Building outdoor climate comprehensive representation method based on multi-element meteorological parameter fusion

Abstract

The application discloses a kind of building outdoor climate comprehensive characterization method based on multi-element meteorological parameter fusion, meteorological data is divided according to day-night time scale, periodic trend of each meteorological parameter is extracted using ensemble empirical mode decomposition (EEMD), and periodic variation sequence (PVS) is obtained;Different meteorological elements are standardized by applying Z-score method, and the dimensional difference is eliminated;Further, a seasonal difference autoregressive moving average out-of-band regression (SARIMAX) model is constructed to quantify the contribution weight of each meteorological parameter to wall temperature;According to the weight coefficient, the weighted sum of each standardized meteorological element is formed, and the comprehensive characterization index (CMIP) of outdoor meteorological parameter is formed;Finally, by applying the CMIP method to the improvement of typical meteorological year (TMY) data, the problems of poor continuity, difficulty in removing outliers and insufficient long-term trend expression of traditional TMY data are effectively solved, and the prediction accuracy of building energy consumption is significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of outdoor calculation parameters for building energy conservation, specifically involving a comprehensive characterization method for building outdoor climate based on the fusion of multiple meteorological parameters. By using this comprehensive characterization method of outdoor meteorological parameters, the multiple characteristics of the outdoor climate environment are depicted, thereby improving the overall quality of input data for energy consumption models. Background Technology

[0002] In the field of built environment, outdoor climate is typically broken down into individual parameters such as temperature and humidity for analysis. While this method can reveal local mechanisms, it struggles to reflect the comprehensive impact of climate on buildings and makes research increasingly complex. Although meteorology has proposed several comprehensive climate indices to characterize outdoor climate, their application to built environment analysis is limited due to the different target groups. This invention, based on the intersection of meteorology, building heat transfer, and building climatology, and combined with a data-driven approach, constructs a dimensionless representation that simultaneously reflects the comprehensive effects of multiple meteorological elements by quantifying the weights of factors such as temperature, humidity, radiation, and wind speed, based on the mechanism of outdoor climate acting on the exterior walls of buildings. Furthermore, this representation is used to improve the traditional Typical Meteorological Year (TMY) database. Summary of the Invention

[0003] The purpose of this invention is to provide a comprehensive characterization method for building outdoor climate based on the fusion of multiple meteorological parameters, which solves the problems of difficult characterization methods and limited meteorological parameters in the existing technology.

[0004] The technical solution adopted in this invention is:

[0005] A comprehensive building outdoor climate characterization method based on multivariate meteorological parameter fusion is proposed. The method divides each meteorological parameter into diurnal time scales, and then uses ensemble empirical mode decomposition to extract the fluctuation characteristics of the meteorological parameters. A seasonal difference autoregressive moving average with exogenous regression model is introduced to determine the weighting factors of each meteorological parameter, and a comprehensive characterization model of outdoor meteorological parameters is obtained to obtain a comprehensive building outdoor climate characterization model. The low-frequency signal in the comprehensive building outdoor climate characterization model is used to reflect the long-term climate fluctuation characteristics and replace the low-frequency signal in TMY.

[0006] Optional, the specific steps for comprehensive characterization of outdoor meteorological parameters are as follows: S1: Data Cleaning Obtain raw meteorological parameters, including outdoor temperature, total solar radiation, wind speed, wind direction, atmospheric pressure, and relative humidity, and perform outlier processing on the meteorological parameter dataset; S2: Dynamic heat transfer through the wall A dynamic heat transfer model of the wall, consisting of the heat conduction equation and boundary conditions, is established, and the inner and outer wall temperatures of the wall are calculated using the dynamic heat transfer model. S3: Temporal Characteristics of Meteorological Data Using a conditional selection method, the time series and data of solar radiation not equal to zero were divided into the total number of daytime hours D and the daytime dataset D. a,m The time series data where solar radiation is zero is divided into the total number of nighttime hours N and the daytime dataset N. a,m ;in a Numbering meteorological parameters a =1,2,3,4,5 m For the number of days; S4: Extraction of meteorological data fluctuation characteristics For the daytime and nighttime datasets, EEMD was used to decompose the time series data of various outdoor meteorological parameters to obtain multiple intrinsic mode functions and a residual sequence for each meteorological parameter. Four low-frequency components, IMF5 to IMF8, were selected and added to the residual sequence to form a periodic variation sequence PVS. The correlation coefficient R was used to evaluate the correlation between the obtained periodic fluctuation expressions of each meteorological element and the original data. S5: Dimensional Standardization of Meteorological Parameters The Z-score standardization method is used to process the meteorological parameters, resulting in a standardized daytime calculation formula; the same applies to nighttime calculations. S6: Determination of weights for various meteorological parameters The diurnal weighting coefficients of each meteorological parameter are determined by extending the statistical regression model using a seasonal autoregressive integral moving average model. and ; S7: Comprehensive Characterization of Outdoor Meteorological Parameters Determining the weighting factors for each meteorological parameter and Then, a comprehensive representation based on the annual outdoor climate cycle fluctuation characteristics is obtained by weighted summation, and a dimensionless index representing the outdoor climate cycle fluctuation is constructed.

