A method for optimizing a building envelope heat transfer performance parameter

Abstract

The application provides a building envelope heat transfer performance parameter optimization method, comprising the following steps: firstly, obtaining initial envelope information of a target building envelope and performing envelope partitioning, constructing a multi-working-condition thermal excitation working condition sequence, collecting envelope thermal response data and boundary input data under the action of the working condition sequence; performing time alignment, abnormality processing and working condition segmentation on the envelope thermal response data, extracting temperature response characteristics, heat flow response characteristics and time delay response characteristics, and forming a corresponding characteristic set; establishing a partition heat transfer mechanism model based on the corresponding characteristic set, inverting and verifying a time-varying heat transfer characteristic parameter set of each envelope partition; under the boundary constraint of adjustable parameters, generating a parameter search prior strategy by using imitating learning and constructing a candidate optimization parameter set, obtaining a predicted heat transfer performance result by using an envelope heat transfer prediction model; and constructing a heat transfer performance objective function based on building energy consumption, indoor thermal stability deviation and envelope parameter adjustment engineering quantity.

Description

Technical Field

[0001] This invention relates to the field of building thermal analysis and energy-saving optimization technology, and in particular to a method for optimizing the heat transfer performance parameters of building envelopes. Background Technology

[0002] Building envelopes are crucial for controlling building energy consumption and maintaining the indoor thermal environment. For existing buildings, the actual heat transfer performance often deviates from the nominal design value due to long-term effects such as material aging, construction deviations, and variable operating conditions. Therefore, accurately obtaining and optimizing the heat transfer performance parameters of the building envelope is an important foundation for building energy conservation diagnosis and retrofitting.

[0003] Existing technologies for analyzing and optimizing building envelope parameters typically employ steady-state heat transfer calculations, single-condition simulations, or local adjustments based on empirical rules, which have the following significant shortcomings:

[0004] First, there is insufficient extraction of dynamic response features: existing methods are mostly based on static parameters or single boundary inputs, which fail to fully stimulate and capture the dynamic thermal response features of buildings under real multiple disturbances such as shading opening and closing, ventilation starting and stopping, etc., resulting in a serious disconnect between the basic parameters on which optimization depends and the actual operating state.

[0005] Secondly, it is difficult to decouple the heat transfer mechanism components. Existing coarse-grained identification often confuses the complex heat transfer process of the building envelope, making it difficult to effectively separate different mechanism components such as conduction, heat storage, wind infiltration disturbance and solar heat gain. This makes it impossible to accurately locate the main sources of heat loss, resulting in poor accuracy and interpretability of parameter inversion.

[0006] Therefore, the present invention proposes a method for optimizing the heat transfer performance parameters of building envelopes; the information disclosed in the background section is only used to enhance the understanding of the background of this disclosure, and therefore may contain prior art information that is not common knowledge to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for optimizing the heat transfer performance parameters of building envelopes, thereby solving the technical problems mentioned in the background section.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A method for optimizing the heat transfer performance parameters of a building envelope includes the following steps:

[0010] S1. Obtain the initial enclosure information of the target building enclosure structure, divide the target building enclosure structure into zones, construct a multi-condition thermal excitation condition sequence corresponding to each zone, and collect enclosure thermal response data and boundary input data.

[0011] S2. Perform time alignment, anomaly removal, and operating condition segmentation on the thermal response data of the building envelope, and extract temperature response features, heat flow response features, and time delay response features to form corresponding feature sets.

[0012] S3. Establish a partitioned heat transfer mechanism model based on the corresponding feature set, and invert and verify the time-varying heat transfer characteristic parameter set of each partition;

[0013] S4. Based on the time-varying heat transfer characteristic parameter set and adjustable parameter boundary, use imitation learning to generate parameter search prior strategy and candidate optimization parameter set, and obtain the predicted heat transfer performance results corresponding to each candidate optimization parameter set.

[0014] S5. Based on the predicted heat transfer performance results, construct the heat transfer performance objective function and perform iterative solution to output the optimized heat transfer performance parameters of the target building envelope.

[0015] S1 specifically includes: acquiring the initial envelope information of the target building envelope; dividing the target building envelope into zones according to the envelope structure type, orientation, floor level, and window type; establishing envelope zoning information; constructing a multi-condition thermal excitation sequence for each zone by combining meteorological forecast data, building operation schedule, and indoor equipment operation records; the multi-condition thermal excitation sequence includes at least shading opening and closing conditions, ventilation start and stop conditions, and air conditioning start and stop conditions; and simultaneously collecting envelope thermal response data and boundary input data for each zone under the action of the multi-condition thermal excitation sequence. The envelope thermal response data includes at least internal surface temperature data, external surface temperature data, indoor air temperature data, outdoor air temperature data, heat flux density data, and condition execution status data. The boundary input data includes at least one of the following: outdoor solar irradiance data, outdoor wind speed data, indoor-outdoor pressure difference data, and window opening status data.

[0016] S2 specifically includes: performing time alignment, missing data completion, and anomaly removal on the thermal response data of the enclosure to obtain thermal response data for cleaning the enclosure; segmenting the thermal response data of cleaning the enclosure according to the start and end times of the working conditions in the multi-working-condition thermal excitation sequence to obtain sub-working-condition response data; extracting temperature response features, heat flux response features, and time delay response features based on the sub-working-condition response data, wherein the temperature response features include at least the temperature rise amplitude and temperature fall amplitude, the heat flux response features include at least the heat flux peak value and the heat flux change slope, and the time delay response features include at least the thermal response delay time; and binding each extracted response feature with the partition number, working condition type identifier, and time period identifier to form a corresponding feature set.

