A digital-twin-based heating power module operation modeling method and system

Abstract

The present application relates to the technical field of digital twinning, in particular to a warm power module operation modeling method and system based on digital twinning, comprising the following steps: obtaining entity temperature and power data to generate operation state identification, extracting non-steady-state peak value to construct thermal stress characteristics, building network from steady data and generating branch thermal resistance characteristics, reconnecting over-limit branches to generate a reconstructed topology model, and combining feature evaluation with heat dissipation output modeling results. In the present application, by dynamically dividing non-steady-state impact and steady-state operation according to the current change rate index, extracting peak data to construct thermal stress limit characteristics, and dynamically disconnecting and reassociating virtual nodes based on the actual thermal resistance over-limit situation to generate a reconstructed heat conduction network topology model, the adaptive dynamic update and accurate topology reorganization of the heat conduction path in the digital space are realized, the thermal limit characteristics under extreme working conditions are accurately described, and the accuracy of the comprehensive heat dissipation evaluation of the warm power module is improved, and the role of outputting effective control instructions is improved.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and in particular to a method and system for modeling the operation of HVAC power modules based on digital twins. Background Technology

[0002] Digital twin technology involves constructing virtual mappings of physical entities and reflecting the operational status and lifecycle evolution of corresponding physical devices in real time in digital space. Its main purpose is to monitor the state and predict the behavior of real-world systems through multi-source data fusion and analysis, and it is widely used in industries such as industrial manufacturing, urban management, aerospace, and the built environment. Traditional HVAC power module operation modeling methods refer to the process of establishing basic models for core power components in heating, ventilation, and air conditioning systems. These models are primarily used to analyze the thermodynamic characteristics and energy consumption variations of HVAC power modules under different operating conditions. They typically employ empirical formula derivation, finite element analysis, or multiple linear regression techniques based on historical operating data to construct basic models of the equipment's heat conduction and power consumption, thereby achieving the goal of understanding the operational characteristics of HVAC power modules.

[0003] Traditional HVAC power module operation modeling methods, while able to analyze the thermodynamic characteristics and energy consumption changes of equipment under different operating conditions using empirical formulas, finite element analysis, or multiple linear regression techniques based on historical data, suffer from problems such as rigid model structure, lagging parameter updates, and disconnect between thermodynamic mapping and actual conditions when the equipment faces drastic changes in transient current and dynamic evolution of local thermal resistance. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a digital twin-based HVAC power module operation modeling method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for modeling the operation of HVAC power modules based on digital twins, comprising the following steps:

[0006] S1: Obtain the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determine the current change evaluation index, add non-steady-state impact state label or stable operation state label, and generate operation state label data.

[0007] S2: Extract the first surface temperature value and the first input current associated with the unsteady-state impact state identifier from the operating status identifier data, and filter the peak values ​​of the first surface temperature value and the first input current value respectively to construct thermal stress limit characteristics;

[0008] S3: Extract the second surface temperature value and the second input current value associated with the stable operating state identifier from the operating state identifier data, construct an initial heat conduction network model, simultaneously extract the corresponding working voltage value to perform working power evaluation and thermodynamic mapping analysis, and generate branch actual thermal resistance characteristics;

[0009] S4: If the actual thermal resistance characteristic of the branch is greater than the preset upper limit threshold of thermal resistance, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset upper limit threshold of thermal resistance from the initial heat conduction network model, extract the first spatial coordinates of the associated virtual nodes, extract the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generate a digital twin reconstructed heat conduction network topology model.

[0010] S5: Extract the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, perform a comprehensive heat dissipation assessment and construct control commands, and output the HVAC power module operation modeling results based on the digital twin.

[0011] As a further aspect of the present invention, during the process of adding the non-steady-state impact state identifier or the stable operation state identifier, if the current change rate evaluation index is greater than the preset current change rate threshold, then the non-steady-state impact state identifier is added; if it is not greater than the preset current change rate threshold, then the stable operation state identifier is added.

[0012] As a further aspect of the present invention, the operating status identification data includes time-series feature records, operating condition classification labels, and dynamic parameter arrays; the thermal stress limit features include maximum heat load coefficient, thermal fatigue damage limit, and critical value for drastic temperature change; the branch actual thermal resistance features include interface contact thermal resistance, material thermal conductivity attenuation rate, and heat transfer hindrance coefficient; the reconstructed heat conduction network topology model based on the digital twin includes a newly constructed heat flow transmission path, an adaptive node matrix, and a dynamic structure mapping diagram; and the HVAC power module operation modeling results based on the digital twin include equipment cooling strategies, power allocation schemes, and module operation optimization reports.

[0013] As a further aspect of the present invention, the step of obtaining the running status identifier data specifically includes:

[0014] S111: Obtain the time node of the same data acquisition process, obtain the surface temperature value, input current value, and working voltage value under the time node, obtain the historical input current value corresponding to the previous historical acquisition process, obtain the time interval value associated with adjacent acquisition processes, and generate a basic parameter sequence.

[0015] S112: Based on the input current value and historical input current value in the basic parameter sequence, extract the current dynamic deviation feature from the input current value and historical input current value, calculate the rate of change by division based on the current dynamic deviation feature and the time interval value, and generate the current rate of change evaluation index.

[0016] S113: Compare the current change rate evaluation index with the preset current change rate threshold. If the current change rate evaluation index is greater than the preset current change rate threshold, an unsteady-state impact state identifier is added. If the current change rate evaluation index is not greater than the preset current change rate threshold, a stable operation state identifier is added, and operation state identifier data is generated.

[0017] As a further aspect of the present invention, the step of obtaining the thermal stress limit characteristic specifically comprises:

[0018] S211: Based on the operating status identification data, extract the first surface temperature value associated with the unsteady-state impact status identification, extract the associated first input current value, filter the peak surface temperature value in the first surface temperature value and the peak input current value in the first input current value, and generate the peak parameter extreme value.

[0019] S212: Based on the peak surface temperature value in the extreme values ​​of the peak parameters, obtain a preset normalization function to map the peak surface temperature value to a temperature characteristic coefficient, extract the peak input current value in the extreme values ​​of the peak parameters, map the peak input current value to a current characteristic coefficient according to the preset normalization function, and generate a set of normalization coefficients.

