confirmed to be caused by single-point electromagnetic interference, the flow rate adjustment output by the method of this invention ( ) significantly smaller than traditional methods ( This demonstrates that the system effectively filters out disturbances from abnormal readings, avoiding overcooling and ineffective valve vibration caused by blindly following local false signals. Conversely, if the aforementioned spatial consistency check is triggered and confirmed to be a multi-point real high-temperature event, the system will increase the cooling control intensity within the cooling system's capacity constraints, thereby balancing the ability to resist interference from abnormal data with the safe response capability to real high-temperature events.

[0157] To verify the multi-cycle fault tolerance mechanism, the continuous deviation processing trajectory of measuring point P2 was recorded as follows:

[0158] In the first calculation cycle, the deviation was 0.48℃, the fluctuation trend was low, the health status remained high, and the measurement point participated in the fusion normally.

[0159] In the second calculation cycle, the deviation increased slightly to 0.52℃, with low fluctuations, and the system maintained normal control.

[0160] In the third calculation cycle, the deviation rose to 0.57℃, and the fluctuation reached a moderate level. The system automatically reduced the local weight of that point.

[0161] In the fourth calculation cycle, the deviation suddenly increased to 1.10℃, and the fluctuation increased significantly, triggering a data quality alarm in the system;

[0162] In the fifth calculation cycle, the deviation reached a high level of 1.35℃, the health status value dropped to a low threshold, and the system officially started safety constraint control.

[0163] The above multi-period observation data shows that the system will not immediately trigger a mandatory alarm due to small noise in a single period, nor will it allow continuously increasing deviations to enter the control with high weight for a long period. If the deviation subsequently falls back and the fluctuation decreases, the health status quantity gradually recovers according to the same rules, and the control gain rises as the quality index of the zone monitoring data recovers.

[0164] Finally, verify the execution feedback. If the system issues a traffic adjustment amount... However, if the actual flow rate feedback only reflects a partial change, and the valve feedback shows that the opening degree has not reached the target, then the system will reduce its execution reliability index in the next cycle. The digital twin model uses actual flow rate as the cooling boundary, rather than a theoretically set flow rate. If the valve opening changes normally but the flow rate does not change accordingly, the system flags the hydraulic loop as abnormal; if the flow rate changes normally but the outlet water temperature rise is abnormal, the system flags the heat exchange status as abnormal. These abnormalities are then processed according to reliability indicators, without directly reducing the health status of the temperature sensor, thus avoiding confusion in the diagnostic path.

[0165] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for monitoring water conservancy and hydropower construction based on digital twins, applied to the temperature control construction of large-volume concrete in hydraulic structures, characterized in that... include: A digital twin temperature field model is constructed based on concrete structure information, construction compartment data, and cooling system layout, and a mapping relationship is established between monitoring points, model calculation units, and cooling zones. Acquire temperature monitoring data, cooling water operation data, and environmental boundary data, and drive the digital twin temperature field model to output the virtual temperature of the monitoring points; The health status of the monitoring data is generated based on the deviation and fluctuation of the measured temperature and the virtual temperature; the measured temperature is corrected based on the health status of the monitoring data, and representative temperatures for each zone are formed. The cooling water flow rate setpoint is generated based on the zone representative temperature, control target, and zone monitoring data quality indicators. The set value is sent to the execution device, and the execution feedback is used as the input for subsequent model calculations and control decisions.

2. The method according to claim 1, characterized in that, The steps of constructing a digital twin temperature field model and establishing a mapping relationship specifically include: establishing a set of concrete structure calculation units based on the geometric boundaries, construction joint locations, and pouring sequence of the hydraulic structure; configuring thermal property parameters and age-related heat source parameters for the calculation units; mapping the spatial location, flow direction, inlet and outlet water locations, and independent electromagnetic regulating valves of the parallel branch cooling pipes of each corresponding cooling zone to the corresponding calculation units; and writing the installation coordinates, burial depth, and compartment to which the temperature sensor belongs into the monitoring point index table, so that the monitoring points, model calculation units, and cooling zones form a callable spatial relationship.

3. The method according to claim 1, characterized in that, The steps for driving the digital twin temperature field model to output the virtual temperature of the monitoring points specifically include: receiving the internal temperature of the concrete, the surface temperature of the concrete, the ambient temperature, the cooling water flow rate, the inlet temperature of the cooling water, and the outlet temperature of the cooling water under the same time reference; performing time stamp verification, missing data identification, and out-of-bounds identification on the received data; using the data that passes the verification as the model boundary input for the current calculation cycle; and extracting the virtual temperature of the corresponding monitoring point from the digital twin temperature field model according to the established mapping relationship.