[0017] S3 specifically includes: establishing a zoned heat transfer mechanism model based on the corresponding feature set, dividing the thermal response of the building envelope into four mechanism components: conduction effect, heat storage effect, wind infiltration disturbance effect, and solar heat gain effect; performing window-by-window inversion on each zone according to the preset inversion window to obtain the equivalent heat transfer parameters, equivalent heat storage parameters, wind infiltration disturbance parameters, and solar heat gain correction parameters for each zone, and forming a time-varying heat transfer characteristic parameter sequence; substituting the time-varying heat transfer characteristic parameter sequence back into the zoned heat transfer mechanism model, calculating the model reconstruction response result, and comparing the deviation between the model reconstruction response result and the measured response result; when the deviation is not greater than the preset inversion error threshold, solidifying the corresponding time-varying heat transfer characteristic parameter sequence into a time-varying heat transfer characteristic parameter set.

[0018] S4 specifically includes: generating preset optimized boundary conditions based on the modification conditions of the target building envelope, the preset optimized boundary conditions including at least the adjustable range of the envelope structure, the range of material replacement, the feasible range of construction, and the range of engineering resource constraints, and forming adjustable parameter boundaries; based on the time-varying heat transfer characteristic parameter set, the adjustable parameter boundaries, and the current performance gap, calling the imitation learning model to generate a parameter search prior strategy, the parameter search prior strategy including at least the parameter items to be adjusted in each partition, the priority adjustment direction, the priority adjustment range, and the adjustment order; generating an initial candidate optimized parameter set according to the parameter search prior strategy, and performing parameter perturbation expansion on the initial candidate optimized parameter set within the adjustable parameter boundaries to form a candidate optimized parameter set; inputting each candidate optimized parameter set into the envelope heat transfer prediction model to obtain the corresponding predicted heat transfer performance results.

[0019] S5 specifically includes: constructing a heat transfer performance objective function based on the predicted heat transfer performance results. The heat transfer performance objective function includes at least the building energy consumption evaluation value, the indoor thermal stability deviation evaluation value, and the evaluation value of the building envelope parameter adjustment engineering quantity. After normalizing the building energy consumption evaluation value, the indoor thermal stability deviation evaluation value, and the evaluation value of the building envelope parameter adjustment engineering quantity, the heat transfer performance objective function value corresponding to each candidate optimization parameter set is calculated. The candidate optimization parameter set is iteratively screened according to the heat transfer performance objective function value, and the target parameter combination is output when the preset convergence condition is met. The target parameter combination is mapped back to each partition according to the partition number to obtain the optimized heat transfer performance parameters of the target building envelope.

[0020] The beneficial effects of this invention are as follows:

[0021] This invention constructs a multi-condition thermal excitation sequence and collects building envelope thermal response data and boundary input data, which can more realistically characterize the dynamic heat transfer state of the target building envelope under operating conditions such as shading opening and closing, ventilation starting and stopping, and air conditioning starting and stopping, thereby improving the accuracy of heat transfer performance parameter optimization.

[0022] This invention extracts temperature response, heat flow, and time-delay response features from building envelope thermal response data, and uses these features to invert and verify a time-varying heat transfer characteristic parameter set. This effectively distinguishes the effects of conduction, heat storage, wind infiltration disturbance, and solar heat gain, improving the accuracy and interpretability of building envelope thermal characteristics. Furthermore, by establishing separate heat transfer mechanism models for different building envelope zones and optimizing them using a time-varying heat transfer characteristic parameter set, this invention avoids the coarse-grained optimization problems caused by using overall static parameters, thereby improving the matching degree between the optimized building envelope parameters and actual building conditions.

[0023] This invention utilizes imitation learning to generate a priori strategy for parameter search and forms a candidate optimization parameter set within adjustable parameter boundaries. This reduces the invalid search range, improves the efficiency of candidate solution generation, and thus enhances the optimization solution speed and the stability of the optimal results. By constructing a heat transfer performance objective function that includes building energy consumption, indoor thermal stability deviation, and the cost of adjusting building envelope parameters, and performing iterative solutions, a comprehensive balance can be achieved between energy-saving effects, thermal environment improvement effects, and engineering implementation resource consumption, thereby increasing the engineering application value of optimized heat transfer performance parameters. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a method for optimizing the heat transfer performance parameters of a building envelope according to the present invention;

[0025] Figure 2 This is a schematic diagram of the enclosure thermal response data acquisition and control interface in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of the multi-condition thermal excitation condition sequence configuration interface in an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram illustrating the real-time trend of thermal response data of the building envelope in an embodiment of the present invention.

[0028] Figure 5 This is a schematic diagram of the configuration interface for the enclosure thermal response feature extraction algorithm in an embodiment of the present invention.

[0029] Figure 6 This is a schematic diagram of the heat transfer parameter inversion execution control interface in an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Example 1: As Figure 1 As shown in the figure, this embodiment provides a method for optimizing the heat transfer performance parameters of a building envelope, including the following steps:

[0032] S1. Obtain the initial enclosure information of the target building enclosure structure, divide the target building enclosure structure into zones, construct a multi-condition thermal excitation condition sequence corresponding to each zone, and collect enclosure thermal response data and boundary input data.

[0033] S2. Perform time alignment, anomaly removal, and operating condition segmentation on the thermal response data of the building envelope, and extract temperature response features, heat flow response features, and time delay response features to form corresponding feature sets.

[0034] S3. Establish a partitioned heat transfer mechanism model based on the corresponding feature set, and invert and verify the time-varying heat transfer characteristic parameter set of each partition;

[0035] S4. Based on the time-varying heat transfer characteristic parameter set and adjustable parameter boundary, use imitation learning to generate parameter search prior strategy and candidate optimization parameter set, and obtain the predicted heat transfer performance results corresponding to each candidate optimization parameter set.

[0036] S5. Based on the predicted heat transfer performance results, construct the heat transfer performance objective function and perform iterative solution to output the optimized heat transfer performance parameters of the target building envelope.

[0037] S1 specifically includes the following sub-steps:

[0038] S110. Obtain the initial enclosure information of the target building's enclosure structure. The initial enclosure information shall include at least: building floor plan, elevation plan, enclosure construction method table, door and window table, site survey record, room function record, and existing operation and maintenance record. Among them, the enclosure construction method table is used to characterize the construction layers of the exterior walls, roof, and exterior windows; the door and window table is used to characterize the exterior window type, opening form, and glass configuration; and the room function record is used to characterize the room usage period, personnel density, and indoor temperature control boundary.