[0020] S213: Based on the temperature characteristic coefficient and current characteristic coefficient in the normalized coefficient set, perform a weighted superposition and fusion operation, and output the thermal stress correlation extreme value. Based on the thermal stress correlation extreme value, perform thermal evaluation analysis to extract the extreme value parameters and generate the thermal stress limit characteristics.

[0021] As a further aspect of the present invention, the step of obtaining the actual thermal resistance characteristics of the branch specifically includes:

[0022] S311: Based on the second surface temperature value associated with the stable operation status identifier in the operation status identifier data, extract the pre-constructed initial heat conduction network model containing virtual nodes and virtual branch segments, map the second surface temperature value to the initial heat conduction network model, calculate the difference of the second surface temperature values ​​at both ends of the virtual branch segment, and generate node temperature distribution characteristics.

[0023] S312: Extract the second input current value and working voltage value associated with the stable operation status identifier based on the operation status identifier data, perform multiplication operation on the second input current value and working voltage value to obtain the power output value, perform power status conversion processing to extract parameter ratio, and generate working power evaluation index.

[0024] S313: Perform thermodynamic mapping analysis on the node temperature distribution characteristics and the working power evaluation index, convert the working power evaluation index into heat flow divergence parameters, calculate the ratio of the node temperature distribution characteristics to the heat flow divergence parameters to extract the resistance attribute values, use the resistance attribute values ​​as impedance parameters to assign and record branch path parameters, and generate the actual thermal resistance characteristics of the branch.

[0025] As a further aspect of the present invention, the steps for obtaining the reconstructed heat conduction network topology model of the digital twin are specifically as follows:

[0026] S411: Based on the preset thermal resistance upper limit threshold, compare the actual thermal resistance characteristics of the branch with the preset thermal resistance upper limit threshold. If the actual thermal resistance characteristics of the branch are greater than the preset thermal resistance upper limit threshold, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset thermal resistance upper limit threshold from the initial heat conduction network model, extract the local node information of the disconnected area, and generate a disconnected network topology map.

[0027] S412: Based on the first spatial coordinates of the associated virtual nodes in the disconnected network topology diagram, extract the second spatial coordinates of the remaining virtual nodes in the initial heat conduction network model, obtain the coordinate difference between the first spatial coordinates and the second spatial coordinates, calculate the coordinate difference according to the Euclidean distance formula, extract the extreme distance value, and generate the node spatial proximity index.

[0028] S413: Extract the virtual nodes associated with the minimum value according to the node spatial proximity index, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node associated with the minimum node spatial proximity index, generate new virtual branch segments, combine multiple virtual branch segments to reorganize the graph parameters, establish three-dimensional connection association, and generate a digital twin reconstructed heat conduction network topology model.

[0029] As a further aspect of the present invention, the step of obtaining the HVAC power module operation modeling results based on digital twin specifically includes:

[0030] S511: Based on the reconstructed heat conduction network topology model of the digital twin, extract the actual thermal resistance characteristics of the remaining branches associated with each virtual branch segment that are less than the preset thermal resistance upper limit threshold, combine the thermal stress limit characteristics to perform system heat dissipation state evaluation and conversion processing, extract the heat dissipation quantitative evaluation score of each node area, and generate the comprehensive heat dissipation efficiency evaluation index of the HVAC power module.

[0031] S512: Compare the comprehensive heat dissipation efficiency evaluation index of the HVAC power module with the preset lower limit threshold of heat dissipation efficiency. If the comprehensive heat dissipation efficiency evaluation index of the HVAC power module is less than the preset lower limit threshold of heat dissipation efficiency, retrieve the input working voltage of the HVAC power module, arrange the reduced duty cycle control operation command stream for the input working voltage combined level sequence, and generate equipment control command.

[0032] S513: According to the equipment control command, perform information assembly and aggregation operation with the reconstructed heat conduction network topology model of the digital twin to form a set of operating configuration parameters, and perform file format normalization and merging on the set of operating configuration parameters to obtain the operating modeling result of the HVAC power module based on the digital twin.

[0033] A digital twin-based HVAC power module operation modeling system includes:

[0034] The operation status identification module obtains the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determines the current change evaluation index, adds non-steady-state impact status identification or stable operation status identification, and generates operation status identification data.

[0035] The limit feature extraction module extracts the first surface temperature value and the first input current associated with the unsteady impact state identifier from the operating state identifier data, and filters the peak values ​​of the first surface temperature value and the first input current value respectively to construct thermal stress limit features;

[0036] The branch thermal resistance analysis module extracts the second surface temperature value and the second input current value associated with the stable operating state identifier from the operating state identifier data, constructs an initial heat conduction network model, simultaneously extracts the corresponding operating voltage value to perform operating power evaluation and thermodynamic mapping analysis, and generates the actual thermal resistance characteristics of the branch.

[0037] The topology network reconstruction module, if the actual thermal resistance characteristic of the branch is greater than the preset thermal resistance upper limit threshold, disconnects the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset thermal resistance upper limit threshold from the initial heat conduction network model, extracts the first spatial coordinates of the associated virtual nodes, extracts the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establishes a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generates a digital twin reconstructed heat conduction network topology model;

[0038] The integrated heat dissipation control module extracts the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, performs a comprehensive heat dissipation assessment, constructs control commands, and outputs the HVAC power module operation modeling results based on the digital twin.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In this invention, by dynamically dividing unsteady-state impact and stable operation states according to the current change rate index, extracting peak data to construct thermal stress limit characteristics, and dynamically disconnecting and re-associating virtual nodes based on the actual thermal resistance exceeding the limit of branches to generate a reconstructed heat conduction network topology model, the invention achieves adaptive dynamic updating and precise topology reorganization of heat conduction paths in digital space, accurately characterizes thermal limit characteristics under extreme conditions, improves the accuracy of comprehensive heat dissipation assessment of HVAC power modules, and outputs effective control commands. Attached Figure Description

[0041] Figure 1 This is a flowchart of the main steps of the present invention;

[0042] Figure 2 This is a flowchart of the process for obtaining the running status identification data of the present invention;

[0043] Figure 3 This is a flowchart of the process for obtaining the thermal stress limit characteristics of the present invention;

[0044] Figure 4 This is a flowchart illustrating the actual thermal resistance characteristics obtained in this invention.