[0007] Optionally, in step S2, the heat transfer equation for the dynamic heat transfer model of the wall is:

[0008] After performing implicit difference on it, we get the following equation: ; Summarized as follows: ; In the formula, express i Node at n Temperature at any given time, in °C;K This represents thermal conductivity, measured in W / (m·K). This represents the thermal diffusivity, expressed in W / (m·K). S Wall area, unit m 2 x represents the spatial variable of the temperature field, in meters (m); t represents the time variable in the temperature field, in seconds (s). x represents the spatial step size, the distance between two adjacent discrete spatial points, in meters; t represents the time step, the time interval between two adjacent time steps, in seconds; ρ represents the density of the wall material, in kg / m³; c represents the specific heat capacity of the wall material, in J / (kg·K); The heat transfer equation for the dynamic heat transfer model of the wall is: ; ; ; = + ; ; In the formula, This represents the evaporative and convective heat transfer coefficient between the inner wall surface and the interior space, expressed in W / (m²). 2 ·K); This represents the evaporative and convective heat transfer coefficient between the exterior wall surface and the outside, expressed in W / (m²). 2 ·K); express t Evaporative heat transfer coefficient of the outer wall at any given time, in W / (m²) 2 ·K); This represents the convective heat transfer coefficient of the outer wall at time t, in W / (m²). 2 ·K); P Atmospheric pressure, unit: hPa; Relative humidity, expressed in % %. This represents the long-wave radiation per unit area of ​​the wall, measured in W / m². 2 ; 1 represents the wall's absorption rate; 2 represents the emissivity of the wall; The Stefan-Boltzmann constant is 5.67 × 10⁻⁶. -8 (W / m) 2 / K 4 ).

[0009] Optionally, in step S3, the meteorological data diurnal time characteristic segmentation model is as follows: ; In the formula, m For the number of days sequence number ( m =1,2,...,365); For the first m The number of daytime hours in a day, in hours (h). For the first m The number of night hours in a day, in hours (h). D Total sunshine hours, in hours (h). N Total nighttime hours, in hours (h). Q Solar radiation, measured in W / m.

[0010] Optionally, in step S4, the model for extracting the temporal fluctuation features of meteorological data is: ; In the formula, The original signal; For modal functions; It is a residual sequence; The periodic transformation sequence model is as follows: ; In the formula, These are the fitted values ​​of the annual average periodic variation sequence of various meteorological parameters, in °C. These are the annual average characteristic values ​​of various meteorological elements, in °C. This represents the annual average periodic fluctuation amplitude of various meteorological parameters, in °C. Indicates the sequence number of days; The constant represents the initial phase of the characteristic expression; It represents half a period of fluctuation in meteorological parameters, with the unit being seconds (s). The correlation coefficient R is calculated as follows: ; In the formula, and These are the calculated value and the measured value, respectively. and These represent the average of the calculated value and the measured value, respectively.

[0011] Optionally, in step S5, the Z-score standardized model is: ; ; ; In the formula ，y a Indicates the values ​​of different meteorological parameters; This represents the sample average value of different meteorological parameters; This represents the standardized value of meteorological parameter a based on daytime data. Daytime hourly serial number =1,2,..., 4384.

[0012] Optionally, in step S6, the SARIMAX model is: ; In the formula The autoregressive part aims to capture the difference between the current value and the previous value. The difference part is used to transform a non-stationary time series into a stationary series; This is the moving average component, used to capture the relationship between the current value and past error terms in the data; For the exogenous variable part, other factors that affect the regression results were added; This indicates the effect of the lagged term on the current value; L is the lag operator, representing the dependency on past time points; s Indicates the length of a season (cycle); d Represents non-seasonal differences d Secondly, it is used to eliminate trends (to stabilize the mean of the sequence); D Indicates seasonal difference D This is used to eliminate periodic seasonal effects; Indicates in t The raw time series values ​​observed at each moment; q Indicates non-seasonal AR order; Q Indicates the seasonal AR level; Represents the seasonal AR coefficient, with the f-th season lag for the... The impact; Represents white noise; Represents an exogenous variable vector of length n (such as temperature, humidity, customs, etc.); This indicates the marginal effect of each exogenous variable corresponding to the regression coefficient.

[0013] Optionally, in step S7, the comprehensive characterization model for outdoor meteorological parameters is: ; Where 'a' is the meteorological parameter number, a = 1, 2, 3, 4, 5, n = 5; For the hourly sequence number, T=1,2,...,8760; Weighting factors for various meteorological parameters during the day; These are the weighting factors for various meteorological parameters at night.

[0014] The beneficial effects of this invention are: This invention focuses on the comprehensive characterization of outdoor meteorological parameters of buildings, overcoming the limitations of traditional meteorological analysis methods in terms of dimensional differences, parameter heterogeneity, and dynamic fluctuation extraction. It presents a comprehensive characterization model of outdoor meteorological parameters based on exterior wall temperature, constructing a method that can represent the long-term fluctuations of building outdoor climate. Based on long-term meteorological observation data (taking Xi'an as an example), the Ensemble Empirical Mode Decomposition (EEMD) method is applied to decompose the multiple frequency components of the meteorological data, extracting the periodic variation characteristics of elements such as temperature, humidity, wind speed, and solar radiation. Then, the PVS data obtained from the decomposition is standardized using Z-score to eliminate dimensional differences, making the various meteorological elements comparable. To quantitatively measure the influence weight of each meteorological element on the thermal response of the building exterior wall, the SARIMAX model is used to analyze the coupling relationship between exterior wall temperature and meteorological data, thereby obtaining objective weight coefficients. and Based on this, the standardized multi-dimensional meteorological elements are weighted and summed to form a comprehensive meteorological information parameter (CMIP), which provides a dimensionless and operable meteorological driving factor for the subsequent optimization design of the building envelope. Attached Figure Description

