Fabricated building energy consumption monitoring method and system based on digital twinning

Abstract

The invention discloses an assembled building energy consumption monitoring method and system based on digital twinning, and relates to the field of simulation optimization, and the method comprises the steps: building a digital twinning model through a finite element method, initializing key state parameters, and carrying out the high-precision synchronous simulation of a temperature field and energy consumption through a sensor real-time data driving model. Based on the difference between simulation data and actual measurement data, a multi-temporal-spatial-scale hierarchical collaborative inversion strategy is adopted, and a Levenberg-Marquardt algorithm is combined to carry out progressive inversion on key parameters, so that fault diagnosis, model self-checking and performance degradation tracking are realized. And when the parameter degradation exceeds the limit, the system issues graded early warning, predicts future energy consumption based on historical rules, simulates energy-saving effects of different maintenance schemes, and finally generates a quantitative decision support report, thereby realizing full-chain intelligent operation and maintenance from energy consumption monitoring and performance degradation diagnosis to predictive maintenance decision.

Description

Technical Field

[0001] This invention relates to the field of simulation optimization, and in particular to a method and system for monitoring the energy consumption of prefabricated buildings based on digital twins. Background Technology

[0002] In the field of energy-saving monitoring and operation and maintenance management of prefabricated buildings, high-precision thermal simulation and digital twin technology provide core decision-making support. However, most current technical solutions still suffer from the following key bottlenecks: Existing energy consumption monitoring methods mostly rely on simplified lumped parameter models or black-box statistical models, making it difficult to accurately characterize key local physical details such as the contact thermal resistance of prefabricated component joints and thermal bridges at nodes, resulting in insufficient accuracy in thermal simulation. More importantly, existing digital twin models generally lack the ability to dynamically correct parameters based on measured data. Traditional parameter inversion methods often treat the building as a homogeneous whole or only isolate and identify local parts, failing to reconcile the contradiction between global performance consistency and local spatial heterogeneity, and also failing to consider the multi-scale evolution characteristics of parameters over time, making it difficult for the model to accurately reflect the complex performance degradation caused by multiple sources such as material aging and construction deviations.

[0003] Furthermore, the existing energy consumption assessment system has significant shortcomings: on the one hand, it is mostly limited to the passive recording of meter data and fails to be deeply integrated with high-precision temperature field simulation, making it difficult to achieve dynamic benchmarking and attribution analysis of theoretical energy consumption and actual energy consumption while ensuring indoor thermal comfort; on the other hand, due to the lack of ability to continuously track key state parameters across multiple spatiotemporal scales, it is impossible to establish a quantitative correlation model between parameter degradation and performance decay, which means it cannot predict future performance based on the current state, nor can it quantitatively simulate the energy-saving effects of different maintenance schemes, severely restricting the implementation effect of refined energy consumption management and preventive maintenance decision-making throughout the entire life cycle of prefabricated buildings. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a digital twin-based method for monitoring the energy consumption of prefabricated buildings, which solves the problem that traditional methods cannot support refined energy consumption management and preventive maintenance throughout the entire life cycle due to model inaccuracies and lack of parameter adaptive capabilities.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for monitoring the energy consumption of prefabricated buildings based on digital twins, which includes constructing a digital twin model of the prefabricated building using the finite element method and initializing key state parameters.

[0008] Real-time data obtained through sensors is used as boundary conditions to drive the model to perform high-precision synchronous simulation and evolution of temperature field and energy consumption.

[0009] Based on the real-time data and simulation results, a multi-temporal-scale hierarchical collaborative inversion strategy is adopted. The Levenberg-Marquardt algorithm is used to progressively invert the key state parameters preset in the model. Based on the inversion results, fault diagnosis and model self-verification are performed to track the parameter degradation trend.

[0010] When the parameter degradation trend exceeds the threshold, a graded early warning information is issued based on the diagnostic results. Based on the historical degradation pattern of the parameters, future performance changes and energy consumption are predicted, the energy-saving effects of different maintenance schemes are simulated, and a quantitative decision support report is obtained.