[0039] Based on the initial building envelope information, the target building envelope structure is divided into zones. Zoning is performed using four criteria: envelope structure type, orientation, floor zone, and window type. An envelope unit is assigned to the same envelope zone only if it meets all four criteria; otherwise, it is assigned to different envelope zones. Here, "floor zone" refers to the ground floor zone, middle floor zone, and top floor zone; "window type" refers to consistent window frame material, opening method, and number of glass layers.

[0040] Number each enclosure section as follows ,in, Assign zone numbers. Establish enclosure zone information based on the zone numbers. Enclosure zone information should include at least: zone number, zone location, zone area, zone orientation, zone structural level, and corresponding room number. Output the enclosure zone information with zone numbers for use in step S120.

[0041] In one possible embodiment, the target building envelope can be divided into a south-facing exterior wall zone, a north-facing exterior wall zone, a roof zone, an east-facing exterior window zone, and a west-facing exterior window zone; if the south-facing exterior wall has the same construction method in the middle layer and the top layer, but the top layer is subject to additional thermal disturbance from the roof, then the south-facing exterior wall of the middle layer and the south-facing exterior wall of the top layer are not classified into the same envelope zone.

[0042] S120. Read the building envelope zoning information output in step S110, and combine it with the meteorological forecast data of the target building's location, the building's operating schedule, and the indoor equipment operation records to generate a multi-condition thermal excitation sequence. The multi-condition thermal excitation sequence refers to a set of multiple thermal disturbance conditions arranged in a predetermined order within a preset observation period, used to sequentially excite distinguishable thermal responses caused by changes in shading, ventilation, and air conditioning operation on the same target building envelope.

[0043] The multi-condition thermal excitation sequence includes at least three conditions: shading opening / closing, ventilation start / stop, and air conditioning start / stop. The shading opening / closing condition alters the solar heat gain boundary of the exterior windows, the ventilation start / stop condition alters the indoor / outdoor air exchange boundary, and the air conditioning start / stop condition alters the indoor temperature control boundary. For each condition, a start time, duration, end time, and stable transition period between adjacent conditions are defined. The stable transition period is a buffer period maintained after the end of one condition and before the beginning of the next, used to reduce the cumulative residual thermal effects between different conditions.

[0044] A working condition coding sequence is established for each enclosure zone. The working condition coding sequence is used to record the working condition type and working condition status of the enclosure zone in each observation period. The generated working condition coding sequence is bound one-to-one with the zone number to form a multi-working condition thermal excitation working condition sequence. The multi-working condition thermal excitation working condition sequence associated with each zone number is output for use in steps S130 and S220.

[0045] In one possible embodiment, the preset observation period can be set to 24 consecutive hours; the shading start-stop condition can last for 2 hours, the ventilation start-stop condition can last for 2 hours, and the air conditioning start-stop condition can last for 3 hours, with a 30-minute stable transition period between adjacent conditions. If the duration of a certain condition is insufficient to cause a significant change in the thermal response of the building envelope, the duration of that condition is extended until stable temperature response characteristics, heat flow response characteristics, and time delay response characteristics can be extracted in subsequent steps.

[0046] S130. Under the action of the multi-condition thermal excitation sequence generated in step S120, thermal response data of the building envelope is synchronously collected for each enclosure zone. The thermal response data of the building envelope includes at least: internal surface temperature data, external surface temperature data, indoor air temperature data, outdoor air temperature data, heat flux density data, and operating condition execution status data.

[0047] To support the subsequent inversion of solar heat gain correction parameters and infiltration disturbance parameters, while collecting the aforementioned building envelope thermal response data, boundary input data corresponding to the current operating conditions are also collected or retrieved simultaneously. The boundary input data includes at least one of the following: outdoor solar irradiance data, outdoor wind speed data, indoor-outdoor pressure difference data, and window opening status data; wherein, outdoor solar irradiance data is used to characterize the solar heat gain boundary, and outdoor wind speed data, indoor-outdoor pressure difference data, and window opening status data are used to characterize the infiltration boundary under ventilation start-stop conditions.

[0048] At least one set of corresponding sampling points should be arranged for each building envelope zone. These sampling points include: an inner surface temperature sampling point, an outer surface temperature sampling point, and a heat flux density sampling point within the same building envelope zone. The inner and outer surface temperature sampling points are arranged at corresponding positions on the inner and outer sides of the same building envelope component, and the heat flux density sampling point is arranged near these corresponding positions to ensure that the collected temperature and heat flux responses originate from the same local building envelope path. Indoor air temperature is collected at a representative location within the corresponding room, while outdoor air temperature, outdoor solar irradiance, and outdoor wind speed are collected at unobstructed locations outside the building. Indoor and outdoor pressure differences can be obtained through corresponding differential pressure sensors, and window opening status can be obtained through door and window status sensors or control logs.

[0049] Each sampling point is sampled according to a unified sampling period and written to the same timestamp buffer. If multiple sampling points of the same type are set for a certain enclosure zone, the sampled values ​​of the same type of sampling points within the enclosure zone are first averaged, and then used as the partition response value of the enclosure zone at the corresponding sampling time. During sampling, the partition number, working condition type, working condition status, boundary input data, and timestamp are synchronously written to the buffer to form traceable original enclosure thermal response data. The original enclosure thermal response data with working condition tags and partition numbers is output for use in step S210.

[0050] In one possible embodiment, the uniform sampling period can be set to 5 minutes; each original thermal response data of the building envelope includes at least: zone number, timestamp, inner surface temperature value, outer surface temperature value, indoor air temperature value, outdoor air temperature value, heat flux density value, outdoor solar irradiance value, and current operating status; under ventilation start-stop conditions, outdoor wind speed value and window opening status value can also be recorded simultaneously.