[0015] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a diagram illustrating the characterization steps of the present invention; Figure 2 This is an exploded view of the outdoor air temperature EEMD of the present invention; Figure 3 This is a daytime outdoor air temperature fluctuation diagram of the present invention; Figure 4 This is a nighttime outdoor air temperature fluctuation diagram according to the present invention; Figure 5 This is a quantitative analysis diagram showing the weighting of various meteorological parameters with respect to wall temperature in this invention; Figure 6 This is a correlation coefficient graph between TMY, CMIP and measured data of the present invention. Detailed Implementation

[0016] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0017] By assessing the contribution weights of various meteorological parameters to external wall temperature, outdoor meteorological conditions are comprehensively characterized to better explain the impact of environmental changes on building energy consumption and reflect the macroscopic trends of changes under the coupling effects of multiple parameters. To accurately characterize the temporal variations of outdoor climate, each meteorological parameter is first divided into diurnal timescales, and then its fluctuation characteristics are extracted using Ensemble Empirical Mode Decomposition (EEMD). Given that each meteorological parameter is an important component of outdoor climate and significantly contributes to external wall temperature, a Seasonal Autoregressive Integrated Moving Average with eXogenous Variables (SARIMAX) model is further introduced to determine the weighting factors of each meteorological parameter and achieve a comprehensive characterization of outdoor meteorological parameters. To address the common problems of "difficulty in removing outliers" and "inability of low-frequency information to reflect long-term fluctuations" in the construction and selection of Typical Meteorological Year (TMY) data, a corresponding improvement scheme is proposed: Utilizing the advantage that low-frequency signals in the comprehensive characterization model can reflect the characteristics of long-term climate fluctuations, the low-frequency signals in TMY are replaced, thereby enhancing the TMY's characterization of seasonal changes and diurnal temperature range, making it more accurate and representative in building heat transfer analysis and energy consumption prediction.

[0018] First, meteorological parameters are divided according to diurnal timescales. Then, ensemble empirical mode decomposition (EEMD) is used to extract their fluctuation characteristics. Given that each meteorological parameter is an important component of outdoor climate and significantly contributes to exterior wall temperature, a seasonal autoregressive integrated moving average with exogenous variables model (SARIMAX) is further introduced to determine the weighting factors of each meteorological parameter and achieve a comprehensive representation of outdoor meteorological parameters. Addressing the common problems of "difficulty in removing outliers" and "low-frequency information failing to reflect long-term fluctuations" in the construction and selection of typical meteorological year (TMY) data, a corresponding improvement scheme is proposed: utilizing the advantage of low-frequency signals in the comprehensive representation model, which can reflect long-term climate fluctuation characteristics, the low-frequency signals in TMY are replaced, thereby enhancing TMY's characterization of seasonal variations and diurnal temperature range, making it more accurate and representative in building heat transfer analysis and energy consumption prediction.

[0019] Meanwhile, to further verify the effectiveness of CMIP, the parameters of the traditional TMY database were improved by utilizing its low-frequency signal characteristics. Consistency checks and correlation analyses were then performed between the improved data and measured data. The results showed that CMIP-TMY data better reflects long-term climate fluctuations and has fewer outliers compared to traditional TMY data. Finally, energy consumption analysis validated the effectiveness of CMIP-TMY.

[0020] Specifically, the steps for comprehensive characterization of outdoor meteorological parameters are as follows: S1: Data Cleaning Obtain raw meteorological parameters, including outdoor temperature, total solar radiation, wind speed, wind direction, atmospheric pressure, and relative humidity, and perform outlier processing on the meteorological parameter dataset; S2: Dynamic heat transfer through the wall A dynamic heat transfer model of the wall, consisting of the heat conduction equation and boundary conditions, is established, and the inner and outer wall temperatures of the wall are calculated using the dynamic heat transfer model. S3: Temporal Characteristics of Meteorological Data Using a conditional selection method, the time series and data of solar radiation not equal to zero were divided into the total number of daytime hours D and the daytime dataset D. a,m The time series data where solar radiation is zero is divided into the total number of nighttime hours N and the daytime dataset N. a,m ;in a Numbering meteorological parametersa =1,2,3,4,5 m For the number of days; S4: Extraction of meteorological data fluctuation characteristics For the daytime and nighttime datasets, EEMD was used to decompose the time series data of various outdoor meteorological parameters to obtain multiple intrinsic mode functions and a residual sequence for each meteorological parameter. Four low-frequency components, IMF5 to IMF8, were selected and added to the residual sequence to form a periodic variation sequence PVS. The correlation coefficient R was used to evaluate the correlation between the obtained periodic fluctuation expressions of each meteorological element and the original data.

[0021] S5: Dimensional Standardization of Meteorological Parameters The Z-score standardization method is used to process the meteorological parameters, resulting in a standardized daytime calculation formula; the same applies to nighttime calculations. S6: Determination of weights for various meteorological parameters The diurnal weighting coefficients of each meteorological parameter are determined by extending the statistical regression model using a seasonal autoregressive integral moving average model. and ; S7: Comprehensive Characterization of Outdoor Meteorological Parameters Determining the weighting factors for each meteorological parameter and Then, a comprehensive representation based on the annual outdoor climate cycle fluctuation characteristics is obtained by weighted summation, and a dimensionless index representing the outdoor climate cycle fluctuation is constructed.

[0022] Furthermore, in step S2, to accurately describe the thermal response of the exterior wall under frequent fluctuations in meteorological conditions, the concept of a "dynamic heat transfer coefficient" is introduced into the boundary conditions. This allows the thermal parameters of the wall model to be updated based on meteorological data at each time step, overcoming the shortcomings of previous models that treated the external environment as a fixed or simplified state. This enables a more realistic capture of the transient heat load on the wall surface. The dynamic heat transfer model of the wall is as follows:

[0023] After performing implicit difference on it, we get the following equation: ; Summarized as follows: ; In the formula, express i Node at n Temperature at any given time, in °C; K This represents thermal conductivity, measured in W / (m·K). This represents the thermal diffusivity, expressed in W / (m·K). S Wall area, unit m2 x represents the spatial variable of the temperature field, in meters (m); t represents the time variable in the temperature field, in seconds (s). x represents the spatial step size, the distance between two adjacent discrete spatial points, in meters; t represents the time step, the time interval between two adjacent time steps, in seconds; ρ represents the density of the wall material, in kg / m³; c represents the specific heat capacity of the wall material, in J / (kg·K).

