A photovoltaic integrated curtain wall photovoltaic energy production and cold and heat load dynamic prediction method

Abstract

The present application relates to a kind of photovoltaic integrated curtain photovoltaic energy and cold and heat load dynamic prediction method, comprising the following steps: constructing curtain flow heat transfer fast prediction model;The heat-gas flow network of curtain is constructed;Input meteorological time series data;Obtain the full time series data of local convective heat transfer coefficient and facade wind pressure distribution of outer surface;Obtain the local convective heat transfer coefficient of cavity wall surface at next moment;Solve curtain non-uniform heat field, obtain component cell, wall and cavity air temperature distribution;Obtain photovoltaic array power generation and cumulative power generation;Calculate the additional cold and heat load caused by curtain system and air conditioning energy conversion;Obtain the dynamic distribution of wall surface temperature and component power generation at different time, complete photovoltaic energy and cold and heat load full time series dynamic prediction calculation.The present application is based on the dynamic coupling calculation of component temperature and power generation, realizes the full time series collaborative prediction of photovoltaic energy and cold and heat load, considers simulation accuracy and calculation efficiency, solves the limitation of traditional simulation model.

Description

Technical Field

[0001] This invention relates to a method for predicting building energy consumption, specifically a method for dynamically predicting photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain walls, belonging to the field of building energy consumption prediction and artificial intelligence technology. Background Technology

[0002] Building integrated photovoltaics (BIPV), as an important technological path to achieve building energy self-sufficiency and low-carbon operation, is gradually moving from conceptual demonstrations to large-scale applications. Among them, BIPV double-skin facades (BIPV-DSF) have become one of the key technologies for improving overall energy efficiency in modern public buildings and high-rise office buildings due to their multiple advantages of combining the functions of building envelope, natural ventilation potential, and photovoltaic power generation capabilities.

[0003] However, driven by wind pressure and thermal pressure, the airflow in the BIPV-DSF cavity is constrained by the geometric configuration, resulting in a non-uniform heat-flow field. This leads to significant temperature gradients in both the photovoltaic modules and the interior of the wall, which not only affects the power generation efficiency of the modules and the building's heating and cooling loads, but may also cause local overheating risks under extreme weather conditions, threatening the safety and durability of the system.

[0004] Currently, existing technologies lack applicable and effective tools for predicting the true thermal field inside BIPV-DSFs. This leads to the use of theoretical methods based on homogenization assumptions or simplified steady-state conditions for performance evaluation and system design in engineering practice, which fails to accurately reflect the transient response characteristics and spatial non-uniformity effects under actual meteorological disturbances. Although some research has attempted to develop high-resolution physical models to improve prediction accuracy, such as refined frameworks that couple computational fluid dynamics and building energy consumption simulation, the core bottleneck of such methods lies in the fact that computational costs increase non-linearly with geometric complexity, the number of meteorological conditions, and time scales. This makes it difficult to support the multi-condition year-round dynamic simulation required in the design phase and the large-scale performance evaluation and optimization for diverse facade configurations.

[0005] The aforementioned limitations lead to amplified prediction deviations in key parameters such as cavity microclimate, component operating temperature, power generation, and building envelope thermal load during system coupling. This not only reduces the reliability of comprehensive performance evaluation but also severely restricts the multi-objective collaborative optimization and scientific decision-making of BIPV-DSF systems in practical engineering.

[0006] Therefore, in order to solve the above-mentioned technical problems, it is indeed necessary to provide an innovative method for dynamically predicting photovoltaic capacity and heating / cooling load of integrated photovoltaic curtain walls, so as to overcome the defects in the existing technology. Summary of the Invention

[0007] The purpose of this invention is to provide a dynamic prediction method for photovoltaic power generation and heating / cooling load of integrated photovoltaic curtain walls. By integrating a data-driven convection heat transfer model with a mechanism-driven heat-airflow network and a single-diode electrical model, a dynamic coupled prediction model of multiple physical processes of electricity, heat, and flow is constructed. This method overcomes the limitations of traditional homogenization and steady-state assumptions in the performance prediction of complex building photovoltaic systems, and achieves full-time-series coordinated prediction from component-level wall temperature and power generation to system-level photovoltaic power generation and heating / cooling load, thereby providing strong guidance for the design optimization of integrated photovoltaic curtain walls.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall, which includes the following process steps: 1) Construct a rapid prediction model for convective heat transfer in the curtain wall, which includes a photovoltaic array outer surface model and a cavity wall model; 2) Construct a thermal-airflow network for the curtain wall; 3) Input meteorological time series data into the heat-airflow network model constructed in step 2), set indoor temperature boundary conditions, set the initial values ​​of the temperature field and flow field in the computational domain, and determine the time step parameters; 4) Using the photovoltaic array outer surface model constructed in step 1), combined with the meteorological time series data in step 3) and the facade geometric parameters, obtain the full time series data of the local convective heat transfer coefficient and the facade wind pressure distribution of the outer surface. 5) Call the cavity wall model constructed in step 1) to obtain the local convective heat transfer coefficient of the cavity wall at the next moment; 6) Based on the energy conservation equation, and combined with the convective heat transfer coefficients of the photovoltaic array outer surface and cavity wall obtained in steps 4) and 5), the non-uniform thermal field of the curtain wall is solved to obtain the temperature distribution of the module cells, wall and cavity air. 7) Based on the power of a single diode, combined with the solar irradiance conditions and the module cell temperature obtained in step 6), obtain the photovoltaic array power generation and cumulative power generation; 8) Based on the wall temperature obtained in step 6) and the set indoor temperature, calculate the additional heating and cooling loads and air conditioning energy consumption caused by the curtain wall system; 9) Repeat steps 5) to 8) to obtain the dynamic distribution of wall temperature and module power generation at different times, so as to realize the full-time dynamic prediction of photovoltaic power generation and heating and cooling load.