[0051] S2 specifically includes the following sub-steps:

[0052] S210. Read the raw thermal response data of the enclosure output in step S130 and perform time alignment processing on the data from different sources. During time alignment, a unified time grid is established based on a unified sampling period, and the data from different collection points are resampled onto the unified time grid. If there are multiple sampled values ​​in the same time grid, their average value is taken. If there are no sampled values ​​in a certain time grid, if the missing length does not exceed the preset completion length, linear interpolation between adjacent time points is used for completion. If the continuous missing length exceeds the preset completion length, the time period is marked as an invalid segment and will not participate in feature extraction in subsequent steps.

[0053] Anomaly removal is performed on the time-aligned data. During anomaly removal, the median of adjacent sampled values ​​before and after a certain sampling time is used as a reference value; if the absolute value of the difference between the current sampled value and the reference value exceeds the anomaly threshold corresponding to that sample size, the current sampled value is marked as an anomaly. For single-point anomalies, interpolation with adjacent valid sampled values ​​is used for replacement; for continuous anomaly segments, they are directly marked as invalid segments and are not included in subsequent operating condition segments.

[0054] In one embodiment, the anomaly threshold for the inner surface temperature data can be set to 2 degrees Celsius, and the anomaly threshold for the heat flux density data can be set to 8 watts per square meter; data segments with consecutive missing lengths exceeding 3 sampling periods can be directly marked as invalid segments. After completing time alignment, missing data processing, and anomaly removal, the cleaning enclosure thermal response data is obtained. The cleaning enclosure thermal response data is output for use in step S220.

[0055] S220: Read the working condition coding sequence from step S120 and the cleaning and enclosure thermal response data output from step S210. Using the start and end times of each working condition in the working condition coding sequence as segmentation boundaries, segment the cleaning and enclosure thermal response data into working condition segments. Each segment includes at least three parts: a baseline period before the working condition begins, a working condition action period, and a recovery period after the working condition ends. The baseline period characterizes the initial thermal state before the working condition switch, the working condition action period characterizes the main thermal response under the working condition disturbance input, and the recovery period characterizes the thermal recovery process after the working condition ends.

[0056] Each segmented data point is assigned a partition number, operating condition type identifier, and time period identifier to form sub-operating condition response data. The time period identifier uniquely identifies a complete response segment corresponding to a specific enclosure zone under a specific operating condition. Each subsequently extracted response feature is traced back to the original data source through this time period identifier. If the number of valid sampling points for a sub-operating condition response data is less than the preset minimum number of valid points, the sub-operating condition response data is marked as an invalid operating condition sample and will not participate in this round of feature extraction.

[0057] In one possible embodiment, the 30 minutes before the start of the operating condition can be used as the baseline period, the entire duration of the operating condition can be used as the operating condition action period, and the 60 minutes after the end of the operating condition can be used as the recovery period. The sub-operating condition response data is output for use in step S230.

[0058] S230. Read the sub-condition response data output in step S220. Extract temperature response features, heat flow response features, and time delay response features for each sub-condition response data. Store the extraction results together with the partition number, condition type identifier, and time period identifier to form a corresponding feature set.

[0059] For the temperature response characteristics, at least the temperature rise and temperature fall values ​​should be extracted. The temperature rise value is calculated using the following formula:

[0060]

[0061] in, This represents the temperature rise amplitude; The peak value of the inner surface temperature during the operating period; This represents the initial temperature of the inner surface at the end of the baseline time period.

[0062] The temperature drop amplitude is calculated using the following formula:

[0063]

[0064] in, This represents the temperature drop. This represents the lowest temperature value of the inner surface during the operating period. The definition is the same as above.

[0065] For the heat flux response characteristics, at least the peak heat flux and the slope of the heat flux change should be extracted. The peak heat flux is denoted as... peak heat flux Defined as the maximum absolute value of heat flux density during the operating period. The slope of the heat flux change is calculated using the following formula:

[0066]

[0067] in, The slope of the heat flux change; The starting calculation time The corresponding heat flux density value; To terminate the calculation time The corresponding heat flux density value; and All sampling times were selected within the operating period, and .

[0068] For time-delay response characteristics, at least the thermal response delay time should be extracted. The thermal response delay time is calculated using the following formula:

[0069]

[0070] in, This refers to the thermal response delay time. This is the moment when the heat flux peak occurs; This is the moment when the current operating condition changes.

[0071] In one possible embodiment, if the inner surface temperature of a south-facing exterior window zone rises from 26.0 degrees Celsius to 28.4 degrees Celsius under the condition of shading being closed, then the temperature rise of that zone under that condition is 2.4 degrees Celsius; if the heat flux peak of that zone under the same condition occurs 35 minutes after the condition switch, then the thermal response delay time of that zone under that condition is 35 minutes.

[0072] After feature extraction is completed, each set of features is bound to its corresponding partition number, operating condition type identifier, and time period identifier to obtain a one-to-one correspondence feature set between the thermal excitation condition and the building envelope thermal response. The corresponding feature set is output for use in step S310.

[0073] S3 specifically includes the following sub-steps:

[0074] S310. Read the corresponding feature set output in step S230 and establish a zoned heat transfer mechanism model for each enclosure zone. The zoned heat transfer mechanism model is used to divide the thermal response of the enclosure into four mechanism components: conduction effect, heat storage effect, wind infiltration disturbance effect, and solar heat gain effect, and to establish the constraint relationship between each mechanism component and the corresponding feature set.

[0075] For exterior wall and roof zones, the effects of conduction and heat storage are constructed primarily using exterior and interior surface temperatures, heat flux density, and thermal response delay time. For exterior window zones, in addition to conduction and heat storage effects, the effects of solar heat gain are constructed by combining changes in temperature rise amplitude and outdoor solar irradiance intensity under shading opening and closing conditions. For the synchronous changes in indoor air temperature, interior surface temperature, outdoor wind speed, indoor-outdoor pressure difference, or window opening status under ventilation start-stop conditions, the effects of wind infiltration disturbance are constructed. Thus, the features extracted in different operating conditions are mapped to interpretable physical influence components.

[0076] For each enclosure zone, the temperature response characteristics, heat flux response characteristics, and time delay response characteristics are bound one by one to the four mechanistic components, forming a characteristic-mechanism correspondence table for that enclosure zone under various operating conditions. The heat transfer mechanism model of the zone and its corresponding characteristic-mechanism correspondence table are output for use in step S320.