[0011] As a preferred embodiment of the energy consumption monitoring method for prefabricated buildings based on digital twins according to the present invention, the construction of the digital twin model of the prefabricated building includes generating a finite element mesh of the prefabricated building based on the geometric data in BIM, and assigning physical property parameters to the finite element mesh based on the material property data in BIM to establish a digital twin model.

[0012] The finite element mesh includes tetrahedral elements of prefabricated components and locally refined meshes of connecting nodes;

[0013] In the digital twin model, a set of key state parameters characterizing the thermal performance of the building envelope are preset, including the equivalent thermal conductivity and contact thermal resistance, and are initialized according to the design values.

[0014] Establish a spatial mapping relationship between the finite element mesh nodes and the physical building sensor measurement points, and update the data in real time.

[0015] As a preferred embodiment of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in this invention, the key state parameters are parameters characterizing the thermal performance of the building envelope, which are composed of equivalent thermal conductivity and contact thermal resistance.

[0016] As a preferred embodiment of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in this invention, the synchronous simulation evolution includes using real-time data on indoor and outdoor temperature, humidity, solar radiation intensity, and equipment operating status collected by sensors as dynamic boundary conditions input into the digital twin model.

[0017] Based on the preset initial state parameters and coupled boundary conditions, the finite element method is used to solve the unsteady heat conduction equations of the building envelope and interior space, and to calculate and update the global temperature field distribution in real time.

[0018] Based on the updated temperature field, the real-time energy consumption of the building at the current moment is dynamically calculated using the energy balance equation;

[0019] The key state parameters obtained from the current synchronous simulation evolution are updated to the digital twin model as the initial parameters for the next simulation time step, thereby realizing the synchronous mapping and continuous evolution of the digital twin model and the physical building in terms of temperature field and energy consumption data.

[0020] As a preferred embodiment of the digital twin-based prefabricated building energy consumption monitoring method of the present invention, the multi-temporal-scale hierarchical collaborative inversion strategy includes: dividing the parameters to be inverted into global, regional, and local layers according to spatial scale; performing progressive inversion from global to local; setting the global layer parameters as optimization variables; temporarily fixing the regional and local layer parameters to initial values; using the building's total energy consumption data and the average temperature of key areas as comparison benchmarks; running the inversion algorithm to optimize the global layer parameters; fixing the inverted global layer parameter values ​​after convergence; inheriting the updated and fixed global layer parameter model; and... Regional layer parameters are used as optimization variables. While keeping the global and local layer parameters unchanged for the time being, energy consumption sub-metering data and temperature data from multiple sensors within the region are used as comparison benchmarks. Under the constraints of the fixed global layer, each region is inverted to fix the obtained regional layer parameter values. The updated and fixed regional layer parameter model is inherited. The local layer parameters of the anomaly detection data points are used as optimization variables. High-frequency, high-spatial-density local sensor data are used for fine comparison. Under the constraints of the fixed global and regional layer parameters, high-precision local inversion is performed to locate and quantify local anomalies.

[0021] Based on the different frequency characteristics of the parameters, the system is divided into low-frequency and high-frequency layers. Monitoring data at different time frequencies are matched for layered inversion. The slow changes in thermal conductivity, overall airtightness, and energy efficiency decay are used as matching parameters. Aggregated data over a long time window is used for inversion to filter out daytime fluctuations and instantaneous interference, and capture long-term trends. The indoor convective heat transfer coefficient, internal heat source intensity, and instantaneous equipment operating efficiency are used as matching parameters. Short periods with clear operating conditions are selected, and high-frequency data is used for dynamic inversion to capture the response of parameters to transient excitations.

[0022] In each inversion layer, an adaptive initial value of the Levenberg-Marquardt algorithm damping factor is set according to the characteristics of the current inversion layer, and the initial value is adaptively adjusted based on the error increase / decrease ratio of the iteration. If the error decreases, the initial value is reduced; if the error increases, the initial value is increased. The parameter values ​​after the convergence of the upper inversion layer are directly used as the known fixed input of the lower model and no longer participate in the lower optimization.