[0045] Figure 5 This is a flowchart illustrating the process of obtaining the reconstructed heat conduction network topology model using digital twins according to the present invention.

[0046] Figure 6 This is a flowchart illustrating the process of obtaining the operational modeling results of the HVAC power module based on digital twins, as described in this invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0048] Please see Figure 1 A method for modeling the operation of HVAC power modules based on digital twins includes the following steps:

[0049] S1: Obtain the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determine the current change evaluation index, if the current change rate evaluation index is greater than the preset current change rate threshold, then add an unsteady-state impact state label, if it is not greater than the preset current change rate threshold, then add a stable operation state label, and generate operation state label data.

[0050] S2: Extract the first surface temperature value and the first input current associated with the unsteady-state impact state in the operating status identification data, and filter the peak values ​​in the first surface temperature value and the peak values ​​in the first input current value to construct the thermal stress limit characteristics;

[0051] S3: Extract the second surface temperature value and the second input current value associated with the stable operation status identifier from the operation status identifier data, construct the initial heat conduction network model, simultaneously extract the corresponding working voltage value to perform working power evaluation and thermodynamic mapping analysis, and generate the actual thermal resistance characteristics of the branches.

[0052] S4: If the actual thermal resistance characteristic of a branch is greater than the preset upper limit threshold of thermal resistance, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branches that are greater than the preset upper limit threshold of thermal resistance from the initial heat conduction network model, extract the first spatial coordinates of the associated virtual nodes, extract the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generate a digital twin reconstructed heat conduction network topology model.

[0053] S5: Extract the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, conduct a comprehensive heat dissipation assessment and construct control commands, and output the HVAC power module operation modeling results based on the digital twin.

[0054] The operational status identification data includes time-series feature records, operating condition classification labels, and dynamic parameter arrays; thermal stress limit characteristics include maximum heat load coefficient, thermal fatigue damage limit, and critical value for drastic temperature change; branch actual thermal resistance characteristics include interface contact thermal resistance, material thermal conductivity attenuation rate, and heat transfer hindrance coefficient; the reconstructed heat conduction network topology model based on digital twins includes newly constructed heat flow transmission paths, adaptive node matrices, and dynamic structure mapping diagrams; the operational modeling results of HVAC power modules based on digital twins include equipment cooling strategies, power allocation schemes, and module operation optimization reports.

[0055] Please see Figure 2 Step S1 is as follows:

[0056] S111: Obtain the time node of the same data acquisition process, obtain the surface temperature value, input current value, and working voltage value under the time node, obtain the historical input current value corresponding to the previous historical acquisition process, obtain the time interval value associated with adjacent acquisition processes, and generate a basic parameter sequence.

[0057] To construct a high-fidelity digital twin sensing layer, high-precision thermal sensors and Hall current / voltage sensors deployed on the surface of HVAC power semiconductors collect real-time operational data. This data serves as the initial input source for modeling the HVAC power module's operation, extracting the timestamp corresponding to the first data acquisition process. At this timestamp, the surface temperature value transmitted by the sensors is extracted as 65.2℃, the associated input current value as 12.4A, and the operating voltage value as 220.5V. Subsequently, the previous round of acquisition tags recorded in the historical database are retrieved, extracting the historical input current value corresponding to the previous historical acquisition process as 11.8A. The time interval between the current timestamp and the previous historical timestamp is obtained; the current timestamp is 10:05:12, and the previous timestamp is 10:05:10. The time interval is calculated as 2 seconds. Then, the surface temperature value, input current value, operating voltage value, historical input current value, and time interval value are combined and concatenated according to a fixed data frame format to generate a basic parameter sequence containing five physical parameters, thereby establishing a rigorous parameter benchmark for digital twin entity mapping. This basic data acquisition process effectively avoids the problem of time sequence disorder of scattered data, ensures the synchronization of feature extraction, and significantly improves the timeliness of subsequent HVAC power module operation modeling.

[0058] Table 1. Examples of Basic Parameter Sequence Acquisition

[0059] 1 65.2℃ 12.4A 220.5V 2 66.8℃ 13.1A 220.3V 3 64.5℃ 11.9A 220.6V

[0060] Table 1 shows a summary of the sensor monitoring data from the first three batches.

[0061] S112: Based on the input current value and historical input current value in the basic parameter sequence, extract the current dynamic deviation characteristics of the input current value and historical input current value, calculate the rate of change by division based on the current dynamic deviation characteristics and time interval values, and generate the current rate of change evaluation index.

[0062] The system directly reads the input current value and historical input current value from the aforementioned basic parameter sequence, and performs a difference calculation to extract the current dynamic deviation feature. This feature serves as a key variable for characterizing the transient evolution boundary in the HVAC power module operation modeling. Specifically, the calculation logic involves subtracting the current input current value from the historical input current value from the previous acquisition process to obtain the absolute deviation. Substituting the aforementioned parameter values, the input current value is 12.4A, and the historical input current value is 11.8A; the difference yields a current dynamic deviation feature of 0.6A. Subsequently, the time interval value from the basic parameter sequence is used to calculate the quotient between the obtained current dynamic deviation feature and this time interval value, thereby extracting the current change rate reflecting transient current fluctuations. Based on this value, the 0.6A deviation feature is divided by the 2-second time interval value, resulting in a change rate of 0.3A per second. This result is directly designated as the current change rate evaluation index, thus enriching the dynamic response dimension of the digital twin system. The advantage of this operational logic is that it effectively quantifies the transient steepness of the current surge through joint calculation of the difference quotient. Compared with the traditional single-point amplitude judgment mechanism, it can more accurately reflect the intensity of energy jumps during the operation of the HVAC module, ensuring the synchronization rate between the digital twin mapping and the real physical mutation.

[0063] S113: Compare the current change rate evaluation index with the preset current change rate threshold. If the current change rate evaluation index is greater than the preset current change rate threshold, an unsteady-state impact state label is added. If the current change rate evaluation index is not greater than the preset current change rate threshold, a stable operation state label is added, and operation state label data is generated.