[0077] In this embodiment, the partitioned heat transfer mechanism model uses an equivalent thermal resistance-capacitance (RC) network to construct its underlying thermodynamic equations. For a given enclosure partition, the energy balance differential equation for its inner surface nodes is expressed as:

[0078] in, and These are the inner and outer surface temperatures, respectively. Outdoor solar irradiance; and These are outdoor and indoor air temperatures, respectively. The specific heat capacity of air at constant pressure; The term refers to the radiative heat source absorbed by the inner surface; and These are the time-varying heat transfer characteristic parameters to be inverted.

[0079] S320. Read the partitioned heat transfer mechanism model and feature-mechanism correspondence table output in step S310, and perform window-by-window inversion of the parameters of each enclosure partition under different operating conditions. During window-by-window inversion, each sub-operating condition response data is divided into multiple adjacent inversion windows according to the preset inversion window length. In each inversion window, the equivalent heat transfer parameters, equivalent heat storage parameters, wind infiltration disturbance parameters, and solar heat gain correction parameters of the enclosure partition are solved respectively. Then, the solution results of adjacent inversion windows are concatenated in time order to form the time-varying parameter sequence of the enclosure partition.

[0080] Each enclosure zone at any time The parameters are expressed as follows:

[0081]

[0082] in, The partition number is Enclosure partitions at all times The time-varying heat transfer characteristic parameter vector; For this enclosure zone at time The equivalent heat transfer parameters; For this enclosure zone at time The equivalent heat storage parameters; For this enclosure zone at time The parameters of air seepage disturbance; For this enclosure zone at time The solar heat gain correction parameters This refers to a time variable or the corresponding sampling time.

[0083] To ensure the reproducibility of the inversion process, within each inversion window, the parameters are updated iteratively based on a weighted combination of the deviations between the model-reconstructed inner surface temperature response and the measured inner surface temperature response, as well as the deviations between the model-reconstructed heat flux response and the measured heat flux response. , , and This continues until the overall deviation of the current inversion window meets the preset window error condition. Here, "weighted combination" means: setting the weights for temperature response deviation and heat flow response deviation according to the current operating condition type, and then forming the overall deviation index for the current window accordingly.

[0084] To ensure continuity between adjacent windows, the initial parameter values ​​are preferentially adopted from the solution results of the previous adjacent inversion window; when there is no result from the previous adjacent inversion window, the nominal thermal parameters of the corresponding structural method for this enclosure zone are used as the initial values. The nominal thermal parameters refer to the initial parameters obtained based on the enclosure structural method table, the nominal thermal conductivity of the material, the nominal thickness, and the empirical values ​​of the standard construction.

[0085] In one possible embodiment, the preset inversion window length can be set to 60 minutes, and the overlap length of adjacent inversion windows can be set to 30 minutes. If a certain external window zone exhibits a larger temperature rise and higher outdoor solar irradiance response under shading-off conditions, the solar heat gain correction parameter obtained in that condition window will be larger than that under shading-on conditions; if a certain external wall zone experiences a rapid change in indoor air temperature and a simultaneous increase in indoor and outdoor pressure difference when switching between ventilation start-stop conditions, the air infiltration disturbance parameter obtained in that condition window will be correspondingly larger.

[0086] After solving the parameters for all inversion windows, the parameter results for the same enclosure zone in adjacent inversion windows are processed to be continuous. For parameter points whose abrupt changes exceed the preset smoothing threshold, it is checked whether there are invalid samples or abnormal fluctuations in their corresponding operating conditions. If so, the results of adjacent valid windows are used for correction. The time-varying heat transfer characteristic parameter sequence of each enclosure zone is output for use in step S330.

[0087] S330. Read the time-varying heat transfer characteristic parameter sequence output in step S320, and resubmit it into the partitioned heat transfer mechanism model established in step S310 to calculate the model reconstruction response result. The model reconstruction response result includes at least: the model reconstruction inner surface temperature response value and the model reconstruction heat flux response value. To avoid the influence of different dimensions of temperature and heat flux on error evaluation, the model reconstruction response value and the corresponding measured response value are first normalized within their respective operating conditions, and then the deviation is compared.

[0088] Inversion error is calculated using the following formula:

[0089]

[0090] in, This is the inversion error; The total number of data points participating in the comparison; For the first A normalized model reconstructs the response value; For the first A normalized measured response value, For the data point index number, and .

[0091] Compare the inversion error with the preset inversion error threshold. In comparison, among which, This is a preset inversion error threshold. When... When, the corresponding time-varying heat transfer characteristic parameter sequence is solidified into a time-varying heat transfer characteristic parameter set; when If the parameter inversion is not performed, return to step S320 to re-execute the parameter inversion, and prioritize adjusting the initial parameter values, window boundaries, or feature weights of the current operating condition window.

[0092] In one possible embodiment, It can be set to 0.08; if the normalized comprehensive response error of a certain enclosure zone is 0.05, then the parameter inversion result of that enclosure zone is considered to have passed the verification. After completing the verification of all enclosure zones, the set of time-varying heat transfer characteristic parameters that have passed the verification is output for use by steps S410 and S420.

[0093] S4 specifically includes the following sub-steps:

[0094] S410. Read the time-varying heat transfer characteristic parameter set output in step S330, and generate preset optimized boundary conditions based on the renovation conditions of the target building. The preset optimized boundary conditions include at least: the adjustable range of the building envelope, the range of material replacement, the feasible range of construction, and the range of engineering resource constraints.

[0095] Among them, the adjustable range of the enclosure structure is used to limit the types of parameters that can be adjusted in each enclosure zone and their upper and lower limits; the material replacement range is used to limit the types of replaceable materials and their thermal parameter ranges; the feasible construction range is used to limit the constructable thickness, replaceable parts and non-modifiable nodes.

[0096] The engineering resource constraint range is used to limit the maximum allowable increase in material per unit area or the upper limit of the total amount of renovation work. Preset optimized boundary conditions are associated with the time-varying heat transfer characteristic parameter sets of each enclosure zone according to the zone number, forming an adjustable parameter boundary. The adjustable parameter boundary is output for use in steps S420 and S430.