[0009] The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall of the present invention is further described in step 1) as follows: 1-1) The outer surface of the photovoltaic array and the cavity wall are divided into sub-regions; the facade unit consisting of the window opening and its adjacent photovoltaic modules is used as the basic division object. The outer surface of the photovoltaic array is divided into N rows and M columns of facade units according to a unified facade module, forming a total of N×M facade units, where N and M are both positive integers, such as Figure 2 As shown; each facade unit includes a window opening and a photovoltaic module area located around the window opening, and the facade unit is used as the basic unit for calculating the local heat transfer characteristics of the outer surface of the photovoltaic array; For a cavity wall, a single facade column is vertically divided into a top end module A, a middle repeating module B, and a bottom end module C; for a facade column containing N vertical facade units, its cavity wall is formed by vertically combining 1 module A, N-1 modules B, and 1 module C; for a complete double-layer curtain wall facade, M facade columns are horizontally spliced ​​to form a complete cavity system division, such as... Figure 3 As shown; Modules A, B, and C are divided into corresponding cavity wall sub-regions based on their geometric contours, window opening boundaries, velocity inlet boundaries, and pressure outlet boundaries, respectively forming sub-region sets for Module A, Module B, and Module C. Through the vertical combination and horizontal splicing of the above module sub-region sets, the sub-region division result of the complete double-layer curtain wall facade cavity wall is obtained.

[0010] 1-2), perform multi-condition CFD simulation, obtain key thermal flow field characteristic data such as local convective heat transfer coefficient, wall temperature, and cavity flow velocity according to the partition, and build a machine learning model training set; 1-3) Construct a data-driven proxy model framework based on multi-output XGBoost. Five input features are used for the outer surface of the photovoltaic array and nine input features are used for the cavity wall. The thermal-fluid coupling characteristics of their respective systems are efficiently predicted and simulated through three independently trained output regressors. (1-4) Input the constructed training data into the XGBoost model, use the mean absolute error as the loss function, calculate the error between the model's predicted value and the target value, and iteratively optimize the model parameters based on the error; then, optimize the key hyperparameters through the system and evaluate the model performance by combining cross-validation, and finally obtain the photovoltaic array outer surface model and cavity wall model.

[0011] The photovoltaic integrated curtain wall photovoltaic capacity and cooling and heating load dynamic prediction method of the present invention is further as follows: Step 2) specifically includes: setting the geometric dimensions of the enclosure structure, the thermal property parameters of the materials and the optical and electrical performance parameters of the photovoltaic modules, establishing the thermal network and airflow network of the structural units such as photovoltaic modules, cavities, walls and window frames, and defining the node connection relationship and the heat transfer, convection heat transfer and air flow mechanism. The governing equations of the network are based on the energy conservation of each node:

[0012] in, Let node i be the heat capacity; and Let the temperatures of node i and node j be denoted as 'i' and 'j' respectively. Let i be the set of adjacent nodes connected to node i via thermal conduction. Let be the thermal resistance between node i and its adjacent node j; This is a solar shortwave radiation heat source.

[0013] The photovoltaic integrated curtain wall photovoltaic capacity and cooling and heating load dynamic prediction method of the present invention is further as follows: Step 3) specifically involves: inputting meteorological time series data such as solar irradiance, ambient temperature, wind speed and wind direction, setting indoor temperature boundary conditions, setting initial values ​​of temperature field and flow field in the calculation domain, and determining time step parameters.

[0014] The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall of the present invention is further described in step 5): 5-1) Based on the facade unit zoning, the upper and lower openings corresponding to each column of units are regarded as independent flow channels, and the pressure difference-flow relationship is used to describe the air exchange intensity under the combined action of wind pressure and thermal pressure:

[0015] in, The pressure difference between the upper and lower openings; and The air pressure at the bottom opening and the air pressure at the top opening; Outdoor air density; The difference in height between the top and bottom openings; Ambient temperature; The temperature of the air in the cavity; Treating the opening as a thin orifice, the pressure difference-flow rate relationship is established using the orifice flow rate equation:

[0016]

[0017] in, The airflow velocity at the opening; The open flow coefficient; For open-circuit mass flow rate; Let the area be the opening area; 5-2), based on the current facade opening wind pressure and cavity air temperature and velocity, call the cavity wall model constructed in step 1) to obtain the cavity wall local convective heat transfer coefficient at the next moment.

[0018] The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall of the present invention is further described in step 6): 6-1), based on the energy conservation equation and the current wall and cavity air temperature distribution, combined with the convective heat transfer coefficients of the photovoltaic array outer surface and cavity wall obtained by steps 4) and 5) respectively, the module cell temperature and wall temperature distribution at the next moment are obtained. 6-2), based on the cavity air energy equation, combined with the cavity side surface temperature of the component. Temperature of the side surface of the wall cavity To obtain the cavity air temperature at the next moment. .

[0019] The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall of the present invention is further described in step 7): 7-1) A single-diode model is used to describe the IV relationship of a photovoltaic module under a given irradiance and cell temperature, and the module output power is obtained at each time step through maximum power point tracking:

[0020]

[0021]

[0022]

[0023] in, and The maximum power point voltage and the component terminal voltage are at time t. and Let be the maximum power point current and the component terminal current at time t; Let be the value of the photogenerated current at time t; Let be the value of the diode's reverse saturation current at time t; It is a series resistor; Let be the value of the parallel resistor at time t; For diode factor parameters; Reference temperature Diode factor parameters under; For reference temperature; and Let be the incident irradiance of the component and the reference irradiance at time t; The short-circuit current temperature coefficient; 7-2), after obtaining the instantaneous power of each partition, the total array power is discretely integrated over time to obtain the cumulative power generation during the statistical period:

[0024]

[0025] in, Let k be the power generation of the photovoltaic module in section k at time t; This refers to the overall loss coefficient of photovoltaic power generation. Let be the power generation efficiency of the photovoltaic module in partition k at time t; Let be the incident irradiance of the photovoltaic module in section k at time t; The effective radiation area of ​​photovoltaic module partition k; This refers to the cumulative power generation of the photovoltaic modules.