[0023] After the inversion of all layers is completed, a final forward simulation is performed using the complete model with updated inversion parameters. The simulation results are then compared with the most comprehensive measured data. If the error meets the requirements, the process ends; otherwise, the error is fed back to the corresponding layer, triggering a limited round of re-inversion and parameter adjustment, forming a self-verifying and self-correcting closed loop.

[0024] As a preferred embodiment of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in this invention, the fault diagnosis and model self-verification include: identifying parameter offsets based on the inversion results, matching the abnormal changes in the inverted parameters with the fault knowledge base, automatically diagnosing the type and location of hidden faults, quantifying the impact of the diagnosed faults on the current energy consumption, and tracking the parameter degradation trend; if the inversion fails to converge or the results are abnormal, the verification process for the model itself and the input data is automatically triggered.

[0025] As a preferred embodiment of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in this invention, the tracking parameter degradation trend includes constructing time series data of key state parameters by sorting them by timestamp based on historical inversion results;

[0026] Trend analysis is performed on the time series data to extract feature quantities that characterize the degradation rate and pattern of the parameters;

[0027] Based on the aforementioned features, the current degradation mode of the parameters is identified, and a corresponding time-varying prediction model is established.

[0028] Using the time-varying prediction model, the future values ​​of the key state parameters are extrapolated, and the remaining time to reach the preset failure threshold is evaluated.

[0029] Different levels of warnings are triggered based on whether the remaining time exceeds the corresponding threshold.

[0030] Secondly, the present invention provides a digital twin-based energy consumption monitoring system for prefabricated buildings, including a modeling unit that uses the finite element method to construct a digital twin model of the prefabricated building and initialize key state parameters.

[0031] The driving unit uses real-time data obtained from sensors as boundary conditions to drive the model to perform high-precision synchronous simulation evolution of temperature field and energy consumption.

[0032] The diagnostic unit, based on the real-time data and simulation results, adopts a multi-temporal-scale hierarchical collaborative inversion strategy, uses the Levenberg-Marquardt algorithm to progressively invert the key state parameters preset in the model, performs fault diagnosis and model self-verification based on the inversion results, and tracks the parameter degradation trend.

[0033] When the parameter degradation trend exceeds the threshold, the decision-making unit issues graded early warning information based on the diagnostic results, predicts future performance changes and energy consumption based on the historical degradation pattern of the parameters, simulates the energy-saving effect of different maintenance schemes, and obtains a quantitative decision support report.

[0034] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in the first aspect of the present invention.

[0035] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the energy consumption monitoring method for prefabricated buildings based on digital twins as described in the first aspect of the present invention.

[0036] The beneficial effects of this invention are as follows: By constructing a digital twin model of prefabricated buildings based on the finite element method, this invention achieves high-precision dynamic simulation and real-time monitoring of thermal performance. Employing a multi-temporal-scale hierarchical collaborative inversion strategy, it can automatically identify and correct key state parameters based on measured data, accurately reflecting performance degradation caused by factors such as material aging and construction deviations. This method not only enables early fault diagnosis and precise location but also predicts future building energy consumption trends and quantitatively evaluates the energy-saving effects of different maintenance schemes. It provides complete technical support for refined energy consumption management throughout the entire lifecycle of prefabricated buildings, from state perception and fault early warning to optimization decision-making. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of a digital twin-based method for monitoring the energy consumption of prefabricated buildings. Detailed Implementation

[0039] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0040] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0041] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0042] Reference Figure 1 As one embodiment of the present invention, this embodiment provides a method for monitoring the energy consumption of prefabricated buildings based on digital twins, comprising the following steps:

[0043] S1: Construct a digital twin model of the prefabricated building using the finite element method and initialize key state parameters.

[0044] To obtain accurate and complete information on building geometry, topology, material properties, and component relationships, which forms the basis for driving a high-fidelity physical simulation model, geometric data is extracted from the BIM model to generate a finite element mesh for prefabricated buildings. The finite element mesh includes tetrahedral elements of prefabricated components and locally refined meshes of connecting nodes. Based on the material property data in the BIM model, physical property parameters are assigned to the finite element mesh, including thermal physical parameters, material and structural layer parameters, boundary conditions and interaction parameters, system and equipment parameters, and initial and state parameters, thus establishing a digital twin model.