[0024] According to claim 2, the comprehensive characterization method for building outdoor climate based on the fusion of multiple meteorological parameters is characterized in that, in step S2, the heat conduction equation of the dynamic heat transfer model of the wall is:

[0025] After performing implicit difference on it, we get the following equation: ; Summarized as follows: ; In the formula, express i Node at n Temperature at any given time, in °C; K This represents thermal conductivity, measured in W / (m·K). This represents the thermal diffusivity, expressed in W / (m·K). S Wall area, unit m 2 x represents the spatial variable of the temperature field, in meters (m); t represents the time variable in the temperature field, in seconds (s). x represents the spatial step size, the distance between two adjacent discrete spatial points, in meters; t represents the time step, the time interval between two adjacent time steps, in seconds; ρ represents the density of the wall material, in kg / m³; c represents the specific heat capacity of the wall material, in J / (kg·K).

[0026] The heat transfer equation for the dynamic heat transfer model of the wall is: ; ; ; = + ; ; In the formula, This represents the evaporative and convective heat transfer coefficient between the inner wall surface and the interior space, expressed in W / (m²). 2 ·K); This represents the evaporative and convective heat transfer coefficient between the exterior wall surface and the outside, expressed in W / (m²). 2 ·K); express t Evaporative heat transfer coefficient of the outer wall at any given time. Unit: W / (m²) 2 ·K); This represents the convective heat transfer coefficient of the outer wall at time t. The unit is W / (m²). 2 ·K); P Indicates atmospheric pressure. Unit: hPa; Relative humidity. Unit: % This represents the long-wave radiation per unit area of ​​the wall, measured in W / m². 2 ; 1 represents the wall's absorption rate; 2 represents the emissivity of the wall; This represents the Stefan-Boltzmann constant, approximately 5.67 × 10⁻⁶. -8 (W / m) 2 / K 4 ); Furthermore, step S3 involves dividing the time scale into day and night analyses, effectively distinguishing the different mechanisms of action of various meteorological elements during the day and night, and improving the accuracy of model predictions. By finely dividing and analyzing various meteorological parameters on the day-night time scale, the dynamic changes in building thermal performance can be more systematically assessed and predicted, providing a new research perspective for the comprehensive characterization of outdoor meteorological conditions and building energy consumption calculation. The meteorological data day-night time characteristic division model is as follows: ; In the formula, m For the number of days sequence number ( m =1,2,...,365); For the first m The number of daytime hours in a day, in hours (h). For the first m The number of night hours in a day, in hours (h). D Total sunshine hours, in hours (h). N Total nighttime hours, in hours (h). Q Solar radiation, measured in W / m.

[0027] Furthermore, in step S4, EEMD is used to effectively remove high-frequency interference signals, highlighting long-term trends and periodic changes, and more accurately capturing the core fluctuation characteristics of meteorological data. The meteorological data temporal fluctuation feature extraction model is as follows: ; In the formula, The original signal; For modal functions; It is a residual sequence.

[0028] The periodic transformation sequence model is as follows:

[0029] In the formula, These are the fitted values ​​of the annual average periodic variation sequence of various meteorological parameters, in °C. These are the annual average characteristic values ​​of various meteorological elements, in °C. This represents the annual average periodic fluctuation amplitude of various meteorological parameters, in °C. Indicates the sequence number of days; The constant represents the initial phase of the characteristic expression; It represents half a period of fluctuation in meteorological parameters, with the unit being seconds (s).

[0030] The correlation coefficient R is calculated as follows:

[0031] In the formula, and These are the calculated value and the measured value, respectively. and These represent the average of the calculated value and the measured value, respectively.

[0032] Furthermore, in step S5, the Z-score normalization method is used to perform dimensionless processing on the meteorological data. This method not only preserves the original distribution characteristics of the data but also enhances the comparability between different parameters, providing a reliable data foundation for the subsequent expression of the comprehensive characterization formula. The Z-score normalization model is as follows: ; ;

[0033] In the formula, y a Indicates the values ​​of different meteorological parameters; This represents the sample average value of different meteorological parameters; This represents the standardized value of meteorological parameter a based on daytime data. Daytime hourly serial number =1,2,..., 4384.

[0034] In step S6, the SARIMAX model considers both historical data and external meteorological factors, enabling it to more comprehensively and accurately reflect the relationship between meteorological parameters and wall temperature. Statistical methods are used to objectively determine the contribution weights of each meteorological parameter, avoiding interference from subjective experience and improving the model's scientific rigor. The SARIMAX model is as follows: ; In the formula The autoregressive part aims to capture the difference between the current value and the previous value. The difference part is used to transform a non-stationary time series into a stationary series; This is the moving average component, used to capture the relationship between the current value and past error terms in the data; For the exogenous variable part, other factors that affect the regression results were added.