4. The method according to claim 1, characterized in that, The steps for generating the health status quantity of monitoring data specifically include: pairing the measured temperature and virtual temperature of the same monitoring point under the same time reference; generating a continuous deviation sequence based on the paired data; obtaining a deviation fluctuation index to characterize the stability of the monitoring point from the continuous deviation sequence; and inputting the deviation magnitude and deviation fluctuation index into a preset monitoring data health status quantity calculation rule so that the monitoring data health status quantity of the corresponding monitoring point is reduced when the deviation magnitude or deviation fluctuation increases.

5. The method according to claim 1, characterized in that, The step of correcting the measured temperature and forming a representative temperature for the zone specifically includes: using the health status of the monitoring data as the fusion weight between the measured temperature and the virtual temperature to generate the corrected measured temperature of the corresponding monitoring point; calling the set of monitoring points in the same cooling zone according to the cooling zone and the established mapping relationship; and weighting and summarizing the corrected measured temperatures in the set of monitoring points according to the health status of the monitoring data to obtain the representative temperature of the cooling zone. When the health status data of a single monitoring point decreases in isolation, reduce the impact of that monitoring point on the representative temperature of the zone. When multiple spatially adjacent monitoring points within the same cooling zone simultaneously show an increase in the same direction of deviation, and the heating rate exceeds a preset safety threshold, a spatial consistency check is triggered. The calculation of the health status quantity of the monitoring data of the multiple spatially adjacent monitoring points is suspended, the fusion weight of the measured temperature of the multiple spatially adjacent monitoring points is maintained or increased, and a real high temperature warning is triggered.

6. The method according to claim 1, characterized in that, After the step of forming the representative temperature of the zone, the method further includes: inputting the calibrated measured temperature and the representative temperature of the zone as observation constraints into the digital twin temperature field model; correcting the boundary heat transfer state, cooling water action state and volume heat source evolution state in the digital twin temperature field model according to the observation constraints; and ensuring that the temperature field output by the digital twin temperature field model and the calibrated observed temperature are kept within the preset model calibration tolerance while maintaining consistency in the construction stage, concrete age and cooling system operation state.

7. The method according to claim 1, characterized in that, The step of generating the cooling water flow setpoint specifically includes: determining the temperature control error signal based on the relationship between the representative temperature of the zone and the control target; forming a zone monitoring data quality index based on the health status of monitoring data from each monitoring point within the same cooling zone; adjusting the basic control gain according to the zone monitoring data quality index to obtain an adaptive control gain corresponding to the monitoring data quality; generating a cooling water flow adjustment amount based on the temperature control error signal and the adaptive control gain, and obtaining the cooling water flow setpoint within the cooling system capacity constraints.

8. The method according to claim 1, characterized in that, The steps of issuing setpoints and processing execution feedback specifically include: converting the cooling water flow rate setpoint into pump station speed, pump unit operation status, or opening control commands for independent electromagnetic regulating valves on parallel branches of the corresponding cooling zone; issuing the control commands to the field controller via an industrial communication interface; collecting actual cooling water flow rate, cooling water inlet and outlet temperatures, and operating status of the executing equipment to form execution feedback data; when the quality indicators of the zone monitoring data meet the anomaly judgment conditions, applying safety constraints to the cooling water flow rate setpoint and generating a review prompt; and when the anomaly judgment conditions are resolved, using the execution feedback data for model boundary updates and control decisions in the next calculation cycle.

9. A water conservancy and hydropower construction monitoring system based on digital twins, used to execute the method according to any one of claims 1 to 8, characterized in that, include: The monitoring and acquisition module is used to acquire temperature monitoring data, cooling water operation data, and environmental boundary data; The digital twin model module is used to construct a digital twin temperature field model and output the virtual temperature of the monitoring points; The data health calculation module is used to generate monitoring data health status quantities based on the deviation and fluctuation of the measured temperature and the virtual temperature. The temperature correction module is used to generate correction temperatures and zone representative temperatures based on the health status data monitored. The adaptive control module is used to generate cooling water flow setpoints based on the zone representative temperature, control objectives, and zone monitoring data quality indicators. The execution feedback module is used to issue control commands, collect execution feedback data, and perform safety constraint processing.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method described in any one of claims 1 to 8, or causes the computing device to perform the following processes: constructing a digital twin temperature field model; establishing a mapping relationship between monitoring points, model calculation units, and cooling zones; acquiring temperature monitoring data, cooling water operation data, and environmental boundary data; outputting virtual temperatures of monitoring points; generating health status quantities of monitoring data; generating correction temperatures and representative temperatures of zones; generating cooling water flow rate setpoints; issuing control commands and using execution feedback as input for subsequent model calculations and control decisions.

Images