[0045] In the digital twin model, a set of key state parameters characterizing the thermal performance of the building envelope are preset, including the equivalent thermal conductivity of the walls, floors, and roof, as well as the contact thermal resistance of the joints of prefabricated components and structural connection nodes. The key state parameters are initialized according to the design values ​​to form a basic thermal performance benchmark consistent with the design state, providing an initial reference system for subsequent dynamic inversion and performance degradation analysis based on measured data.

[0046] A spatial mapping relationship is established between the finite element mesh nodes and the physical building sensor measurement points. Through spatial coordinate matching algorithm and topology association analysis, the three-dimensional installation coordinates of each physical sensor are bound to the nearest finite element mesh node or its element, forming a mapping relationship table containing sensor ID, spatial coordinates, corresponding mesh node and weight coefficient. This mapping table constitutes a bidirectional transmission interface for real-time monitoring data to drive the simulation model update, realizing the accurate transmission and interactive verification of measured data such as temperature and energy consumption with the physical quantities of the simulation nodes.

[0047] The construction of the digital twin model has physical consistency, data compatibility, and state evolution capability, providing a reliable computing foundation for subsequent accurate energy consumption monitoring, fault diagnosis, degradation prediction, and maintenance decision-making.

[0048] S2: Real-time data obtained through sensors is used as boundary conditions to drive the model to perform high-precision synchronous simulation evolution of temperature field and energy consumption.

[0049] Sensors deployed inside and outside the building and in equipment systems collect multi-dimensional physical field data in real time, including indoor and outdoor temperature, humidity, solar radiation intensity, and equipment operating status data. After data cleaning and time synchronization processing, the data is used as dynamic boundary conditions input into the digital twin model.

[0050] Based on the preset initial state parameters and coupled boundary conditions, the finite element method is used to solve the unsteady heat conduction equations of the building envelope and interior space, and to calculate and update the global temperature field distribution in real time.

[0051] Based on the updated temperature field, the real-time energy consumption of the building at the current moment is dynamically calculated through the energy balance equation. The key state parameters obtained from the current synchronous simulation evolution are then updated to the digital twin model as the initial parameters for the next simulation time step, thereby realizing the synchronous mapping and continuous evolution of the digital twin model and the physical building in terms of temperature field and energy consumption data.

[0052] This provides a reliable objective function for subsequent multi-temporal scale hierarchical collaborative inversion based on measured-simulation residuals, exposes hidden faults such as local thermal bridges, joint leakage, and material performance degradation in the building envelope in real time, supports continuous self-correction of model parameters and accurate tracking of performance degradation trends, and ultimately transforms passive energy consumption statistics into proactive, predictive, and precise full life cycle energy consumption management and fault pre-maintenance.

[0053] S3: Based on the real-time data and simulation results, a multi-temporal-scale hierarchical collaborative inversion strategy is adopted. The Levenberg-Marquardt algorithm is used to progressively invert the key state parameters preset in the model. Based on the inversion results, fault diagnosis and model self-verification are performed to track the parameter degradation trend.

[0054] The parameters to be inverted are divided into different layers according to spatial scale and different frequency characteristics for progressive inversion. This decomposes the original 30-50 dimensional, strongly nonlinear, and highly ill-conditioned parameter inversion problem into a series of low-dimensional, well-conditioned sub-problems with strong prior constraints, so that the inversion process has global convergence, local high accuracy and computational efficiency.

[0055] The parameters to be inverted are divided into global, regional, and local layers according to spatial scale, and the inversion is performed progressively from global to local.

[0056] Global layer inversion: Set the global layer parameters as optimization variables, temporarily fix the regional and local layer parameters to their initial values, use the building's total energy consumption data and the average temperature of key areas as comparison benchmarks, and execute the Levenberg-Marquardt algorithm (damping factor λ initial value 10). -2 The process is iterated until convergence, and the optimal parameters of the global layer are obtained and fixed.