[0064] The derived current change rate evaluation index is extracted, and a preset current change rate threshold is retrieved from the local configuration file. A numerical comparison is performed to provide accurate operational condition classification for the state machine distribution of the digital twin model. This preset current change rate threshold is obtained by averaging the extreme values ​​of safety test fluctuations under 100 historical full-load operating conditions, with a specific value set at 0.25A per second. The calculated current change rate evaluation index of 0.3A per second is compared with the preset threshold of 0.25A per second. Since 0.3A is greater than the threshold limit of 0.25A, it is determined that the current device has experienced a sudden power overload, and a non-steady-state impact state identifier is written to the end of the current data frame. If the current change rate evaluation index of a subsequent batch is 0.15A per second, it is determined that it is not greater than the preset threshold, and a stable operating state identifier is appended to the end of the data frame. This tag allocation action generates operating state identifier data. The advantage of this logic is that it achieves a binary, fine-grained isolation of operating states through the boundary division of hard thresholds.

[0065] Please see Figure 3 Step S2 is as follows:

[0066] S211: Based on the operating status identifier data, extract the first surface temperature value associated with the unsteady-state impact status identifier, extract the associated first input current value, filter the peak surface temperature value in the first surface temperature value and the peak input current value in the first input current value, and generate the peak parameter extreme value;

[0067] The generated operational status identification data is traversed, and data slicing is performed on data frames carrying non-steady-state impact status identification. All surface temperature records during the impact period are extracted as the first set of surface temperature values, and all input current records within this interval are extracted as the first set of input current values ​​to construct the performance degradation boundary of the digital twin model under extreme conditions. Subsequently, a local maximum search process is performed on the two sets of data, comparing the values ​​of each record to filter out the highest point in the fluctuation process. Assuming data is collected at 15 time points within a non-steady-state interval, the highest value in the first set of surface temperature values ​​is extracted, yielding a peak surface temperature of 78.5℃, and the highest value in the first set of input current values ​​is extracted, yielding a peak input current of 18.2A. The extracted highest temperature and highest current records are paired and packaged to generate peak parameter extrema specifically characterizing short-term overload intensity. The advantage of this filtering operation is that it accurately identifies the most severe operating parameters that the power module experiences at the moment of electromagnetic shock.

[0068] S212: Based on the peak surface temperature value in the extreme values ​​of the peak parameters, obtain a preset normalization function to map the peak surface temperature value to a temperature characteristic coefficient, extract the peak input current value in the extreme values ​​of the peak parameters, map the peak input current value to a current characteristic coefficient according to the preset normalization function, and generate a set of normalization coefficients.

[0069] The generated peak parameter extreme values ​​are read, and the peak surface temperature and peak input current values ​​are extracted and fed into the data scaling processing stream to meet the standardized ingestion requirements of multi-source heterogeneous data in the digital twin environment. A linear mapping operation is performed by calling the system's underlying preset normalization function, which is configured with temperature and current ranges. The actual peak parameter is subtracted from the theoretical lower limit, and then divided by the difference between the theoretical upper and lower limits to complete the conversion. Setting the theoretical lower limit to 20℃ and the upper limit to 100℃, the aforementioned peak surface temperature value of 78.5℃ is substituted into this logic, subtracted by 20 to obtain 58.5, and then divided by the range difference of 80 to obtain the temperature characteristic coefficient of 0.73. Similarly, setting the theoretical current lower limit to 0A and the upper limit to 25A, the peak input current value of 18.2A is substituted into the calculation, and divided by 25 to obtain the current characteristic coefficient of 0.728. The calculated temperature and current characteristic coefficients are then aggregated to generate a dimensionless set of normalized coefficients, laying a numerical foundation for the unified feature space processing of HVAC power module operation modeling. The advantage of this mapping process is that it greatly reduces the absolute order-of-magnitude influence of multi-source heterogeneous parameters by eliminating the metric differences between different physical dimensions.

[0070] S213: Perform a weighted superposition and fusion operation based on the temperature characteristic coefficients and current characteristic coefficients in the normalized coefficient set, and output the thermal stress correlation extreme value. Perform thermal evaluation and analysis based on the thermal stress correlation extreme value to extract the extreme value parameters and generate the thermal stress limit characteristics.

[0071] A set of normalized coefficients, including temperature and current characteristic coefficients, is retrieved. Independent influence weights are assigned to each characteristic, and a weighted superposition and fusion operation is performed. The resulting composite index deeply supports the physical damage prediction mechanism of digital twins. The weight values ​​are set based on material thermal conductivity testing experience, with a temperature weight of 0.6 and a current weight of 0.4. Multiplying the temperature characteristic coefficient (0.73) by the temperature weight (0.6) yields 0.438, and multiplying the current characteristic coefficient (0.728) by the current weight (0.4) yields 0.2912. The products are then summed, ultimately outputting a thermal stress correlation extreme value of 0.7292. Based on this thermal stress correlation extreme value, a thermodynamic assessment analysis is performed. The correlation extreme value is multiplied by the initial material hardness coefficient to extract an extreme parameter characterizing the degree of micro-damage to the device. With a hardness coefficient set to 120 MPa, the calculated extreme parameter is 87.504 MPa, which is then converted into a thermal stress limit characteristic reflecting the device's critical pressure point and fed back to the HVAC power module's operational modeling module as a fatigue threshold. The advantage of this weighted superposition fusion logic is that it rationally allocates the contribution ratio of heat source and heat accumulation effect to the overall stress.

[0072] Please see Figure 4 Step S3 is as follows:

[0073] S311: Based on the second surface temperature value associated with the stable operation status identifier in the operation status identifier data, extract the pre-constructed initial heat conduction network model containing virtual nodes and virtual branch segments, map the second surface temperature value to the initial heat conduction network model, calculate the difference of the second surface temperature value at both ends of the virtual branch segment, and generate node temperature distribution characteristics.