[0035] This indicates the effect of the lagged term on the current value; L is the lag operator, representing the dependency on past time points; s Indicates the length of a season (cycle); d Represents non-seasonal differences d Secondly, it is used to eliminate trends (to stabilize the mean of the sequence); D Indicates seasonal difference D This is used to eliminate periodic seasonal effects; Indicates in t The raw time series values ​​observed at each moment; q Indicates non-seasonal AR order; Q Indicates the seasonal AR level; Represents the seasonal AR coefficient, with the f-th season lag for the... The impact; Represents white noise; Represents an exogenous variable vector of length n (such as temperature, humidity, customs, etc.); This indicates the marginal effect of each exogenous variable corresponding to the regression coefficient.

[0036] Furthermore, step S7, which integrates the CMIP low-frequency signal, overcomes the problems of discontinuity and insufficient low-frequency information representation in traditional TMY data, improving data representativeness and accuracy. The optimized CMIP-TMY data more realistically reflects long-term meteorological trends and reduces interference from anomalous data. The comprehensive characterization model for outdoor meteorological parameters is as follows: ; Where 'a' is the meteorological parameter number (a=1,2,3,4,5); For hourly serial numbers (T=1,2,...,8760); Weighting factors for various meteorological parameters during the day; These are the weighting factors for various meteorological parameters at night.

[0037] In summary, this invention overcomes the limitations of traditional meteorological analysis methods in terms of dimensional differences, parameter heterogeneity, and dynamic fluctuation extraction, and obtains a comprehensive characterization model of outdoor meteorological parameters based on the outer wall temperature, thus constructing a characterization method that can represent the long-term fluctuations of the building's outdoor climate.

[0038] Example 1: Part One: Please refer to Figure 1 This invention, based on the intersection of meteorology, building heat transfer, and building climatology, and combined with a data-driven approach, provides a dimensionless representation that simultaneously reflects the comprehensive effects of multiple meteorological elements by quantifying the weights of factors such as temperature, humidity, radiation, and wind speed, based on the mechanism of outdoor climate's influence on the exterior wall. First, each meteorological parameter is divided according to a diurnal timescale, and then its fluctuation characteristics are extracted using ensemble empirical mode decomposition. Given that each meteorological parameter is an important component of the outdoor climate and significantly contributes to the exterior wall temperature, a SARIMAX model is further introduced to determine the weighting factors of each meteorological parameter and achieve a comprehensive representation of outdoor meteorological parameters. Addressing the common problems of "difficulty in removing outliers" and "low-frequency information failing to reflect long-term fluctuations" in the construction and selection of TMY data, a corresponding improvement scheme is proposed: utilizing the advantage of low-frequency signals in the comprehensive representation model, which can reflect the long-term fluctuation characteristics of climate, the low-frequency signals in TMY are replaced, thereby enhancing TMY's characterization of seasonal changes and diurnal temperature range, making it more accurate and representative in building heat transfer analysis and energy consumption prediction.

[0039] First, the dynamic heat transfer model is solved using the finite difference method. In the implicit difference method, the solution for the future time step is obtained by solving a system of linear equations that include unknowns for the future time step. This means that the solution for the future time step is implicit in the equations and needs to be solved. Although the implicit difference method performs well in terms of stability and allows for larger time steps, it usually requires iterative solutions due to the need to solve a system of linear or nonlinear equations, resulting in high computational complexity. Furthermore, the implementation is relatively difficult due to the involvement of solving a system of equations.

[0040] This invention addresses the large data volume and long-time domain problems of long-term heat conduction by employing a more stable implicit difference method suitable for long-term simulations. By introducing an adaptive time step control strategy (ATSCM) into the implicit difference model, not only is the stability of the numerical solution maintained, but computational efficiency is also optimized, thereby improving the overall versatility and robustness of the model. ; ; The numerical format for implicit difference methods can be expressed as: ; thermal diffusivity With the density of the material and specific heat capacity The relationship between them can be expressed by the following formula: ; Summarized as follows: ; ; S ; Summarized as follows: In the formula, express i Node at n Temperature at any given time, in °C; K This represents thermal conductivity, measured in W / (m·K). This represents the thermal diffusivity, expressed in W / (m·K). S Wall area, unit m 2 x represents the spatial variable of the temperature field, in meters (m); t represents the time variable in the temperature field, in seconds (s). x represents the spatial step size, the distance between two adjacent discrete spatial points, in meters; t represents the time step, the time interval between two adjacent time steps, in seconds; ρ represents the density of the wall material, in kg / m³; c represents the specific heat capacity of the wall material, in J / (kg·K); It can be explained in a physical sense This represents the energy transformation between node i-1 and node i at time k+1. Let i be the energy change between node i and node i+1 at time k+1. Let i be the energy transformation of node i from time k to time k+1.