[0057] Regional layer inversion: Inheriting the updated and fixed global layer parameter model, releasing 5-6 regional layer parameters for different orientations as optimization variables, keeping the global and local layer parameters temporarily unchanged, using energy consumption sub-metering data for each orientation or floor and temperature data from multiple sensors within the region as comparison benchmarks, and reducing the initial value of the LM damping factor λ to 10. -3 Under the strong constraints of the first layer parameters, the parameters of each region are optimized in parallel or serially, and the parameters of all regions are fixed after convergence.

[0058] Local layer inversion: Inheriting, updating, and fixing the regional layer parameter model, only releasing the locally sensitive parameters (usually ≤5, such as the thermal resistance of a certain joint, the thermal bridge ψ value of a certain window corner) selected by anomaly detection as optimization variables, using high-frequency, high-spatial-density local sensor data for fine comparison, and reducing the initial value of the LM damping factor λ to 10. -4 Under the constraint of keeping the global and regional layer parameters fixed, high-precision local inversion is performed to achieve quantitative location and separation of heat loss contribution of millimeter-level thermal bridges or joint defects.

[0059] Based on the different frequency characteristics of the parameters, the data is divided into low-frequency and high-frequency layers, and monitoring data of different time frequencies are matched for hierarchical inversion.

[0060] Low-frequency layer inversion: Slow parameters with a change period of more than 7 days are classified as low-frequency layers. Slow changes in thermal conductivity, overall air tightness, and slow energy efficiency decay are used as matching parameters. Aggregated energy consumption over a long window and outdoor average temperature data are used for inversion. Daytime fluctuations and instantaneous interference are filtered out, making it specifically designed to capture long-term degradation trends.

[0061] High-frequency layer inversion: Instantaneous parameters with a variation period of less than 24 hours are classified as high-frequency layers. Indoor convective heat transfer coefficient, internal heat source intensity, and instantaneous equipment operating efficiency are used as matching parameters. A short period of time with clear operating conditions is selected, and high-frequency data is used for dynamic inversion to capture the response of parameters to transient excitation.

[0062] After each inversion layer is completed, the optimal parameter values ​​and covariance matrix of the previous layer are directly used as the initial values ​​for the next layer's optimization, along with the Tikhonov regularization prior. An adaptive Levenberg-Marquardt algorithm damping factor initial value is set based on the characteristics of the current inversion layer, and this initial value is adaptively adjusted based on the error increase / decrease ratio during iteration.

[0063] If the objective function decreases by more than 30% in this iteration, then λ ← λ × 0.33.

[0064] If the objective function increases or decreases by less than 5%, then λ ← λ × 4.

[0065] And λ is always limited to

[10] -8 10 2 Within the range.

[0066] All inter-layer constraints are hard constraints (fixed values) and will not participate in subsequent optimization, ensuring that the inversion process is unidirectional and oscillating.

[0067] After the inversion of all layers is completed, a full-field forward transient simulation is performed using a complete digital twin model with updated parameters. The simulation results are then compared with the most comprehensive measured data. If the error meets all convergence criteria, the inversion ends and the parameter set is archived for degradation trend analysis. If any index exceeds the threshold, the corresponding layer is automatically backtracked according to the residual space distribution. After relaxing the range of that layer, a maximum of two rounds of finite re-inversion are triggered until the requirements are met or the maximum number of correction rounds is reached, forming a complete closed-loop self-verification mechanism.

[0068] Based on the inversion results, the parameter offset is identified, and the abnormal changes in the inverted parameters are matched with the fault knowledge base to automatically diagnose the type and location of hidden faults, quantify the impact of the diagnosed faults on the current energy consumption, and track the parameter degradation trend. If the inversion fails to converge or the results are abnormal, the verification process of the model itself and the input data is automatically triggered.

[0069] Through fault diagnosis and self-verification mechanisms, the system can automatically detect and accurately locate latent thermal faults, accurately quantify the energy consumption impact of faults, and close-loop correct the abnormal parameters without human intervention, thereby ensuring the robustness and physical interpretability of the digital twin model in long-term operation.

[0070] S4: When the parameter degradation trend exceeds the threshold, a graded early warning information is issued based on the diagnostic results, and based on the historical degradation pattern of the parameter, future performance changes and energy consumption are predicted, the energy-saving effect of different maintenance schemes is simulated, and a quantitative decision support report is obtained.