[0026] The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall of the present invention is further described in step 8): 8-1), Calculate the instantaneous heat transfer load on the indoor side caused by the building envelope:

[0027]

[0028] in, Let be the heat transfer load caused by the building envelope at time t; Let be the cooling load caused by the building envelope at time t; Let be the heating load caused by the building envelope at time t; 8-2), convert the heat load into the corresponding air conditioning input power, i.e.

[0029]

[0030] in, The power required by the air conditioning system at time t to meet the load of the building envelope; and The coefficient of performance (COP) for air conditioning systems; This represents the cumulative energy consumption of the air conditioning system.

[0031] The photovoltaic integrated curtain wall photovoltaic power generation and cooling and heating load dynamic prediction method of the present invention can also be: step 9) specifically is: through steps 5) to 8), a complete dynamic coupling prediction process of non-uniform thermal field and power generation of photovoltaic integrated curtain wall at a certain time is constructed, the calculation time is continuously updated, each time t is updated to t+dt, and steps 5) to 8) are repeated to obtain the dynamic distribution of wall temperature and component power generation at different times.

[0032] Compared with the prior art, the present invention has the following beneficial effects: 1. The photovoltaic integrated curtain wall photovoltaic power generation and cooling and heating load dynamic prediction method of the present invention is based on the XGBoost algorithm, constructs a fast prediction model of convection heat transfer of double curtain wall, and combines it with the heat-airflow network model to realize the dynamic prediction of non-uniform thermal field caused by complex geometric features in photovoltaic integrated double curtain wall system. It can accurately obtain the temperature and flow velocity distribution of module cells, wall and cavity air, and finally couple a single diode power model to realize the dynamic coupling and fast calculation of module temperature and power generation, and simultaneously predict the additional cooling and heating load caused by double curtain wall system.

[0033] 2. The photovoltaic power generation and dynamic prediction method for heating and cooling load of photovoltaic integrated curtain walls of the present invention can be used to predict the non-uniform thermal field distribution and system dynamic performance caused by complex structures such as window openings and structural keel in photovoltaic integrated double-layer curtain walls. Compared with traditional CFD simulation or simplified thermal network model, it solves the problems that the influence of complex geometric features on thermal field distribution cannot be considered in engineering applications and that traditional methods are difficult to balance between calculation accuracy and efficiency. It provides an effective tool for performance optimization and energy consumption assessment of photovoltaic building integrated systems. Attached Figure Description

[0034] Figure 1 This is a flowchart of the method for dynamically predicting photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain walls according to the present invention.

[0035] Figure 2 This is a schematic diagram of the generation scheme for the photovoltaic array outer surface model dataset in step 1) of the present invention.

[0036] Figure 3 This is a schematic diagram of the generation scheme for the cavity wall model dataset in step 1) of the present invention.

[0037] Figure 4 This is a schematic diagram of the cavity inner surface convection heat transfer partition coding in step 1) of the present invention.

[0038] Figure 5 This is a schematic diagram of the photovoltaic array outer surface model in step 1) of the present invention.

[0039] Figure 6 This is a structural diagram of the cavity wall model in step 1) of the present invention.

[0040] Figure 7 This is a schematic diagram of the curtain wall partitioning discretization used for thermal network solution in step 2) of the present invention.

[0041] Figure 8 This is a schematic diagram of the heat transfer mechanism and thermal network nodes of the photovoltaic integrated curtain wall in step 2) of the present invention.

[0042] Figure 9 This refers to the RC network constructed in the thickness direction in step 2) of the present invention.

[0043] Figure 10 This is a schematic diagram of the airflow network structure inside the cavity in step 2) of the present invention. Detailed Implementation

[0044] Please refer to the instruction manual appendix. Figure 1 As shown, this invention is a method for dynamically predicting photovoltaic capacity and heating / cooling load of integrated photovoltaic curtain walls, which includes the following process steps: 1) Constructing a rapid prediction model for convection heat transfer in curtain walls: Based on the XGBoost algorithm and CFD simulation dataset, and considering the influence of complex geometric features such as facade window frames and structural keels, a rapid prediction model for convection heat transfer in double-layer curtain walls is constructed. This model includes a photovoltaic array outer surface model and a cavity wall model.

[0045] This step is specifically as follows: 1-1) The heat transfer characteristics of the outer surface of the photovoltaic array and the cavity wall are analyzed separately, and sub-regions are divided according to the differences in their structural geometry and heat transfer characteristics. The facade unit composed of the window opening and its adjacent photovoltaic modules is used as the basic division object. The outer surface of the photovoltaic array is divided into N rows and M columns of facade units according to a unified facade module, forming a total of N×M facade units, where N and M are both positive integers. Figure 2 As shown; each facade unit includes a window opening and a photovoltaic module area located around the window opening, and the facade unit is used as the basic unit for calculating the local heat transfer characteristics of the outer surface of the photovoltaic array; For a cavity wall, a single facade column is vertically divided into a top end module A, a middle repeating module B, and a bottom end module C; for a facade column containing N vertical facade units, its cavity wall is formed by vertically combining 1 module A, N-1 modules B, and 1 module C; for a complete double-layer curtain wall facade, M facade columns are horizontally spliced ​​to form a complete cavity system division, such as... Figure 3 As shown; Modules A, B, and C are divided into corresponding cavity wall sub-regions based on their geometric contours, window opening boundaries, velocity inlet boundaries, and pressure outlet boundaries, respectively forming sub-region sets for Module A, Module B, and Module C. Through the vertical combination and horizontal splicing of the above module sub-region sets, the sub-region division result of the complete double-layer curtain wall facade cavity wall is obtained.