[0074] Data frames with stable operating status indicators are separated from the operating status identification data, and the second surface temperature values ​​under normal operating conditions are parsed from them. An initial heat conduction network model stored on a cloud server is synchronously loaded. This model consists of virtual nodes representing local heat points of the device and virtual branch segments representing heat transfer paths, which are interwoven. Multiple sets of parsed second surface temperature values ​​are assigned to the corresponding virtual nodes according to their physical spatial coordinates to complete the digital twin thermal field mapping, achieving precise state synchronization from physical to virtual. Subsequently, for any two adjacent virtual nodes in the model, their respective bound second surface temperature values ​​are extracted and the difference is calculated to evaluate the thermal diffusion gradient along a specific virtual branch segment. For example, the temperature at node 1 is 45.2℃, and the temperature at node 2 is 42.1℃. A subtraction operation is performed between these two values ​​to obtain a temperature difference of 3.1℃. The differences in the second surface temperature values ​​at both ends of all virtual branch segments are then summarized to generate the node temperature distribution characteristics. The advantage of this heat network mapping and temperature difference calculation process is that it discretizes and refines macroscopic temperature data into microscopic temperature gradient fields, establishes a gridded analytical framework for modeling the operation of HVAC power modules, and effectively reveals the risk of local heat accumulation inside the modules.

[0075] S312: Extract the second input current value and working voltage value associated with the stable operation status identifier based on the operation status identifier data, perform multiplication operation on the second input current value and working voltage value to obtain the power output value, perform power status conversion processing to extract parameter ratio, and generate working power evaluation index.

[0076] Within the stable operating state data segment, the second input current value fed back from the device end side and the corresponding operating voltage value are extracted synchronously to drive the load state machine deduction process of the HVAC power module operation modeling. The module performs a multiplication operation on the two extracted electrical parameters to obtain the overall power output value. Substituting the second input current value of 8.5A and the operating voltage value of 220V, the multiplication yields a power output value of 1870 watts (i.e., the calculation logic). Subsequently, this power output value undergoes power state conversion processing, using the device's rated power upper limit parameter as a reference benchmark. Setting the rated power upper limit parameter to 2500 watts, the actual power output value of 1870 watts is divided by the upper limit parameter of 2500 watts, and the parameter ratio of the current operating load to the theoretical limit is calculated to be 0.748. This ratio is directly defined as the stable operating power evaluation index, providing a real reference anchor for energy injection within the digital twin space. The advantage of this logical operation is that by replacing the nominal power with the real-time product of voltage and current and introducing the upper limit ratio conversion, it accurately depicts the actual load level of the device under the current operating environment.

[0077] Table 2. Example of correlation calculation for various parameters under steady state.

[0078]

[0079] Table 2 shows the specific impact of different input parameters on the final proportional calculation results during continuous and stable operation.

[0080] S313: Perform thermodynamic mapping analysis on the node temperature distribution characteristics and working power evaluation index, convert the working power evaluation index into heat flow divergence parameters, calculate the ratio of node temperature distribution characteristics to heat flow divergence parameters to extract resistance attribute values, use the resistance attribute values ​​as impedance parameters to assign and record branch path parameters, and generate actual thermal resistance characteristics of the branch.

[0081] The core step in establishing the thermal impedance matrix for the HVAC power module operation model is to retrieve the node temperature distribution characteristics and operating power evaluation index calculated in the previous stage and perform a mapping analysis process from electrical energy consumption to thermal conversion. The operating power evaluation index of 0.748 is selected and multiplied by the inherent heat conversion rate of the device to obtain the corresponding heat flow dispersion parameter. The heat conversion rate is set to 12 percentage points of the total power consumption. After multiplication, the heat flow dispersion parameter is obtained as 0.0897 joules per second. Then, the temperature difference between the two ends of the aforementioned virtual branch segment, i.e., the node temperature distribution characteristic, is extracted as 3.1℃. This distribution characteristic is divided by the heat flow dispersion parameter of 0.0897 joules per second, and the resistance attribute value along the heat transfer path is extracted as 34.55℃ per watt. This obtained resistance attribute value of 34.55 is used as the impedance parameter and directly written into the attribute configuration table of the corresponding virtual branch segment for path parameter assignment. This process is repeated for all paths to generate the actual thermal resistance characteristics of the entire network branches, thereby dynamically assigning physical boundary attributes to each virtual branch in the digital twin model. The advantage of this ratio calculation is that it couples the abstract temperature difference and heat flow into an intuitive thermal resistance index.

[0082] Please see Figure 5 Step S4 is as follows:

[0083] S411: Based on the preset thermal resistance upper limit threshold, compare the actual thermal resistance characteristics of the branch with the preset thermal resistance upper limit threshold. If the actual thermal resistance characteristics of the branch are greater than the preset thermal resistance upper limit threshold, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset thermal resistance upper limit threshold from the initial heat conduction network model, extract the local node information of the disconnected area, and generate a disconnected network topology map.

[0084] A preset thermal resistance upper limit threshold is obtained from the system to perform network topology pruning, dynamically pruning the digital twin network to match the actual degradation characteristics of the physical entity. This preset thermal resistance upper limit threshold is extracted based on fatigue limit test data of 50 aging devices, with the average thermal resistance at the interface peeling of various materials set at 30°C per watt. The actual thermal resistance characteristic of the branch with a specific value of 34.55°C per watt is compared with the threshold of 30°C per watt. Given that 34.55 is significantly greater than the threshold limit of 30, it is determined that the actual physical heat conduction channel corresponding to the virtual branch segment has undergone severe degradation or even breakage. Subsequently, in the graphical data structure of the initial heat conduction network model, an edge element deletion operation is performed to cut off the virtual branch segment associated with the actual thermal resistance characteristic exceeding the threshold. Then, the coordinate information of the isolated local nodes at both ends of the cut segment and the records of missing connectivity are extracted. This topology culling action generates a disconnected network topology map reflecting the internal thermal bottleneck region of the device, providing an accurate damage mapping basis for the heat island effect simulation of subsequent HVAC power module operation modeling. The advantage of this severance decision logic is that it simulates the thermal disconnection phenomenon at the real material interface. Compared with the static and invariant network model, this dynamic pruning operation reduces the localization error of damage.

[0085] S412: Based on the first spatial coordinates of the associated virtual nodes in the disconnected network topology diagram, extract the second spatial coordinates of the remaining virtual nodes in the initial heat conduction network model, obtain the coordinate difference between the first and second spatial coordinates, calculate the coordinate difference according to the Euclidean distance formula, extract the extreme distance value, and generate the node spatial proximity index.