[0097] In one embodiment, the thickness of the insulation layer in the exterior wall partition can be adjusted from 20 mm to 80 mm, the heat transfer performance level of the glass in the exterior window partition can be limited to between double-glazed and triple-glazed glass, and the maximum allowable material increment per unit area can be set to 50 kg per square meter.

[0098] S420: Read the time-varying heat transfer characteristic parameter set output in step S330 and the adjustable parameter boundary output in step S410, and call the imitation learning model to generate a parameter search prior strategy. Here, the "imitation learning model" refers to a model trained by learning the mapping relationship between "state input" and "optimization adjustment action" in the historical optimal enclosure parameter adjustment trajectory; the "historical optimal enclosure parameter adjustment trajectory" refers to the multi-round parameter adjustment sequence and its corresponding performance results in historical building enclosure optimization cases, which have been solved through simulation and verified by technical personnel.

[0099] To ensure the reproducibility of training samples, each historically preferred envelope parameter adjustment trajectory includes at least: an initial time-varying heat transfer characteristic parameter set, adjustable parameter boundaries, multiple rounds of parameter adjustment actions, predicted heat transfer performance results for each round, and a final objective function evaluation result. Among these, the multiple rounds of parameter adjustment actions characterize the direction, magnitude, and order of parameter adjustments performed on each envelope zone in each iteration round; the predicted heat transfer performance results characterize the building energy consumption, indoor thermal stability deviation, and renovation workload obtained after the parameter adjustment in that round; and the final objective function evaluation result characterizes whether the trajectory belongs to the historically preferred trajectory.

[0100] To ensure that the current target building envelope can be directly mapped to historical optimal behavior, a current performance gap needs to be constructed. The current performance gap refers to the set of differences between the current scheme's building energy consumption evaluation value, indoor thermal stability deviation evaluation value, and envelope parameter adjustment quantity evaluation value and the target evaluation threshold; where the target evaluation threshold consists of preset energy-saving targets, preset thermal stability targets, and preset upper limits for implementation quantities. The current performance gap, along with the time-varying heat transfer characteristic parameter set and adjustable parameter boundaries of the current target building envelope, are used as the state input for the current round.

[0101] During the model training phase, the state input and preferred adjustment action are extracted for each historical optimal building envelope parameter adjustment trajectory, and the mapping relationship between the state input and the preferred adjustment action is learned. During the model invocation phase, the time-varying heat transfer characteristic parameter set of the current target building envelope, the adjustable parameter boundary, and the current performance gap are input into the model to simulate learning, and the parameter search prior strategy is output. The parameter search prior strategy includes at least: the parameter items to be adjusted first in each envelope zone, the first adjustment direction, the first adjustment range, and the order of adjustment between envelope zones.

[0102] The prior strategy for parameter search is used to generate an initial set of candidate optimization parameters. This initial set of candidate optimization parameters refers to one or more combinations of parameters that preferentially fall within the recommendation direction and recommendation range of the imitation learning model. The prior strategy for parameter search and the initial set of candidate optimization parameters are output for use in step S430.

[0103] In one possible embodiment, if the imitation learning model outputs "prioritize adjusting the solar heat gain parameters of the west-facing window zone, then adjust the insulation thickness parameters of the roof zone, and prioritize adjusting the parameters of the west-facing window zone in the direction of reducing solar heat gain", then the initial candidate optimization parameter set is first generated around the recommended interval of the west-facing window zone, and then around the recommended interval of the roof zone.

[0104] In one specific embodiment, the imitation learning model is constructed using a behavior cloning algorithm based on deep neural networks. The model's network structure includes: an input layer (receiving the concatenated state input vector), several fully connected hidden layers (e.g., 2 or 3 layers, using the ReLU activation function to extract state features), and an output layer (using the Softmax function to output the probability distribution of discrete adjustment actions, or using a linear layer to output the mean and variance of continuous adjustment amplitudes). During model training, the state inputs from historical optimal enclosure parameter adjustment trajectories are used as network inputs, and the corresponding optimal adjustment actions are used as labels. Cross-entropy or mean squared error (MSE) is used as the loss function, and the gradient descent algorithm is used to continuously update the network weights until the model's action prediction accuracy on the validation set meets a preset threshold.

[0105] S430: Read the adjustable parameter boundary output in step S410 and the parameter search prior strategy and initial candidate optimization parameter set output in step S420. Perform parameter perturbation expansion on the initial candidate optimization parameter set to form a candidate optimization parameter set. During parameter perturbation expansion, at least a first preset proportion of candidate solutions fall within the priority adjustment direction and priority adjustment range indicated by the parameter search prior strategy. The remaining candidate solutions are explored within the boundary neighborhood allowed by the adjustable parameter boundary to avoid missing potential better solutions.

[0106] For each set of candidate optimization parameters, it is mapped to the corresponding building envelope zone according to the zone number and input into the building envelope heat transfer prediction model to obtain the predicted heat transfer performance results. The building envelope heat transfer prediction model is a prediction model used to output heat transfer results and indoor thermal environment results under each working condition after inputting the current parameters of the target building envelope, candidate optimization parameters, multi-condition thermal excitation sequence, and corresponding boundary conditions. The predicted heat transfer performance results include at least: cumulative heat transfer load, indoor air temperature fluctuation results, and corresponding renovation work quantities.

[0107] In one possible embodiment, if the initial candidate optimization parameter set contains 10 parameter combinations, then 5 local perturbation schemes can be preferentially expanded near each parameter combination to form 60 candidate optimization parameter sets. The first preset ratio can be set to 70%, meaning that at least 42 candidate schemes must fall within the priority direction and priority range indicated by the parameter search prior strategy. If a candidate optimization parameter set exceeds the adjustable parameter boundary, it is directly eliminated and not included in the enclosure heat transfer prediction model. After predicting all candidate optimization parameter sets, the predicted heat transfer performance results corresponding one-to-one with each candidate optimization parameter set are output for use in step S510.