[0046] 1-2), conduct multi-condition CFD simulations. The dataset generation scheme for the photovoltaic array outer surface model is shown in Table 1 and 2. Figure 4 The scheme for generating the cavity wall model dataset is shown in Table 2 and Figure 5 Based on the partitioning, key thermal flow field characteristic data such as local convective heat transfer coefficient, wall temperature, and cavity flow velocity are obtained, and a training set for machine learning models is constructed.

[0047] Table 1: Dataset Generation Scheme for Photovoltaic Array Outer Surface Model

[0048] Table 2: Cavity Wall Model Dataset Generation Scheme

[0049] Before model training, all input features and output variables are scaled to the interval [0, 1] using the min-max normalization method:

[0050] The dataset was randomly divided into a training set (70%), a validation set (15%), and a test set (15%) to ensure that each subset is representative of the parameter space distribution.

[0051] (1-3) To address the convective heat transfer process on the outer surface and cavity walls of the photovoltaic array, a data-driven proxy model framework based on multi-output XGBoost is constructed. Five input features are used for the outer surface of the photovoltaic array (see...). Figure 6 Nine input features are used for the cavity wall (see...). Figure 7 Each system uses three independently trained output regressors to efficiently predict and simulate the thermal-fluid coupling characteristics of its respective system.

[0052] (1-4) The constructed training data is input into the XGBoost model. The mean absolute error is used as the loss function to calculate the error between the model's predicted value and the target value, and the model parameters are iteratively optimized based on the error. Subsequently, the model performance is evaluated by systematically optimizing key hyperparameters and combining cross-validation to ensure that the model has stable and reliable prediction capabilities on different data subsets.

[0053] 2) Constructing the thermal-airflow network of the curtain wall: Setting the geometric dimensions of the building envelope, the thermal properties of the materials, and the optical and electrical performance parameters of the photovoltaic modules. Establishing the thermal and airflow networks of the structural units such as photovoltaic modules, cavities, walls, and window frames, defining the node connection relationships and the heat transfer, convection heat transfer, and airflow mechanisms.

[0054] This step is specifically as follows: 2-1) Establishing a data-driven construction rule and node index system for the heat-airflow network: Based on the geometric dimensions of the building envelope, component types, material thermal properties, and optical and electrical performance parameters of the photovoltaic modules (see Tables 3 to 5), a mapping relationship is established between the curtain wall discrete units, internal calculation partitions, node types, and global node numbers. Each node includes its row number, column number, component type, partition location, node category, and corresponding parameters. Subsequently, node sets and branch sets are automatically generated according to preset heat transfer paths and airflow paths. The node sets include photovoltaic module nodes, cavity air nodes, wall nodes, window frame nodes, and indoor / outdoor boundary nodes. The branch sets include solid heat conduction branches, surface convection heat transfer branches, long-wave radiation heat transfer branches, solar radiation heat source branches, and air mass flow and energy transport branches.

[0055] In practical implementation, the above-mentioned heat-airflow network can be automatically constructed using programming languages ​​such as Python. Specifically, data structures such as arrays, lists, or dictionaries are used to store the node number, spatial location, component attributes, and thermal property parameters. A connection table is used to store the heat transfer relationship or airflow relationship between adjacent nodes, and the corresponding equivalent thermal conductivity, convective heat transfer coefficient, radiative heat transfer coefficient, and airflow resistance parameters are calculated according to the branch type. Furthermore, the node information and branch connection relationships are converted into a node-branch connection table, an equivalent thermal conductivity matrix, a node heat capacity vector, a heat source vector, and an airflow connection matrix, thereby forming a unified calculation network that can be used for the automatic assembly of subsequent heat network equations and airflow network equations. When the number of rows and columns of the curtain wall, structural dimensions, material parameters, or photovoltaic module parameters change, the number of nodes, branch connection relationships, and network matrix can be automatically updated by modifying the input parameters, without the need to manually rebuild the heat-airflow network model.

[0056] Table 3: Basic Information of Building Facade in the Example

[0057] Table 4: Thermophysical properties of building envelope materials in the examples

[0058] Table 5: Basic Information of Photovoltaic Modules in the Example

[0059] 2-2), the curtain wall is discretized into component units in the vertical direction (N rows) and horizontal direction (M columns) to clarify the differences in heat transfer response of different units under different irradiation, boundary conditions and structural differences; then, within each facade unit, calculation sub-zones are further divided and temperature nodes are set (e.g. Figure 2As shown in the figure, this allows for the description of the heat redistribution process through equivalent conduction, convection, and radiation branches between zones, while maintaining a clear understanding of the interlayer heat transfer paths and boundary heat transfer mechanisms. This discrete framework provides a unified spatial index and node organization basis for the subsequent assembly of the thermal network equations and the solution of the zoned air-wall coupling.

[0060] 2-3) Establish a heat-airflow network for structural units such as photovoltaic modules, cavities, walls, and window frames, defining the node connection relationships and heat transfer, convection heat transfer, and airflow mechanisms, such as... Figures 8 to 10 As shown; The fundamental governing equation of the network is that the energy of each node is conserved:

[0061] in, Let node i be the heat capacity; and Let the temperatures of node i and node j be denoted as 'i' and 'j' respectively. Let i be the set of adjacent nodes connected to node i via thermal conduction. Let be the thermal resistance between node i and its adjacent node j; This is a solar shortwave radiation heat source.