[0041] Boundary conditions are a crucial step in solving difference equations. This invention employs dynamic boundary conditions to simulate indoor and outdoor environments. Indoor boundary conditions: In practice, the dynamic heat transfer boundary conditions of the building envelope are mixed conditions, namely, Type II and Type III boundary conditions. Considering that the temperature of the inner wall is relatively stable due to radiation and convection, based on the empirical formula in the ASHRAE standard, the indoor flow heat transfer coefficient is taken as 3.5. The indoor boundary conditions are illustrated using Type III boundary conditions as an example, as shown in the following equation: ; Outdoor boundary conditions: Due to the lack of solar radiation data, traditional outdoor boundary condition calculation methods utilize the comprehensive outdoor temperature. T sa This paper calculates solar radiation heat flux, simplifying the role of solar radiation in boundary conditions. It differs from the traditional method of directly using outdoor composite temperature. T saAs a boundary condition, solar radiation is considered as a separate heat flow term. During the dynamic energy transfer process on the exterior wall surface, heat transfer includes evaporation, convection, and radiation heat exchange with the air, as well as heat transfer with adjacent nodal walls. Heat transfer coefficients that vary with climate change, such as the convective heat transfer coefficient, are introduced. The value is taken from the empirical formula in the ASHRAE standard, evaporative heat transfer coefficient. The values ​​are taken from the Lewis number (Lewis Relation), the general relationship between evaporation and convective heat transfer coefficients on a water-bearing surface, and the long-wave radiation from the external wall surface to the environment. Integrating three types of boundary conditions, the temperature equilibrium equation for the external wall surface is established as follows: ; ; ; = + ; ; In the formula, This represents the evaporative and convective heat transfer coefficient between the inner wall surface and the interior space, expressed in W / (m²). 2 ·K); This represents the evaporative and convective heat transfer coefficient between the exterior wall surface and the outside, expressed in W / (m²). 2 ·K); express t Evaporative heat transfer coefficient of the outer wall at any given time. Unit: W / (m²) 2 ·K); This represents the convective heat transfer coefficient of the outer wall at time t. The unit is W / (m²). 2 ·K); P Indicates atmospheric pressure. Unit: hPa; Relative humidity. Unit: % This represents the long-wave radiation per unit area of ​​the wall, measured in W / m². 2 ; 1 represents the wall's absorption rate; 2 represents the emissivity of the wall; This represents the Stefan-Boltzmann constant, approximately 5.67 × 10⁻⁶. -8 (W / m) 2 / K 4 ); Part Two: After obtaining the wall temperature according to the above steps, EEMD was used to process the two sets of meteorological data for Xi'an area, one for day and one for night. Different meteorological parameters from the two data sets were decomposed into several Intrinsic Mode Functions (IMFs) and a residual sequence, as shown below. Figure 2 As shown in the diagram. IMF1-IMF4 are high-frequency components, capable of capturing intraday temperature fluctuations or changes caused by local weather events; IMF5-IMF8 are low-frequency components, capable of capturing monthly temperature fluctuations and revealing seasonal temperature variations; the residual sequence represents long-term climate trends. Considering that low-frequency components typically reflect long-term trends and periodic variations in the data, this paper selects the four low-frequency IMFs (IMF5 to IMF8) and accumulates them with the residual sequence to obtain the Periodic Variation Series (PVS) of daytime outdoor temperature parameters. Figure 3 and Figure 4 The two PVS sets shown effectively preserved the main periodic variation characteristics of diurnal meteorological parameters, while removing high-frequency noise and short-term fluctuations.

[0042] After obtaining the PVS (Performance Variation Score), to further quantify and describe the variation patterns of various outdoor meteorological parameters during the daytime in Xi'an, the Least Squares Method was used to fit the PVS. The Least Squares Method is a commonly used parameter estimation method; by minimizing the sum of squared errors between the fitted function and the observed data, the optimal parameters of the fitted model can be effectively solved. The specific steps are as follows.

[0043] Step S1: First, collect daytime outdoor temperature and other relevant meteorological parameters data for the Xi'an area to ensure data integrity and accuracy. Then, apply the EEMD method to decompose these time series data, extracting individual IMFs and residual sequences.

[0044] ; In the formula, The original signal; For modal functions; It is a residual sequence.

[0045] Step S2: Given that IMF5 to IMF8 contain the main low-frequency information in the data, representing the long-term trend and periodic changes of meteorological parameters, these four IMFs are selected and accumulated with the residual sequence to form PVS. This process can effectively remove high-frequency noise and retain the main variation characteristics of meteorological parameters.

[0046] The periodic transformation sequence model is as follows: ; In the formula, These are the fitted values ​​of the annual average periodic variation sequence of various meteorological parameters, in °C. These are the annual average characteristic values ​​of various meteorological elements, in °C. This represents the annual average periodic fluctuation amplitude of various meteorological parameters, in °C. Indicates the sequence number of days; The constant represents the initial phase of the characteristic expression; It represents half a period of fluctuation in meteorological parameters, with the unit being seconds (s).

[0047] Step S3: Based on the characteristics of PVS, capture the periodic changes of meteorological parameters. By comparing the fitted curve with the measured data, use the Pearson correlation coefficient R to evaluate the periodic characteristics of each outdoor meteorological parameter. If the fitting effect meets expectations, it indicates that the constructed general expression can well describe the periodic variation characteristics of outdoor meteorological parameters during the daytime in Xi'an; otherwise, it is necessary to reselect the model or adjust the parameters to improve the fitting accuracy.

[0048] ; In the formula Indicates the measured value. This represents the calculated value. This represents the average value of the measurements. This indicates that the average value has been calculated.

[0049] Step S4: The meteorological data are dimensionless by using Z-score normalization. This method not only preserves the original distribution characteristics of the data but also enhances the comparability between different parameters, providing a reliable data foundation for the subsequent expression of the comprehensive characterization formula.

[0050] ; ;

[0051] In the formula, a The meteorological parameters are numbered, a = 1, 2, 3, 4, 5; y a Indicates the values ​​of different meteorological parameters; This represents the sample average value of different meteorological parameters; This represents the standardized value of meteorological parameter a based on daytime data. Daytime hourly serial number =1,2,..., 4384.

[0052] Step S5: Considering the influence of historical data and external meteorological factors, the SARIMAX model is adopted to more comprehensively and accurately reflect the relationship between meteorological parameters and wall temperature. The contribution weights of each meteorological parameter are objectively determined using statistical methods, avoiding interference from subjective experience and improving the scientific rigor of the model. The SARIMAX model is as follows: ; In the formula The autoregressive part aims to capture the difference between the current value and the previous value. The difference part is used to transform a non-stationary time series into a stationary series; This is the moving average component, used to capture the relationship between the current value and past error terms in the data; For the exogenous variable part, other factors that affect the regression results were added.