[0071] To upgrade instantaneous parameter inversion to long-term performance evolution management, and to transform the operation and maintenance of prefabricated buildings from passive response to proactive, quantitative, and plannable preventive maintenance, thereby reducing unplanned downtime and additional energy consumption caused by material and structural aging.

[0072] All key state parameters obtained from each complete inversion, along with their covariance, timestamp, and meteorological normalization identifier, are uniformly stored in a time series database to construct time series data of key state parameters. Trend analysis is then performed on the time series data to extract feature quantities that characterize the degradation rate of parameters and the model.

[0073] Based on the aforementioned features, the current degradation mode of the parameters is identified, and a corresponding time-varying prediction model is established. The future values ​​of the key state parameters are extrapolated using the time-varying prediction model, and the remaining time to reach the preset failure threshold is evaluated. The remaining time is then used to determine whether the corresponding threshold is exceeded.

[0074] In this embodiment, the time-varying prediction model refers to a dynamic prediction equation that automatically identifies the current degradation mode based on the historical time series of parameters and selects the corresponding mathematical function or statistical model. A linear model can be used to extrapolate the future values ​​of key thermal parameters deterministically or probabilistically. In other embodiments, the time-varying prediction model refers to a mathematical model that can explicitly characterize and handle the dynamic changes of the system's intrinsic parameters, structure, or relationships over time. This can employ aging models based on physical mechanisms, machine learning models, advanced statistical time series models, etc.

[0075] When preset performance safety or economic thresholds are exceeded, an early warning mechanism is automatically triggered. Warnings are categorized based on the degree of exceedance and the scope of impact (e.g., alert, warning, critical alarm) and distributed via a visual interface, SMS, or work order system. Abstract numerical changes are transformed into specific, timely, and actionable maintenance instructions, enabling managers to focus on areas of performance abnormality immediately and preventing minor issues from escalating into major failures or energy waste.

[0076] In digital twin models, different maintenance or renovation schemes are virtually implemented for the degraded parts or systems indicated by the early warning. Simulation calculations are performed on each virtual scheme, and the differences in energy consumption, peak load, indoor environment, etc., between it and the baseline scenario without maintenance under the same future boundary conditions are compared, thereby quantifying the energy-saving effect, investment payback period, and environmental benefits of each scheme.

[0077] The system automatically integrates all the above analysis results, including current problem diagnosis, future risk prediction, and quantitative comparison of various simulation solutions, to generate a structured decision support report. The report clearly presents the recommended solution, expected energy savings, cost estimates, investment payback period, and implementation priority suggestions, realizing the core value of asset preservation, energy efficiency improvement, and operation and maintenance cost optimization.

[0078] This embodiment also provides a digital twin-based energy consumption monitoring system for prefabricated buildings, including:

[0079] The modeling unit uses the finite element method to construct a digital twin model of the prefabricated building and presets and initializes key state parameters.

[0080] The driving unit uses real-time data obtained from sensors as boundary conditions to drive the model to perform high-precision synchronous simulation evolution of temperature field and energy consumption.

[0081] The diagnostic unit, based on the real-time data and simulation results, adopts a multi-temporal-scale hierarchical collaborative inversion strategy and uses the Levenberg-Marquardt algorithm to progressively invert the key state parameters preset in the model. Based on the inversion results, it performs fault diagnosis and model self-verification, and tracks the parameter degradation trend.

[0082] When the parameter degradation trend exceeds the threshold, the decision-making unit issues graded early warning information based on the diagnostic results, predicts future performance changes and energy consumption based on the historical degradation pattern of the parameters, simulates the energy-saving effect of different maintenance schemes, and obtains a quantitative decision support report.

[0083] This embodiment also provides a computer device applicable to the energy consumption monitoring method for prefabricated buildings based on digital twins, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the energy consumption monitoring method for prefabricated buildings based on digital twins as proposed in the above embodiment.