[0062] 3) Input meteorological time series data into the heat-airflow network model constructed in step 2), set indoor temperature boundary conditions, set the initial values ​​of the temperature field and flow field in the computational domain, and determine the time step parameters.

[0063] In this embodiment, the calculation period is set to the typical winter weather day (December 4th) of a certain city, from 06:00 to 18:00. Meteorological time series data are input, including horizontal solar irradiance, ambient temperature, wind speed and wind direction (see Table 6). The indoor temperature is set to 20 degrees. Table 6: Input conditions for calculation during the runtime of Example 1, December 4th, from 06:00 to 18:00

[0064] The initial values ​​of the temperature field and flow field in the computational domain are set according to the ambient temperature and wind speed, and the time step is determined to be 1 hour, with a total simulation duration of 24 hours.

[0065] 4) Call the photovoltaic array outer surface model constructed in step 1), combine it with the meteorological time series data in step 3), and the facade geometric parameters to obtain the full time series data of the local convective heat transfer coefficient and the facade wind pressure distribution of the outer surface.

[0066] 5) Using the cavity wall model constructed in step 1), obtain the local convective heat transfer coefficient of the cavity wall at the next moment. This specifically includes the following steps: 5-1), based on the wind pressure difference at the lower opening of the facade at the current moment. Top-opening air pressure and cavity air temperature Obtain the airflow velocity at the air inlet. .

[0067] Specifically, based on the facade unit zoning, the upper and lower openings corresponding to each column of units are regarded as independent flow channels, and the pressure difference-flow relationship is used to describe the air exchange intensity under the combined action of wind pressure and thermal pressure:

[0068] in, The pressure difference between the upper and lower openings; and The air pressure at the bottom opening and the air pressure at the top opening; Outdoor air density; The difference in height between the top and bottom openings; Ambient temperature; This refers to the temperature of the air in the cavity.

[0069] Treating the opening as a thin orifice, the pressure difference-flow rate relationship is established using the orifice flow rate equation:

[0070]

[0071] in, The airflow velocity at the opening; The open flow coefficient; For open-circuit mass flow rate; Let be the area of ​​the opening.

[0072] 5-2), based on the current facade opening wind pressure and cavity air temperature and velocity, call the cavity wall model constructed in step 1) to obtain the local convective heat transfer coefficient of the cavity wall at the next moment. Specifically, read the building envelope dimensional parameters (window opening height-to-width ratio) from step 2). , facade window-to-wall ratio Cavity thickness keel width and window opening height-to-width ratio The number of cavity modules (1 module A, N-1 modules B, and 1 module C, where n is the number of facade layers) and the corresponding number of measuring points are determined based on the constructed facade matrix. Then, the above dimensional parameters, module and measuring point information are compared with the current... Inlet flow rate and ambient (entrance) temperature The cavity wall convective heat transfer rapid prediction model (i.e., cavity wall model) constructed in step 1) is input together to obtain the local convective heat transfer coefficient of the cavity wall at the corresponding time. Including the local convective heat transfer coefficient on the cavity side of the photovoltaic module and the local convective heat transfer coefficient of the wall cavity side .

[0073] 6) Based on the energy conservation equation, and combined with the convective heat transfer coefficients of the outer surface of the photovoltaic array and the cavity wall obtained in steps 4) and 5), the non-uniform thermal field of the curtain wall is solved to obtain the temperature distribution of the module cells, the wall and the cavity air.

[0074] This step specifically includes the following steps: 6-1) Based on the energy conservation equation and the current wall and cavity air temperature distribution, combined with the convective heat transfer coefficients of the photovoltaic array outer surface and cavity wall obtained by steps 4) and 5) respectively, the module cell temperature and wall temperature distribution at the next moment are obtained.

[0075] Specifically, based on the convective heat transfer coefficient of the photovoltaic module's outer surface obtained in steps 4) and 5), and the convective heat transfer coefficient of the cavity wall Convection heat transfer terms between the outer surface of the photovoltaic module and the outdoor air were established respectively. Convection heat transfer between the cavity side surface of the photovoltaic module and the cavity air convective heat transfer between the side surface of the wall cavity and the air in the cavity The thermal resistance of each convective heat transfer surface is determined based on the local convective heat transfer coefficient and heat transfer area of ​​the corresponding surface.

[0076]

[0077]

[0078] in, and These are the areas of the photovoltaic modules and the wall sub-sections, respectively.

[0079] Based on the aforementioned convective heat transfer terms and the energy conservation equations for each structural layer, heat balance equations are established for each temperature node of the photovoltaic module and the wall, and the wall temperature distribution at the next moment is obtained by solving these equations. The photovoltaic module temperature nodes include the outdoor surface temperature. Battery layer temperature and cavity side surface temperature The wall temperature node includes the surface temperature of the wall cavity side. Temperature of internal joints in the wall and and indoor surface temperature .