[0053] This indicates the effect of the lagged term on the current value; L is the lag operator, representing the dependency on past time points; s Indicates the length of a season (cycle); d Represents non-seasonal differences d Secondly, it is used to eliminate trends (to stabilize the mean of the sequence); D Indicates seasonal difference D This is used to eliminate periodic seasonal effects; Indicates in t The raw time series values ​​observed at each moment; q Indicates non-seasonal AR order; Q Indicates the seasonal AR level; Represents the seasonal AR coefficient, with the f-th season lag for the... The impact; Represents white noise; Represents an exogenous variable vector of length n (such as temperature, humidity, customs, etc.); This indicates the marginal effect of each exogenous variable corresponding to the regression coefficient.

[0054] Step S5: Fusing CMIP low-frequency signals into the SARIMAX model overcomes the problems of discontinuity and insufficient low-frequency information representation in traditional TMY data, improving data representativeness and accuracy. The optimized CMIP-TMY data more realistically reflects long-term meteorological trends and reduces interference from anomalous data. The comprehensive characterization model for outdoor meteorological parameters is as follows: ; Where 'a' is the meteorological parameter number (a=1,2,3,4,5); For hourly serial numbers (T=1,2,...,8760); Weighting factors for various meteorological parameters during the day; These are the weighting factors for various meteorological parameters at night.

[0055] The periodic expressions for the daytime outdoor meteorological parameters in Xi'an are shown in Table 3, and the periodic expressions for the nighttime outdoor meteorological parameters are shown in Table 4. The Pearson correlation coefficient R is used to evaluate the correlation between the raw meteorological data and the periodic expressions of each meteorological parameter.

[0056] Table 1. Expressions of periodic characteristics of meteorological parameters during the day.

[0057] Table 2. Expressions of periodic characteristics of meteorological parameters at night

[0058] When processing the SARIMAX model, the Akaike Information Criterion (AIC) is one of the commonly used information criteria in evaluating time series regression models. The smaller the AIC value, the better the overall performance of the model. Based on the ACF and PACF, candidate model parameters were selected. Then, using the Box-Jenkins tool on the Python platform, the AIC of the candidate models was optimized (see Table 5), ultimately determining the specific model parameters. After model training, predictions were made, and the mean absolute error between the predicted and actual values ​​was calculated to evaluate the model's accuracy. After training, the model coefficients were retrieved as follows: Figure 5 As shown, these coefficients reflect the relative importance of each meteorological parameter in predicting the inner wall temperature.

[0059] Table 3. Variation of AIC values ​​with different models

[0060] Taking Xi'an as an example further, following the above steps for comprehensive characterization, the final comprehensive characterization of outdoor meteorological parameters for the whole year is as follows: ; Table 4 Comprehensive Characterization Parameters of Xi'an During the Day

[0061] Table 5 Comprehensive Characteristic Parameters of Xi'an Nighttime Environment

[0062] Part Three: Please refer to Figure 6After improving the meteorological parameters of TMY using the Xi'an Regional Integrated Characterization Model (CMIP), the correlation coefficients with the measured data were all improved, especially for parameters with significant seasonal variations such as solar radiation and wind speed. These results indicate that the typical meteorological year data improved using CMIP exhibits higher accuracy and can more accurately reflect actual climate change trends. Compared with traditional typical meteorological year data, the corrected data not only significantly improves the temperature fit but also better captures the transitional characteristics between seasons. This correction method effectively compensates for the deficiencies of the original meteorological year data in terms of seasonal variations and outliers, providing a temperature model that more closely matches actual climate characteristics, particularly in detail, more closely reflecting actual climate dynamics.

[0063] In summary, this invention, based on the mechanism of outdoor climate action on the exterior wall, constructs a dimensionless representation that simultaneously reflects the comprehensive effects of multiple meteorological elements by quantifying the weights of factors such as temperature, humidity, radiation, and wind speed. This representation was then used to improve the database of typical meteorological years.

[0064] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A comprehensive method for characterizing building outdoor climate based on the fusion of multiple meteorological parameters, characterized in that, The meteorological parameters are divided into different time scales based on day and night, and then the fluctuation characteristics of the meteorological parameters are extracted using ensemble empirical mode decomposition. A seasonal difference autoregressive moving average with exogenous regression model is introduced to determine the weighting factors of each meteorological parameter and to achieve a comprehensive characterization of outdoor meteorological parameters, thus obtaining a comprehensive characterization model of building outdoor climate. The low-frequency signal in the building outdoor climate integrated characterization model is used to represent the long-term climate fluctuation characteristics, and the low-frequency signal in TMY is replaced.

2. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 1, characterized in that, The specific steps for the comprehensive characterization of outdoor meteorological parameters are as follows: S1: Data Cleaning Obtain raw meteorological parameters, including outdoor temperature, total solar radiation, wind speed, wind direction, atmospheric pressure, and relative humidity, and perform outlier processing on the meteorological parameter dataset; S2: Dynamic heat transfer through the wall A dynamic heat transfer model of the wall, consisting of the heat conduction equation and boundary conditions, is established, and the inner and outer wall temperatures of the wall are calculated using the dynamic heat transfer model. S3: Temporal Characteristics of Meteorological Data Using a conditional selection method, the time series and data of solar radiation not equal to zero were divided into the total number of daytime hours D and the daytime dataset D. a,m The time series data where solar radiation is zero is divided into the total number of nighttime hours N and the daytime dataset N. a,m ;in a Numbering meteorological parameters a =1,2,3,4,5 m For the number of days; S4: Extraction of meteorological data fluctuation characteristics For the daytime and nighttime datasets, EEMD was used to decompose the time series data of various outdoor meteorological parameters to obtain multiple intrinsic mode functions and a residual sequence for each meteorological parameter. Four low-frequency components, IMF5 to IMF8, were selected and added to the residual sequence to form a periodic variation sequence PVS. The correlation coefficient R was used to evaluate the correlation between the obtained periodic fluctuation expressions of each meteorological element and the original data. S5: Dimensional Standardization of Meteorological Parameters The Z-score standardization method is used to process the meteorological parameters, resulting in a standardized daytime calculation formula; the same applies to nighttime calculations. S6: Determination of weights for various meteorological parameters The diurnal weighting coefficients of each meteorological parameter are determined by extending the statistical regression model using a seasonal autoregressive integral moving average model. and ; S7: Comprehensive Characterization of Outdoor Meteorological Parameters Determining the weighting factors for each meteorological parameter and Then, a comprehensive representation based on the annual outdoor climate cycle fluctuation characteristics is obtained by weighted summation, and a dimensionless index representing the outdoor climate cycle fluctuation is constructed.

3. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S2, the heat transfer equation for the dynamic heat transfer model of the wall is: After performing implicit difference on it, we get the following equation: ； Summarized as follows: ; In the formula, express i Node at n Temperature at any given time, in °C; K This represents thermal conductivity, measured in W / (m·K). This represents the thermal diffusivity, expressed in W / (m·K). S Wall area, unit m 2 ; x represents the spatial variable of the temperature field, in meters (m); t represents the time variable in the temperature field, in seconds (s). x represents the spatial step size, the distance between two adjacent discrete spatial points, in meters; t represents the time step, the time interval between two adjacent time steps, in seconds; ρ represents the density of the wall material, in kg / m³; c represents the specific heat capacity of the wall material, in J / (kg·K); The heat transfer equation for the dynamic heat transfer model of the wall is: ； ； ； = + ； ； In the formula, This represents the evaporative and convective heat transfer coefficient between the inner wall surface and the interior space, expressed in W / (m²). 2 ·K); This represents the evaporative and convective heat transfer coefficient between the exterior wall surface and the outside, expressed in W / (m²). 2 ·K); express t Evaporative heat transfer coefficient of the outer wall at any given time, in W / (m²) 2 ·K); This represents the convective heat transfer coefficient of the outer wall at time t, in W / (m²). 2 ·K); P Atmospheric pressure, unit: hPa; Relative humidity, expressed in % %. This represents the long-wave radiation per unit area of ​​the wall, measured in W / m². 2 ; 1 represents the wall's absorption rate; 2 represents the emissivity of the wall; The Stefan-Boltzmann constant is 5.67 × 10⁻⁶. -8 (W / m) 2 / K 4 ).

4. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S3, the model for classifying the diurnal time characteristics of meteorological data is as follows: ； In the formula, m For the number of days sequence number ( m =1,2,...,365); For the first m The number of daytime hours in a day, expressed in hours (h). For the first m The number of night hours in a day, in hours (h). D Total sunshine hours, in hours (h). N Total nighttime hours, in hours (h). Q Solar radiation, measured in W / m.

5. The comprehensive characterization method for building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S4, the model for extracting the temporal fluctuation features of meteorological data is as follows: ； In the formula, The original signal; For modal functions; It is a residual sequence; The periodic transformation sequence model is as follows: ； In the formula, These are the fitted values ​​of the annual average periodic variation sequence of various meteorological parameters, in °C. These are the annual average characteristic values ​​of various meteorological elements, in °C. This represents the annual average periodic fluctuation amplitude of various meteorological parameters, in °C. Indicates the sequence number of days; The constant represents the initial phase of the characteristic expression; It represents half a period of fluctuation in meteorological parameters, with the unit being seconds (s). The correlation coefficient R is calculated as follows: ； In the formula, and These are the calculated value and the measured value, respectively. and These represent the average of the calculated value and the measured value, respectively.

6. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S5, the Z-score standardized model is: ； ； ； In the formula ，y a Indicates the values ​​of different meteorological parameters; This represents the sample average value of different meteorological parameters; This represents the standardized value of meteorological parameter a based on daytime data. Daytime hourly serial number =1,2,..., 4384.

7. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S6, the SARIMAX model is: ; In the formula The autoregressive part aims to capture the difference between the current value and the previous value. The difference part is used to transform a non-stationary time series into a stationary series; This is the moving average component, used to capture the relationship between the current value and past error terms in the data; For the exogenous variable part, other factors that affect the regression results were added; This indicates the effect of the lagged term on the current value; L is a lag operator, representing the dependency on past points in time; s Indicates the length of a season (cycle); d Represents non-seasonal differences d Secondly, it is used to eliminate trends (to stabilize the mean of the sequence); D Indicates seasonal difference D This is used to eliminate periodic seasonal effects; Indicates in t The raw time series values ​​observed at each moment; q Indicates non-seasonal AR order; Q Indicates the seasonal AR level; Represents the seasonal AR coefficient, with the f-th season lag for the... The impact; Represents white noise; Represents an exogenous variable vector of length n (such as temperature, humidity, customs, etc.); This indicates the marginal effect of each exogenous variable corresponding to the regression coefficient.

8. The method for comprehensive characterization of building outdoor climate based on the fusion of multiple meteorological parameters according to claim 2, characterized in that, In step S7, the comprehensive characterization model for outdoor meteorological parameters is as follows: ； Where 'a' is the meteorological parameter number, a = 1, 2, 3, 4, 5, n = 5; For the hourly sequence number, T=1,2,...,8760; Weighting factors for various meteorological parameters during the day; These are the weighting factors for various meteorological parameters at night.

Images