[0086] The previously generated disconnected network topology is retrieved, and the three-dimensional first spatial coordinate data of the isolated virtual nodes exposed due to branch severance are read. The azimuth coordinate components in length, width, and height are extracted. Simultaneously, the unbroken normal flow connectivity areas in the initial heat conduction network model are traversed, and the second spatial coordinate data of all remaining virtual nodes are extracted. For any isolated node and any remaining node, the difference between the coordinate components in each dimension is subtracted. Assuming the first spatial coordinates are 3, 4, and 5, and the second spatial coordinates are 6, 8, and 5, the difference results are -3, -4, and 0, respectively. Following the Euclidean distance calculation rule, the squares of each dimension difference are calculated, and then summed. Finally, the square root of the sum is calculated to obtain the extreme distance value of 5. The extreme distance values ​​of each node pairing calculated through traversal are summarized into a set as a node spatial proximity index characterizing the search range of available heat conduction paths. This provides accurate spatial measurement for the adaptive reconnection and structural repair of the digital twin system, and is also key preliminary data for optimizing the fault tolerance of the HVAC power module operation modeling network. The advantage of this spatial distance calculation is that it accurately quantifies the physical distance that heat needs to travel to achieve flow transfer, providing a reliable spatial search basis for subsequent reconstruction of the heat transfer path and greatly reducing the probability of erroneous reconnection.

[0087] S413: Extract virtual nodes associated with the minimum value based on the node spatial proximity index, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node associated with the minimum node spatial proximity index, generate new virtual branch segments, combine multiple virtual branch segments to reorganize the graph parameters, establish three-dimensional connection association, and generate a digital twin reconstructed heat conduction network topology model.

[0088] A global ascending sort search operation is performed on the aforementioned set of node spatial proximity indices to select the index with the smallest value. Assuming the minimum value found is 2.5, the remaining virtual nodes associated with this minimum value are located. Then, between the isolated virtual node corresponding to the first spatial coordinate and the node associated with the minimum value, a graph matrix edge element insertion operation is performed to open up a backup heat conduction channel, generating a new virtual branch segment. This search and docking operation is repeated until all isolated nodes are reconnected to the main network. Subsequently, the graph parameters of multiple newly created virtual branch segments are recombined with the original normal branch segments, the node connectivity adjacency matrix is ​​refreshed and the spatial direction is updated to establish a three-dimensional connection association that conforms to the current degradation status. After this network topology reconstruction process, a reconstructed heat conduction network topology model that maps to the real damaged physical entity is finally generated, thus perfectly reproducing the damage evolution process of the entity in virtual space, laying a dynamic topology foundation for deeper HVAC power module operation modeling. The advantage of this adjacency reconnection logic is that it adaptively restores the heat flow bypass phenomenon after local damage occurs.

[0089] Please see Figure 6 The S5 steps are as follows:

[0090] S511: Based on the digital twin, the reconstructed heat conduction network topology model extracts the actual thermal resistance characteristics of the remaining branches that are less than the preset thermal resistance upper limit threshold associated with each virtual branch segment. Combined with the thermal stress limit characteristics, the system heat dissipation status is evaluated and converted. The heat dissipation quantification score of each node area is extracted to generate the comprehensive heat dissipation efficiency evaluation index of the HVAC power module.

[0091] Accessing the previously constructed digital twin-based reconstructed heat conduction network topology model, its connectivity matrix is ​​traversed one by one, extracting the actual thermal resistance characteristics of the remaining branches bound to all intact virtual branch segments. The actual thermal resistance of one critical path is selected as 18.4℃ per watt. The previously calculated thermal stress limit characteristic of 87.504 MPa is retrieved, and a comprehensive evaluation and transformation process is performed to deeply drive the thermal evaluation system of the digital twin. The thermal stress limit characteristic is divided by the actual thermal resistance characteristic of the remaining branch to calculate the quotient; substituting the value, 87.504 divided by 18.4, yields a regional dissipation intensity of 4.755. This result is mapped to a 100-point evaluation system, supplemented by basic offset compensation parameters, and the corresponding node region's heat dissipation quantitative evaluation score is extracted as 84 points. Subsequently, the evaluation scores of all node regions within the model are arithmetically averaged to generate a globally applicable comprehensive heat dissipation efficiency evaluation index for HVAC power modules, serving as the most intuitive judgment output of the HVAC power module operation modeling health evaluation mechanism. The advantage of this joint evaluation conversion process is that it achieves cross-dimensional integration of the material's mechanical limit load and dynamically evolving thermal impedance.

[0092] Table 3. Calculation of heat dissipation quantification score for each node region.

[0093] 1 18.4 87.504 84 points 2 22.1 87.504 76 points 3 19.5 87.504 81 points

[0094] Table 3 details the impact of changes in parameters of different branch node regions on the final quantitative evaluation score.

[0095] S512: Compare the comprehensive heat dissipation efficiency evaluation index of the HVAC power module with the preset lower limit threshold of heat dissipation efficiency. If the comprehensive heat dissipation efficiency evaluation index of the HVAC power module is less than the preset lower limit threshold of heat dissipation efficiency, retrieve the input working voltage of the HVAC power module, arrange the reduced duty cycle control operation command stream for the input working voltage combined level sequence, and generate equipment control commands.

[0096] The calculated comprehensive heat dissipation efficiency evaluation index of the HVAC power module is extracted, and a preset lower limit threshold for heat dissipation efficiency is retrieved from the control logic memory for comparison. This preset lower limit threshold is determined by analyzing the critical experimental data of thermal runaway of the equipment, and the average extreme heat dissipation at the moment before the occurrence of thermal escape is set as 80 points. The heat dissipation efficiency evaluation index with a global average score of 78 points obtained from the aforementioned calculation is compared with this 80-point threshold. Since 78 points is less than the preset threshold of 80 points, it is determined that the equipment has a risk of overheating and self-destruction. Subsequently, a bus request is sent to retrieve the current input operating voltage waveform data of the HVAC power module, and the high and low level periods are identified. Pulse width modulation intervention is performed on the input operating voltage, and a new level sequence is combined by reducing the high level duration and increasing the low level waiting time. Based on the deduction results of the HVAC power module operation modeling, closed-loop active intervention is implemented, and the control operation command flow that reduces the current duty cycle by 15% is arranged. The sending address protocol is encapsulated to generate formal equipment control commands. This marks a substantial leap from single monitoring to intelligent-led control of the digital twin system. The advantage of this threshold comparison and duty cycle reduction intervention logic is that it realizes active frequency reduction protection based on digital twin evaluation results, which can complete load stripping intervention within 3 seconds before thermal collapse occurs, ensuring the survival rate of power modules.