[0080] Outdoor surface of photovoltaic modules:

[0081] Photovoltaic module cell layer:

[0082] Cavity side surface of photovoltaic module:

[0083] The side surface of the wall cavity:

[0084] Interior surface of the wall:

[0085] in, and The inner surface of the photovoltaic panel and the outer surface of the wall are zoned and numbered; This refers to the node temperature of the photovoltaic cell layer. Temperature of the nodes on the outer surface of the component; The equivalent thermal resistance between the outer surface of the component and the cell layer node; The equivalent thermal resistance for heat transfer between the outer surface of the component and the outdoor air; The equivalent thermal resistance of the component's outer surface to long-wave radiation from the environment; This is the absorption of heat by shortwave waves in the outer glass layer; The equivalent heat capacity of the photovoltaic cell layer node; The equivalent node temperature on the inner surface of the component; This is the equivalent thermal resistance between the photovoltaic cell and the nodes on the inner surface of the module; This refers to the absorption of heat by shortwave radiation in the battery layer. This is the equivalent heat deduction item corresponding to the electrical energy output; and The equivalent heat capacity of the node between the inner surface of the component and the outer surface of the wall; and The equivalent node temperature between the inner surface of the component and the outer surface of the wall; The equivalent temperature of the joint of the wall insulation layer; The equivalent thermal resistance from the photovoltaic cell to the inner surface node of the module; The equivalent thermal resistance from the outer surface of the wall to the insulation layer node; and The equivalent thermal resistance for convective heat transfer between the two walls of the cavity side; The equivalent thermal resistance to long-wave radiation between the inner surface of the photovoltaic cell and the outer surface of the wall. Temperature of the inner surface nodes of the wall; Temperature at the joint of the concrete layer in the wall; Indoor air temperature; The indoor equivalent radiation node temperature; The equivalent thermal resistance between the wall structure layer and the inner surface node; The equivalent thermal resistance for heat transfer between the indoor inner surface node and the indoor air convection. It represents the equivalent thermal resistance of the indoor inner surface node to the long-wave radiation of the indoor environment.

[0086] To characterize the non-uniform temperature field of the photovoltaic integrated curtain wall, a three-dimensional equivalent discretization approach based on energy conservation is adopted. Energy connections are established between different nodes through equivalent heat conduction, convection, and radiation heat transfer channels, thereby realizing three-dimensional heat transfer coupling calculation.

[0087] in, This represents the heat transfer power between partitions k and l. This is the in-plane thermal conductivity correction factor; The equivalent thermal conductivity between partitions k and l; This represents the equivalent thermal conductivity cross-sectional area between partitions k and l. This is the equivalent heat conduction path length between the corresponding nodes of partition k and l; and This represents the equivalent temperature of corresponding nodes in the same structural layer for partitions k and l.

[0088] A dedicated heat conduction branch is set up between the window frame zone and its adjacent photovoltaic zone or wall zone at the corresponding layer node, and the thermal resistance is constructed using the thermal conductivity parameters of the window frame material. The calculation method is as follows:

[0089]

[0090]

[0091] in, Equivalent heat capacity of window frame nodes; The equivalent node temperature of the window frame; This refers to the heat exchange power between the window frame and the inner surface nodes of adjacent photovoltaic modules. This refers to the heat exchange power between the window frame and adjacent wall surface nodes. The equivalent thermal resistance from the window frame to the inner surface node of the adjacent photovoltaic module; It is the equivalent thermal resistance from the window frame to the adjacent wall surface node.

[0092] The long-wave radiation exchange between the two walls within the cavity employs a locally sparse many-to-many radiation coupling algorithm; for any photovoltaic inner surface partition k, the set of outer wall surface partitions participating in radiation exchange is defined as follows. Its physical meaning is to only consider partition l that contributes significantly to the viewpoint geometrically with partition k, in order to maintain the sparsity of the coupling matrix; the specific calculation method is as follows:

[0093]

[0094]

[0095]

[0096]

[0097] in: The combined effective emissivity of the two walls; The angle coefficient between partition k and partition l; and Let k be the row and column indexes of the elevation partition; and For partition l, specify the row and column indexes within the elevation partition; The neighborhood radius is the number of nodes in the partitioned grid. The equivalent spatial distance between partitions k and l; The distance between the center points of partitions k and l on the same wall surface; The cavity thickness; Unnormalized weights constructed for the angle coefficients; The angle coefficient between partition k and partition l; This represents the sum of the weights of all partitions within the neighborhood of partition k. This represents the total number of distributable angle coefficients for partition k.

[0098] 6-2), based on the cavity air energy equation, combined with the cavity side surface temperature of the component. Temperature of the side surface of the wall cavity To obtain the cavity air temperature at the next moment. .

[0099] Specifically, within the zonal modeling framework, sensible heat exchange occurs between the zoned air and the two side walls through convection; simultaneously, air exchange between zones achieves energy transfer and redistribution between zones through enthalpy transport.

[0100] in, The heat capacity of the k-air node in the partition; air density; The specific heat capacity of air; Let k be the air volume of the partition.

[0101] The inter-zone air exchange term characterizes the effect of airflow transporting sensible heat between zones; when there is mass exchange between adjacent zones, air transfers energy between zones in the form of enthalpy transport; the equivalent thermal resistance of enthalpy transport from zone k to l is defined as follows:

[0102] in, Let be the mass flow rate transported from adjacent partition l into partition k.

[0103] To maintain consistency with the time discretization of the wall thermal network, the transient energy equation of the cavity air is discretized. A backward Eulerian implicit scheme is used to unify the convective heat transfer and enthalpy transport terms. By taking values ​​at specific times, we obtain a discrete equation for the zoned air temperature that is easy to iterate in conjunction with the heat network equation:

[0104] in, Let k be the air volume of the partition; and These are the air node temperatures for partitions k and l, respectively; and The equivalent node temperatures of the inner and outer surfaces of component partition k; The equivalent thermal resistance for convective heat transfer on the cavity side of the component partition k; The equivalent thermal resistance for convective heat transfer on the cavity side of the wall partition; The equivalent thermal resistance for enthalpy transport between partitions k and l is given.

[0105] 7) Based on the power of a single diode, combined with the solar irradiance conditions and the cell temperature of the module obtained in step 6), the power generation of the photovoltaic array and the cumulative power generation are obtained.