[0084] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0085] This embodiment also provides a storage medium on which a computer program is stored. When executed by a processor, the program implements the energy consumption monitoring method for prefabricated buildings based on digital twins as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0086] In summary, this invention achieves dynamic, accurate diagnosis, and proactive optimization of the thermal performance of prefabricated buildings throughout their entire lifecycle by constructing a high-fidelity digital twin that integrates physical mechanisms and real-time data, and by employing a multi-temporal-scale hierarchical collaborative inversion and intelligent early warning decision-making mechanism. Specifically, by combining finite element modeling and parametric inversion, it achieves precise digital characterization and state tracking of complex thermal details such as joint thermal resistance and local thermal bridges in prefabricated components; by combining real-time data-driven and multi-scale inversion, it enables the digital model to adaptively and synchronously update performance degradation due to material aging and construction deviations, overcoming the problem of disconnect between traditional static models and physical entities; and by establishing a closed loop between degradation trend prediction and virtual scheme simulation, it achieves a paradigm shift from passive energy consumption recording to proactive performance maintenance, providing quantitative decision support for preventive maintenance and energy efficiency optimization of prefabricated buildings, thereby improving the refinement, intelligence, and lifecycle economy of building operation and maintenance.

[0087] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for monitoring the energy consumption of prefabricated buildings based on digital twins, characterized in that: This includes using the finite element method to construct a digital twin model of a prefabricated building and initializing key state parameters; Real-time data obtained through sensors is used as boundary conditions to drive the model to perform high-precision synchronous simulation and evolution of temperature field and energy consumption. Based on the real-time data and simulation results, a multi-temporal-scale hierarchical collaborative inversion strategy is adopted. The Levenberg-Marquardt algorithm is used to progressively invert the key state parameters preset in the model. Based on the inversion results, fault diagnosis and model self-verification are performed to track the parameter degradation trend. When the parameter degradation trend exceeds the threshold, a graded early warning information is issued based on the diagnostic results. Based on the historical degradation pattern of the parameters, future performance changes and energy consumption are predicted, the energy-saving effects of different maintenance schemes are simulated, and a quantitative decision support report is obtained.

2. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 1, characterized in that: The construction of the digital twin model of the prefabricated building includes generating a finite element mesh of the prefabricated building based on the geometric data in BIM, and assigning physical property parameters to the finite element mesh based on the material property data in BIM to establish a digital twin model. The finite element mesh includes tetrahedral elements of prefabricated components and locally refined meshes of connecting nodes; In the digital twin model, a set of key state parameters characterizing the thermal performance of the building envelope are preset, including the equivalent thermal conductivity and contact thermal resistance, and are initialized according to the design values. Establish a spatial mapping relationship between the finite element mesh nodes and the physical building sensor measurement points, and update the data in real time.

3. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 2, characterized in that: The key state parameters are parameters characterizing the thermal performance of the building envelope, consisting of the equivalent thermal conductivity and the contact thermal resistance.

4. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 3, characterized in that: The synchronous simulation evolution includes using real-time data on indoor and outdoor temperature, humidity, solar radiation intensity, and equipment operating status collected by sensors as dynamic boundary conditions input into the digital twin model. Based on the preset initial state parameters and coupled boundary conditions, the finite element method is used to solve the unsteady heat conduction equations of the building envelope and interior space, and to calculate and update the global temperature field distribution in real time. Based on the updated temperature field, the real-time energy consumption of the building at the current moment is dynamically calculated using the energy balance equation; The key state parameters obtained from the current synchronous simulation evolution are updated to the digital twin model as the initial parameters for the next simulation time step, thereby realizing the synchronous mapping and continuous evolution of the digital twin model and the physical building in terms of temperature field and energy consumption data.

5. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 4, characterized in that: The multi-temporal-scale hierarchical collaborative inversion strategy includes: dividing the parameters to be inverted into global, regional, and local layers according to spatial scale; performing progressive inversion from global to local; setting global layer parameters as optimization variables; temporarily fixing regional and local layer parameters to initial values; using total building energy consumption data and average temperature of key areas as comparison benchmarks; running an inversion algorithm to optimize global layer parameters; fixing the obtained global layer parameter values ​​after convergence; inheriting, updating, and fixing the global layer parameter model; using regional layer parameters as optimization variables; keeping global and local layer parameters temporarily unchanged; using energy consumption sub-metering data of specific areas in the regional layer and temperature data from multiple sensors within the area as comparison benchmarks; performing inversion on each area under the constraint of the fixed global layer; fixing the obtained regional layer parameter values; inheriting, updating, and fixing the regional layer parameter model; using local layer parameters of anomaly detection data points as optimization variables; using high-frequency, high-spatial-density local sensor data for fine comparison; and performing high-precision local inversion under the constraint of fixed global and regional layer parameters to locate and quantify local anomalies. Based on the different frequency characteristics of the parameters, the data is divided into low-frequency and high-frequency layers. Monitoring data at different time frequencies are matched for hierarchical inversion. The slow change in thermal conductivity, the change in overall air tightness, and the slow decay of energy efficiency are used as matching parameters. Aggregated data over a long time window are used for inversion to filter out daytime fluctuations and instantaneous interference and capture long-term trends. Using indoor convective heat transfer coefficient, internal heat source intensity and instantaneous equipment operating efficiency as matching parameters, a short period of time with clear operating conditions is selected, and high-frequency data is used for dynamic inversion to capture the response of parameters to transient excitation. In each inversion layer, an adaptive initial value of the Levenberg-Marquardt algorithm damping factor is set according to the characteristics of the current inversion layer, and the initial value is adaptively adjusted based on the error increase / decrease ratio of the iteration. If the error decreases, the initial value is reduced; if the error increases, the initial value is increased. The parameter values ​​after the convergence of the upper inversion layer are directly used as the known fixed input of the lower model and no longer participate in the lower optimization. After the inversion of all layers is completed, a final forward simulation is performed using the complete model with updated inversion parameters. The simulation results are then compared with the most comprehensive measured data. If the error meets the requirements, the process ends. If the conditions are not met, the error is fed back to the corresponding layer, triggering a limited round of re-inversion and parameter adjustment, forming a self-verifying and self-correcting closed loop.

6. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 5, characterized in that: The fault diagnosis and model self-verification include identifying parameter offsets based on the inversion results, matching the abnormal changes in the inverted parameters with the fault knowledge base, automatically diagnosing the type and location of hidden faults, quantifying the impact of the diagnosed faults on the current energy consumption, and tracking the parameter degradation trend. If the inversion fails to converge or the results are abnormal, a verification process for the model itself and the input data will be automatically triggered.

7. The energy consumption monitoring method for prefabricated buildings based on digital twins as described in claim 6, characterized in that: The tracking parameter degradation trend includes constructing time series data of key state parameters based on historical inversion results sorted by timestamp; Trend analysis is performed on the time series data to extract feature quantities that characterize the degradation rate and pattern of the parameters; Based on the aforementioned features, the current degradation mode of the parameters is identified, and a corresponding time-varying prediction model is established. Using the time-varying prediction model, the future values ​​of the key state parameters are extrapolated, and the remaining time to reach the preset failure threshold is evaluated. Different levels of warnings are triggered based on whether the remaining time exceeds the corresponding threshold.

8. A prefabricated building energy consumption monitoring system based on digital twins, based on the file encryption method according to any one of claims 1 to 7, characterized in that: This includes modeling units, which utilize the finite element method to construct a digital twin model of the prefabricated building and initialize key state parameters; The driving unit uses real-time data obtained from sensors as boundary conditions to drive the model to perform high-precision synchronous simulation evolution of temperature field and energy consumption. The diagnostic unit, based on the real-time data and simulation results, adopts a multi-temporal-scale hierarchical collaborative inversion strategy, uses the Levenberg-Marquardt algorithm to progressively invert the key state parameters preset in the model, performs fault diagnosis and model self-verification based on the inversion results, and tracks the parameter degradation trend. When the parameter degradation trend exceeds the threshold, the decision-making unit issues graded early warning information based on the diagnostic results, predicts future performance changes and energy consumption based on the historical degradation pattern of the parameters, simulates the energy-saving effect of different maintenance schemes, and obtains a quantitative decision support report.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the prefabricated building energy consumption monitoring method based on digital twins as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the prefabricated building energy consumption monitoring method based on digital twins as described in any one of claims 1 to 7.

Images