[0097] S513: Based on the equipment control instructions, perform information compilation and aggregation operations with the reconstructed heat conduction network topology model of the digital twin to form a set of operating configuration parameters. Perform file format normalization and merging on the set of operating configuration parameters to obtain the operating modeling results of the HVAC power module based on the digital twin.

[0098] The system receives and issues frequency reduction duty cycle control commands and reads the latest digital twin data file of the reconstructed heat conduction network topology model from the memory resident area. It then performs information assembly and aggregation, seamlessly integrating the duty cycle correction magnitude parameters included in the control commands as additional control node attributes into the global configuration form of the topology model. This allows the model to synchronously reflect the heat dissipation trend after frequency reduction, forming a set of operational configuration parameters encompassing physical structural attributes and control logic strategies. Subsequently, it normalizes and merges the various graph coordinate files, attribute key-value pair files, and command control script execution files within this parameter set. It unifies and reshapes the multi-type file structures into a standardized data exchange format text, eliminating data interaction barriers between different subsystems. Finally, it outputs a closed-loop verified, highly coupled physical reality and algorithm strategy-based HVAC power module operational modeling result, achieving intelligent management and self-evolution throughout the entire lifecycle. The benefit of this assembly, aggregation, and format unification operation is that it elevates unidirectional monitoring and analysis to a full-element twin model supporting dynamic response strategy simulation.

[0099] A digital twin-based HVAC power module operation modeling system includes:

[0100] The operation status identification module is used to execute S1: obtain the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determine the current change evaluation index, attach the non-steady-state impact status identification or stable operation status identification, and generate operation status identification data.

[0101] The limit feature extraction module is used to perform S2: extract the first surface temperature value and the first input current associated with the unsteady impact state identifier in the operating status identifier data, filter the peak value in the first surface temperature value and the peak value in the first input current value respectively, and construct the thermal stress limit feature;

[0102] The branch thermal resistance analysis module is used to perform S3: extract the second surface temperature value and the second input current value associated with the stable operating state in the operating state identification data, construct the initial heat conduction network model, simultaneously extract the corresponding operating voltage value to perform operating power evaluation and thermodynamic mapping analysis, and generate the actual thermal resistance characteristics of the branch.

[0103] The topology network reconstruction module is used to execute S4: If the actual thermal resistance characteristic of a branch is greater than the preset upper limit threshold of thermal resistance, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset upper limit threshold of thermal resistance from the initial heat conduction network model, extract the first spatial coordinates of the associated virtual nodes, extract the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generate a digital twin reconstructed heat conduction network topology model.

[0104] The integrated heat dissipation control module is used to execute S5: extract the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, perform integrated heat dissipation assessment and construct control commands, and output the HVAC power module operation modeling results based on the digital twin.

[0105] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for modeling the operation of HVAC power modules based on digital twins, characterized in that, Includes the following steps: S1: Obtain the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determine the current change evaluation index, add non-steady-state impact state label or stable operation state label, and generate operation state label data. S2: Extract the first surface temperature value and the first input current associated with the unsteady-state impact state identifier from the operating status identifier data, and filter the peak values ​​of the first surface temperature value and the first input current value respectively to construct thermal stress limit characteristics; S3: Extract the second surface temperature value and the second input current value associated with the stable operating state identifier from the operating state identifier data, construct an initial heat conduction network model, simultaneously extract the corresponding working voltage value to perform working power evaluation and thermodynamic mapping analysis, and generate branch actual thermal resistance characteristics; S4: If the actual thermal resistance characteristic of the branch is greater than the preset upper limit threshold of thermal resistance, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset upper limit threshold of thermal resistance from the initial heat conduction network model, extract the first spatial coordinates of the associated virtual nodes, extract the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generate a digital twin reconstructed heat conduction network topology model. S5: Extract the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, perform a comprehensive heat dissipation assessment and construct control commands, and output the HVAC power module operation modeling results based on the digital twin.

2. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that: During the process of adding the unsteady-state impact status indicator or the stable operation status indicator, if the current change rate evaluation index is greater than the preset current change rate threshold, then the unsteady-state impact status indicator is added; if it is not greater than the preset current change rate threshold, then the stable operation status indicator is added.

3. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that: The operational status identification data includes time-series feature records, operating condition classification labels, and dynamic parameter arrays; the thermal stress limit features include maximum heat load coefficient, thermal fatigue damage limit, and critical value for drastic temperature change; the branch actual thermal resistance features include interface contact thermal resistance, material thermal conductivity attenuation rate, and heat transfer hindrance coefficient; the reconstructed heat conduction network topology model based on the digital twin includes a newly constructed heat flow transmission path, an adaptive node matrix, and a dynamic structure mapping diagram; the operational modeling results of the HVAC power module based on the digital twin include equipment cooling strategies, power allocation schemes, and module operation optimization reports.

4. The method for modeling the operation of HVAC power modules based on digital twins according to claim 1, characterized in that, The specific steps for obtaining the running status identifier data are as follows: S111: Obtain the time node of the same data acquisition process, obtain the surface temperature value, input current value, and working voltage value under the time node, obtain the historical input current value corresponding to the previous historical acquisition process, obtain the time interval value associated with adjacent acquisition processes, and generate a basic parameter sequence. S112: Based on the input current value and historical input current value in the basic parameter sequence, extract the current dynamic deviation feature from the input current value and historical input current value, calculate the rate of change by division based on the current dynamic deviation feature and the time interval value, and generate the current rate of change evaluation index. S113: Compare the current change rate evaluation index with the preset current change rate threshold. If the current change rate evaluation index is greater than the preset current change rate threshold, an unsteady-state impact state identifier is added. If the current change rate evaluation index is not greater than the preset current change rate threshold, a stable operation state identifier is added, and operation state identifier data is generated.

5. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that, The specific steps for obtaining the thermal stress limit characteristics are as follows: S211: Based on the operating status identification data, extract the first surface temperature value associated with the unsteady-state impact status identification, extract the associated first input current value, filter the peak surface temperature value in the first surface temperature value and the peak input current value in the first input current value, and generate the peak parameter extreme value. S212: Based on the peak surface temperature value in the extreme values ​​of the peak parameters, obtain a preset normalization function to map the peak surface temperature value to a temperature characteristic coefficient, extract the peak input current value in the extreme values ​​of the peak parameters, map the peak input current value to a current characteristic coefficient according to the preset normalization function, and generate a set of normalization coefficients. S213: Based on the temperature characteristic coefficient and current characteristic coefficient in the normalized coefficient set, perform a weighted superposition and fusion operation, and output the thermal stress correlation extreme value. Based on the thermal stress correlation extreme value, perform thermal evaluation analysis to extract the extreme value parameters and generate the thermal stress limit characteristics.

6. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that, The specific steps for obtaining the actual thermal resistance characteristics of the branch are as follows: S311: Based on the second surface temperature value associated with the stable operation status identifier in the operation status identifier data, extract the pre-constructed initial heat conduction network model containing virtual nodes and virtual branch segments, map the second surface temperature value to the initial heat conduction network model, calculate the difference of the second surface temperature values ​​at both ends of the virtual branch segment, and generate node temperature distribution characteristics. S312: Extract the second input current value and working voltage value associated with the stable operation status identifier based on the operation status identifier data, perform multiplication operation on the second input current value and working voltage value to obtain the power output value, perform power status conversion processing to extract parameter ratio, and generate working power evaluation index. S313: Perform thermodynamic mapping analysis on the node temperature distribution characteristics and the working power evaluation index, convert the working power evaluation index into heat flow divergence parameters, calculate the ratio of the node temperature distribution characteristics to the heat flow divergence parameters to extract the resistance attribute values, use the resistance attribute values ​​as impedance parameters to assign and record branch path parameters, and generate the actual thermal resistance characteristics of the branch.

7. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that, The specific steps for obtaining the reconstructed heat conduction network topology model from the digital twin are as follows: S411: Based on the preset thermal resistance upper limit threshold, compare the actual thermal resistance characteristics of the branch with the preset thermal resistance upper limit threshold. If the actual thermal resistance characteristics of the branch are greater than the preset thermal resistance upper limit threshold, disconnect the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset thermal resistance upper limit threshold from the initial heat conduction network model, extract the local node information of the disconnected area, and generate a disconnected network topology map. S412: Based on the first spatial coordinates of the associated virtual nodes in the disconnected network topology diagram, extract the second spatial coordinates of the remaining virtual nodes in the initial heat conduction network model, obtain the coordinate difference between the first spatial coordinates and the second spatial coordinates, calculate the coordinate difference according to the Euclidean distance formula, extract the extreme distance value, and generate the node spatial proximity index. S413: Extract the virtual nodes associated with the minimum value according to the node spatial proximity index, establish a connection between the virtual node corresponding to the first spatial coordinate and the virtual node associated with the minimum node spatial proximity index, generate new virtual branch segments, combine multiple virtual branch segments to reorganize the graph parameters, establish three-dimensional connection association, and generate a digital twin reconstructed heat conduction network topology model.

8. The method for HVAC power module operation modeling based on digital twin according to claim 1, characterized in that, The specific steps for obtaining the operational modeling results of the HVAC power module based on digital twin are as follows: S511: Based on the reconstructed heat conduction network topology model of the digital twin, extract the actual thermal resistance characteristics of the remaining branches associated with each virtual branch segment that are less than the preset thermal resistance upper limit threshold, combine the thermal stress limit characteristics to perform system heat dissipation state evaluation and conversion processing, extract the heat dissipation quantitative evaluation score of each node area, and generate the comprehensive heat dissipation efficiency evaluation index of the HVAC power module. S512: Compare the comprehensive heat dissipation efficiency evaluation index of the HVAC power module with the preset lower limit threshold of heat dissipation efficiency. If the comprehensive heat dissipation efficiency evaluation index of the HVAC power module is less than the preset lower limit threshold of heat dissipation efficiency, retrieve the input working voltage of the HVAC power module, arrange the reduced duty cycle control operation command stream for the input working voltage combined level sequence, and generate equipment control command. S513: According to the equipment control command, perform information assembly and aggregation operation with the reconstructed heat conduction network topology model of the digital twin to form a set of operating configuration parameters, and perform file format normalization and merging on the set of operating configuration parameters to obtain the operating modeling result of the HVAC power module based on the digital twin.

9. A HVAC power module operation modeling system based on digital twin, characterized in that, The system is used to implement the method according to any one of claims 1-8, comprising: The operation status identification module obtains the surface temperature value, input current value and time interval of the physical entity of the HVAC power module, determines the current change evaluation index, adds non-steady-state impact status identification or stable operation status identification, and generates operation status identification data. The limit feature extraction module extracts the first surface temperature value and the first input current associated with the unsteady impact state identifier from the operating state identifier data, and filters the peak values ​​of the first surface temperature value and the first input current value respectively to construct thermal stress limit features; The branch thermal resistance analysis module extracts the second surface temperature value and the second input current value associated with the stable operating state identifier from the operating state identifier data, constructs an initial heat conduction network model, simultaneously extracts the corresponding operating voltage value to perform operating power evaluation and thermodynamic mapping analysis, and generates the actual thermal resistance characteristics of the branch. The topology network reconstruction module, if the actual thermal resistance characteristic of the branch is greater than the preset thermal resistance upper limit threshold, disconnects the virtual branch segments associated with the actual thermal resistance characteristics of the branch that are greater than the preset thermal resistance upper limit threshold from the initial heat conduction network model, extracts the first spatial coordinates of the associated virtual nodes, extracts the second spatial coordinates of each remaining virtual node in the initial heat conduction network model, establishes a connection between the virtual node corresponding to the first spatial coordinate and the virtual node with the smallest association to construct a new virtual branch segment, and generates a digital twin reconstructed heat conduction network topology model; The integrated heat dissipation control module extracts the actual thermal resistance characteristics and thermal stress limit characteristics of the remaining branches in the reconstructed heat conduction network topology model of the digital twin, performs a comprehensive heat dissipation assessment, constructs control commands, and outputs the HVAC power module operation modeling results based on the digital twin.

Images