[0106] This step specifically includes the following steps: 7-1) A single-diode model is used to describe the IV relationship of a photovoltaic module under a given irradiance and cell temperature, and the module output power is obtained at each time step through maximum power point tracking (MPP):

[0107]

[0108]

[0109]

[0110] in, and The maximum power point voltage and the component terminal voltage are at time t. and Let be the maximum power point current and the component terminal current at time t; Let be the value of the photogenerated current at time t; Let be the value of the diode's reverse saturation current at time t; It is a series resistor; Let be the value of the parallel resistor at time t; For diode factor parameters; Reference temperature Diode factor parameters under; For reference temperature; and Let be the incident irradiance of the component and the reference irradiance at time t; This is the temperature coefficient of the short-circuit current.

[0111] (7-2) After obtaining the instantaneous power of each partition, the power of the entire array is further discretely integrated over time to obtain the cumulative power generation during the statistical period.

[0112]

[0113]

[0114] in, Let k be the power generation of the photovoltaic module in section k at time t; This refers to the overall loss coefficient of photovoltaic power generation. Let be the power generation efficiency of the photovoltaic module in partition k at time t; Let be the incident irradiance of the photovoltaic module in section k at time t; The effective radiation area of ​​photovoltaic module partition k; This refers to the cumulative power generation of the photovoltaic modules.

[0115] 8) Based on the wall temperature obtained in step 6) and the set indoor temperature, calculate the additional cooling and heating load and air conditioning energy consumption caused by the curtain wall system. This specifically includes the following steps: The inner surface temperature of the wall obtained based on the aforementioned wall temperature distribution prediction model. It can calculate the instantaneous heat transfer load on the indoor side caused by the building envelope. :

[0116]

[0117] in, Let be the heat transfer load caused by the building envelope at time t; Let be the cooling load caused by the building envelope at time t; Let t be the heating load caused by the building envelope at time t.

[0118] Given the coefficients of performance for cooling and heating, the heat load is further converted into the corresponding air conditioning input power for subsequent energy consumption calculations during the statistical period:

[0119]

[0120] in, The power required by the air conditioning system at time t to meet the load of the building envelope; and The coefficient of performance (COP) for air conditioning systems; This represents the cumulative energy consumption of the air conditioning system.

[0121] 9) Repeat steps 5) to 8) to obtain the dynamic distribution of wall temperature and module power generation at different times, so as to realize the full-time dynamic prediction of photovoltaic power generation and heating and cooling load.

[0122] Specifically, steps 5) to 8) constitute a complete dynamic coupling prediction process of non-uniform thermal field and power generation of photovoltaic integrated curtain wall at a given time. The calculation time is continuously updated, and each time t is updated to t+dt. Steps 5) to 8) are repeated to obtain the dynamic distribution of wall temperature and module power generation at different times, thereby realizing the full-time dynamic prediction of photovoltaic capacity and heating and cooling load.

[0123] Within the set calculation period, the conversion results of the component temperature, photovoltaic power generation, and air conditioning energy consumption on the building envelope side of the photovoltaic integrated curtain wall in this embodiment are shown in Table 7.

[0124] Table 7: Component temperature, photovoltaic power generation, and air conditioning energy consumption conversion for photovoltaic integrated curtain wall in the example.

[0125] The above-described specific embodiments are merely preferred embodiments of this invention and are not intended to limit this invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the protection scope of this invention.

Claims

1. A method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall, characterized in that: The process includes the following steps: 1) Construct a rapid prediction model for convective heat transfer in the curtain wall, which includes a photovoltaic array outer surface model and a cavity wall model; 2) Construct a thermal-airflow network for the curtain wall; 3) Input meteorological time series data into the heat-airflow network model constructed in step 2), set indoor temperature boundary conditions, set the initial values ​​of the temperature field and flow field in the computational domain, and determine the time step parameters; 4) Using the photovoltaic array outer surface model constructed in step 1), combined with the meteorological time series data in step 3) and the facade geometric parameters, obtain the full time series data of the local convective heat transfer coefficient and the facade wind pressure distribution of the outer surface. 5) Call the cavity wall model constructed in step 1) to obtain the local convective heat transfer coefficient of the cavity wall at the next moment; 6) Based on the energy conservation equation, and combined with the convective heat transfer coefficients of the photovoltaic array outer surface and cavity wall obtained in steps 4) and 5), the non-uniform thermal field of the curtain wall is solved to obtain the temperature distribution of the module cells, wall and cavity air. 7) Based on the power of a single diode, combined with the solar irradiance conditions and the module cell temperature obtained in step 6), obtain the photovoltaic array power generation and cumulative power generation; 8) Based on the wall temperature obtained in step 6) and the set indoor temperature, calculate the additional heating and cooling loads and air conditioning energy consumption caused by the curtain wall system; 9) Repeat steps 5) to 8) to obtain the dynamic distribution of wall temperature and module power generation at different times, so as to realize the full-time dynamic prediction of photovoltaic power generation and heating and cooling load.

2. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 1) specifically refers to: 1-1) The outer surface of the photovoltaic array and the cavity wall are divided into sub-regions; wherein, the facade unit composed of the window opening and its adjacent photovoltaic modules is used as the basic division object, the outer surface of the photovoltaic array is divided into N rows and M columns of facade units according to a unified facade module, forming a total of N×M facade units, where N and M are both positive integers; each facade unit includes a window opening and a photovoltaic module area located around the window opening, and the facade unit is used as the basic unit for calculating the local heat transfer characteristics of the outer surface of the photovoltaic array; For a cavity wall, a single facade column is divided vertically into a top end module A, a middle repeating module B, and a bottom end module C; for a facade column containing N vertical facade units, its cavity wall is formed by vertically combining 1 module A, N-1 modules B, and 1 module C; for a complete double-layer curtain wall facade, M facade columns are spliced ​​horizontally to form a complete cavity system division. Modules A, B, and C are divided into corresponding cavity wall sub-regions based on their geometric contours, window opening boundaries, velocity inlet boundaries, and pressure outlet boundaries, respectively forming sub-region sets for Module A, Module B, and Module C. Through the vertical combination and horizontal splicing of the above module sub-region sets, the sub-region division result of the complete double-layer curtain wall facade cavity wall is obtained. 1-2), perform multi-condition CFD simulation, obtain key thermal flow field characteristic data such as local convective heat transfer coefficient, wall temperature, and cavity flow velocity according to the partition, and build a machine learning model training set; 1-3) Construct a data-driven proxy model framework based on multi-output XGBoost. Five input features are used for the outer surface of the photovoltaic array and nine input features are used for the cavity wall. The thermal-fluid coupling characteristics of their respective systems are efficiently predicted and simulated through three independently trained output regressors. (1-4) Input the constructed training data into the XGBoost model, use the mean absolute error as the loss function, calculate the error between the model's predicted value and the target value, and iteratively optimize the model parameters based on the error; then, optimize the key hyperparameters through the system and evaluate the model performance by combining cross-validation, and finally obtain the photovoltaic array outer surface model and cavity wall model.

3. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 2) specifically involves: setting the geometric dimensions of the enclosure structure, the thermal properties of the materials, and the optical and electrical performance parameters of the photovoltaic modules; establishing the thermal and airflow networks of the structural units such as the photovoltaic modules, cavities, walls, and window frames; and defining the node connection relationships and the heat transfer, convection heat transfer, and airflow mechanisms. The governing equations of the network are based on the energy conservation of each node: in, Let node i be the heat capacity; and Let the temperatures of node i and node j be denoted as 'i' and 'j' respectively. Let i be the set of adjacent nodes connected to node i via thermal conduction. Let be the thermal resistance between node i and its adjacent node j; This is a solar shortwave radiation heat source.

4. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 3) specifically involves: inputting meteorological time-series data such as solar irradiance, ambient temperature, wind speed, and wind direction; setting indoor temperature boundary conditions; setting initial values ​​for the temperature field and flow field in the computational domain; and determining the time step parameters.

5. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 5) specifically involves: 5-1) Based on the facade unit zoning, the upper and lower openings corresponding to each column of units are regarded as independent flow channels, and the pressure difference-flow relationship is used to describe the air exchange intensity under the combined action of wind pressure and thermal pressure: in, The pressure difference between the upper and lower openings; and The air pressure at the bottom opening and the air pressure at the top opening; Outdoor air density; The difference in height between the top and bottom openings; Ambient temperature; The temperature of the air in the cavity; Treating the opening as a thin orifice, the pressure difference-flow rate relationship is established using the orifice flow rate equation: in, The airflow velocity at the opening; The open flow coefficient; For open-circuit mass flow rate; Let the area be the opening area; 5-2), based on the current facade opening wind pressure and cavity air temperature and velocity, call the cavity wall model constructed in step 1) to obtain the cavity wall local convective heat transfer coefficient at the next moment.

6. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 6) specifically involves: 6-1), based on the energy conservation equation and the current wall and cavity air temperature distribution, combined with the convective heat transfer coefficients of the photovoltaic array outer surface and cavity wall obtained by steps 4) and 5) respectively, the module cell temperature and wall temperature distribution at the next moment are obtained. 6-2), based on the cavity air energy equation, combined with the cavity side surface temperature of the component. Temperature of the side surface of the wall cavity To obtain the cavity air temperature at the next moment. .

7. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 7) specifically involves: 7-1) A single-diode model is used to describe the IV relationship of a photovoltaic module under a given irradiance and cell temperature, and the module output power is obtained at each time step through maximum power point tracking: in, and The maximum power point voltage and the component terminal voltage are at time t. and Let be the maximum power point current and the component terminal current at time t; Let be the value of the photogenerated current at time t; Let be the value of the diode's reverse saturation current at time t; It is a series resistor; Let be the value of the parallel resistor at time t; For diode factor parameters; Reference temperature Diode factor parameters under; For reference temperature; and Let be the incident irradiance of the component and the reference irradiance at time t; The short-circuit current temperature coefficient; 7-2), after obtaining the instantaneous power of each partition, the total array power is discretely integrated over time to obtain the cumulative power generation during the statistical period: in, Let k be the power generation of the photovoltaic module in section k at time t; This refers to the overall loss coefficient of photovoltaic power generation. Let be the power generation efficiency of the photovoltaic module in partition k at time t; Let be the incident irradiance of the photovoltaic module in section k at time t; The effective radiation area of ​​photovoltaic module partition k; This refers to the cumulative power generation of the photovoltaic modules.

8. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 8) specifically refers to: 8-1), Calculate the instantaneous heat transfer load on the indoor side caused by the building envelope: in, Let be the heat transfer load caused by the building envelope at time t; Let be the cooling load caused by the building envelope at time t; Let be the heating load caused by the building envelope at time t; 8-2), convert the heat load into the corresponding air conditioning input power, i.e. in, The power required by the air conditioning system at time t to meet the load of the building envelope; and The coefficient of performance (COP) for air conditioning systems; This represents the cumulative energy consumption of the air conditioning system.

9. The method for dynamic prediction of photovoltaic capacity and heating / cooling load of photovoltaic integrated curtain wall as described in claim 1, characterized in that: Step 9) specifically involves: constructing a complete dynamic coupling prediction process of non-uniform thermal field and power generation of photovoltaic integrated curtain wall at a given time through steps 5) to 8), continuously updating the calculation time, updating the time t to t+dt each time, and repeating steps 5) to 8) to obtain the dynamic distribution of wall temperature and component power generation at different times.

Images