[0108] Furthermore, the building envelope heat transfer prediction model uses a building thermal dynamics simulation engine (such as EnergyPlus, DeST, or TRNSYS) as its underlying computational kernel. The candidate optimization parameter set is converted into a model parameter file (such as an EnergyPlus idf file) recognizable by the simulation engine, and combined with a multi-condition thermal excitation sequence as the operating weather and control boundaries, the simulation engine's transient solver is invoked to perform hourly iterative calculations, thereby outputting high-precision predicted heat transfer performance results.

[0109] S5 specifically includes the following sub-steps:

[0110] S510. Read the predicted heat transfer performance results output in step S430 and construct the heat transfer performance objective function. The heat transfer performance objective function is used to unify building energy consumption, indoor thermal stability deviation, and envelope parameter adjustment work within the same evaluation framework.

[0111] The objective function for heat transfer performance is constructed as follows:

[0112]

[0113] in, The objective function value for heat transfer performance; This is the building energy consumption weighting coefficient; This is the weighting coefficient for indoor thermal stability deviation; Adjust the weighting coefficients for the quantity evaluation of the retaining structure parameters; This is the building energy consumption evaluation value; This is the evaluation value for indoor thermal stability deviation; Adjust the engineering quantity evaluation value for the retaining structure parameters.

[0114] To ensure that the above three evaluation values ​​can be weighted and summed on the same numerical scale, before calculating the objective function value of heat transfer performance, the following steps are taken: , and Each evaluation value is normalized to the same numerical scale according to a preset evaluation interval. Specifically, for any evaluation value... Normalization is performed using the following formula:

[0115]

[0116] in, The normalized evaluation value; These are the original evaluation values ​​to be normalized; This is the lower limit of the preset evaluation interval corresponding to the evaluation indicator; This represents the upper limit of the preset evaluation range corresponding to this evaluation indicator. In this step, the building energy consumption evaluation value... Indoor thermal stability deviation evaluation value Evaluation value of the quantity of retaining wall parameter adjustment project All samples were normalized using the above method before being substituted into the heat transfer performance objective function.

[0117] Among them, the building energy consumption evaluation value is obtained from the cumulative heat transfer load or cumulative air conditioning demand under the multi-condition thermal excitation sequence; the indoor thermal stability deviation evaluation value is obtained by calculating the absolute deviation of the indoor air temperature relative to the center temperature of the target stable range at each sampling time, and then averaging the absolute deviation over the entire prediction period; the evaluation value of the building envelope parameter adjustment engineering quantity is obtained by statistically analyzing the amount of renovation materials consumed, construction hours and component replacement engineering quantity of each building envelope zone, and then normalizing the data.

[0118] If a candidate set of optimization parameters does not meet the preset comfort lower limit or construction feasibility constraints, its heat transfer performance objective function value will be directly assigned an inferior value and it will not participate in subsequent optimization.

[0119] In one possible embodiment, , and These values ​​can be set to 0.5, 0.3, and 0.2 respectively. After calculating the objective function for all candidate optimization parameter sets, the evaluation results of the objective function corresponding to each candidate optimization parameter set are output for use in step S520.

[0120] S520. Read the objective function evaluation results output in step S510 and iteratively filter the candidate optimization parameter set. During the filtering, first sort the parameters from best to worst according to the objective function values ​​of heat transfer performance, and then retain the top few groups as the preferred parameter combinations for the current iteration round; the retained preferred parameter combinations are returned to step S430 to generate the locally expanded candidate set for the next iteration round.

[0121] The iteration stops when the changes in the objective function value of heat transfer performance in two consecutive iterations satisfy the convergence condition. The convergence condition is determined by the following formula:

[0122]

[0123] in, For the first The objective function value of optimal heat transfer performance in round iteration; For the first The objective function value of optimal heat transfer performance in round iteration; The preset convergence threshold is used; This is the current iteration round number.

[0124] When the above inequality holds, the optimal parameter combination corresponding to the current round is output as the target parameter combination; when the above inequality does not hold, the optimal parameter combination of the current round is returned to step S430 to continue generating the candidate optimization parameter set for the next round and performing prediction and evaluation.

[0125] In one possible embodiment, The value can be set to 0.01; if the objective function values ​​of the optimal heat transfer performance in two consecutive rounds are 0.426 and 0.419 respectively, then the difference is 0.007, which is less than 0.01, indicating that convergence has been achieved. Output the target parameter combination that satisfies the convergence condition for use in step S530.

[0126] S530. Read the target parameter combination output in step S520, map it back to each building envelope zone according to the zone number, and obtain the optimized heat transfer performance parameters of each building envelope zone. The optimized heat transfer performance parameters include at least: the equivalent optimized heat transfer parameters and corresponding recommended insulation thickness of the exterior wall zone, the equivalent optimized heat transfer parameters and corresponding recommended roof insulation configuration of the roof zone, and the equivalent optimized heat transfer parameters and corresponding recommended exterior window configuration or shading configuration of the exterior window zone.

[0127] The optimized heat transfer performance parameters of each building envelope zone are summarized to form the output results of the optimized heat transfer performance parameters of the target building envelope structure. The output results include at least: zone number, final parameter value, corresponding applicable operating condition range, predicted energy saving improvement, predicted indoor thermal stability improvement, and corresponding renovation work volume. Here, "applicable operating condition range" refers to the set of operating conditions in which a target parameter combination meets the preset comfort constraints and construction feasibility constraints in all operating condition segments of a multi-condition thermal excitation sequence.

[0128] In one possible embodiment, the output results can show that: the west-facing window zone prioritizes the use of window configurations that reduce solar heat gain, and the roof zone prioritizes the addition of a certain thickness of insulation layer; after optimization, the cumulative heat transfer load decreases by 12%, the indoor thermal stability deviation decreases by 18%, and the total amount of renovation work remains within the preset upper limit. The final output is the optimized heat transfer performance parameters of the target building envelope, which serves as the final result of this embodiment.

[0129] All the above formulas are performed using dimensionless numerical calculations; the relevant formulas are based on empirical models that approximate the real situation, obtained through extensive data collection and software simulation fitting. The preset parameters and thresholds involved in the formulas can be conventionally set and adjusted by those skilled in the art according to the physical constraints of the actual application scenario.

[0130] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0131] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0132] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0134] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0136] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0137] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0138] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing the heat transfer performance parameters of a building envelope, characterized in that, Includes the following steps: S1. Obtain the initial enclosure information of the target building enclosure structure, divide the target building enclosure structure into zones, construct a multi-condition thermal excitation condition sequence corresponding to each zone, and collect enclosure thermal response data and boundary input data. S2. Perform time alignment, anomaly removal, and operating condition segmentation on the thermal response data of the building envelope, and extract temperature response features, heat flow response features, and time delay response features to form corresponding feature sets. S3. Establish a partitioned heat transfer mechanism model based on the corresponding feature set, and invert and verify the time-varying heat transfer characteristic parameter set of each partition; S4. Based on the time-varying heat transfer characteristic parameter set and adjustable parameter boundary, use imitation learning to generate parameter search prior strategy and candidate optimization parameter set, and obtain the predicted heat transfer performance results corresponding to each candidate optimization parameter set.

2. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 1, characterized in that, S5. Based on the predicted heat transfer performance results, construct the heat transfer performance objective function and perform iterative solution to output the optimized heat transfer performance parameters of the target building envelope.

3. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 1, characterized in that, S1 specifically includes: Obtain the initial enclosure information of the target building's enclosure structure, and divide the target building's enclosure structure into zones according to the enclosure structure type, orientation, floor interval, and window type to establish enclosure zoning information; Combining meteorological forecast data, building operation schedules, and indoor equipment operation records, a multi-condition thermal excitation sequence is constructed for each zone. The multi-condition thermal excitation sequence includes at least the shading opening and closing condition, the ventilation start and stop condition, and the air conditioning start and stop condition. Under the action of a multi-condition thermal excitation sequence, the thermal response data of the enclosure and the boundary input data of each zone are collected synchronously.

4. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 3, characterized in that, The boundary input data shall include at least one of the following: outdoor solar irradiance data, outdoor wind speed data, indoor-outdoor pressure difference data, and window opening status data; Envelope thermal response data should include at least the following: internal surface temperature data, external surface temperature data, indoor air temperature data, outdoor air temperature data, heat flux density data, and operating condition status data.

5. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 1, characterized in that, S2 specifically includes: Time alignment, missing data completion, and anomaly removal are performed on the thermal response data of the enclosure to obtain thermal response data of the cleaning enclosure. Based on the start and end times of the working conditions in the multi-working-condition thermal excitation working condition sequence, the working condition segmentation is performed on the thermal response data of the cleaning enclosure to obtain the sub-working-condition response data. Temperature response features, heat flow response features, and time delay response features are extracted from the response data under different operating conditions. Each extracted response feature is then bound to the partition number, operating condition type identifier, and time period identifier to form a corresponding feature set.

6. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 5, characterized in that, Temperature response characteristics include at least the temperature rise and temperature fall amplitudes; heat flux response characteristics include at least the heat flux peak value and the heat flux change slope; and time delay response characteristics include at least the heat response delay time.

7. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 1, characterized in that, S3 specifically includes: Based on the corresponding feature set, a zoned heat transfer mechanism model is established, and the thermal response of the building envelope is divided into four mechanism components: conduction effect, heat storage effect, wind infiltration disturbance effect, and solar heat gain effect. Perform window-by-window inversion on each partition according to the preset inversion window to obtain the equivalent heat transfer parameters, equivalent heat storage parameters, wind infiltration disturbance parameters and solar heat gain correction parameters of each partition, and form a time-varying heat transfer characteristic parameter sequence. The time-varying heat transfer characteristic parameter sequence is resubmitted into the partitioned heat transfer mechanism model, the model reconstruction response result is calculated, and the deviation between the model reconstruction response result and the measured response result is compared; when the deviation is not greater than the preset inversion error threshold, the corresponding time-varying heat transfer characteristic parameter sequence is solidified into a time-varying heat transfer characteristic parameter set.

8. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 1, characterized in that, S4 specifically includes: Based on the modification conditions of the target building envelope, preset optimization boundary conditions are generated. The preset optimization boundary conditions include at least the adjustable range of the envelope structure, the range of material replacement, the range of construction feasibility, and the range of engineering resource constraints, and form adjustable parameter boundaries. Based on the time-varying heat transfer characteristic parameter set, adjustable parameter boundaries, and current performance gap, an imitation learning model is invoked to generate a parameter search prior strategy. The parameter search prior strategy includes at least the parameter items to be adjusted first in each partition, the direction of adjustment first, the range of adjustment first, and the order of adjustment.

9. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 8, characterized in that, Also includes: An initial set of candidate optimization parameters is generated based on the parameter search prior strategy. Within the adjustable parameter boundary, the initial set of candidate optimization parameters is expanded by parameter perturbation to form a new set of candidate optimization parameters. Each set of candidate optimization parameters is then input into the building envelope heat transfer prediction model to obtain the corresponding predicted heat transfer performance results.

10. The method for optimizing the heat transfer performance parameters of a building envelope according to claim 2, characterized in that, S5 specifically includes: Based on the predicted heat transfer performance results, a heat transfer performance objective function is constructed. The heat transfer performance objective function includes at least the building energy consumption evaluation value, the indoor thermal stability deviation evaluation value, and the evaluation value of the building envelope parameter adjustment engineering quantity. After normalizing the building energy consumption evaluation value, indoor thermal stability deviation evaluation value, and building envelope parameter adjustment engineering quantity evaluation value, the heat transfer performance objective function value corresponding to each candidate optimization parameter set is calculated. The candidate optimization parameter set is iteratively screened based on the objective function value of heat transfer performance, and the target parameter combination is output when the preset convergence condition is met. The target parameter combination is mapped back to each partition according to the partition number to obtain the optimized heat transfer performance parameters of the target building envelope.

Images