Building global thermal state characterization method based on equivalent room temperature model

Abstract

The application discloses a kind of based on equivalent room temperature model's building global thermal state characterization method, it is related to building thermal environment evaluation technical field;The operation data of each partition of building is collected to construct 5-dimensional feature vector and is standardized, clustering algorithm is applied to the operation data of building cooling condition adaptive segmentation and is endowed with each operation period reference weight, through weighted average mode, the correlation coefficient of room temperature-cooling load and the area proportion of partition are fused to calculate integrated weight coefficient, based on the phase lag effect of cooling system, introduce real-time dynamic response factor to carry out adaptive adjustment, fusion operation period reference weight, partition integrated weight coefficient and real-time dynamic response factor are used to construct integrated weight coefficient, and the equivalent room temperature of whole building is calculated by nonlinear weighted fusion calculation of each partition room temperature using integrated weight coefficient;The application uses the above-mentioned one based on equivalent room temperature model's building global thermal state characterization method, realizes the multidimensional description to the complex thermal characteristics of building.

Description

Technical Field

[0001] This invention relates to the field of building thermal environment assessment technology, and in particular to a method for characterizing the global thermal state of a building based on an equivalent room temperature model. Background Technology

[0002] In central air conditioning systems of public buildings, indoor temperature is not only a direct indicator of thermal comfort but also a core thermodynamic state parameter characterizing the building's actual cooling load demand. However, traditional cooling load estimation methods have significant theoretical limitations in actual operation. On the one hand, traditional physical white-box models are mostly based on ideal operating conditions, making it difficult to dynamically capture time-varying factors such as human activity, equipment disturbances, and the degradation of the building envelope's thermal performance, resulting in prediction errors often exceeding 15%. On the other hand, current data-driven studies often use the "cooling capacity method," which directly substitutes the cooling load with the equivalent output cooling capacity of the chiller unit. This method ignores the unsteady heat transfer characteristics between the system's cooling supply and terminal demand, fails to consider the degree of deviation of indoor temperature from the setpoint, and neglects the phase lag effect caused by the thermal inertia of the cooling network and the building. When the indoor thermal environment fluctuates, the transient cooling capacity often fails to accurately reflect the building's dynamic cooling load gap.

[0003] Therefore, relying solely on historical cooling capacity for prediction and control can easily lead to a mismatch between supply and demand. It is necessary to introduce indoor temperature deviations to dynamically correct the cooling load. Summary of the Invention

[0004] The purpose of this invention is to provide a method for characterizing the global thermal state of a building based on an equivalent room temperature model. This method aims to integrate spatially dispersed multi-dimensional room temperature monitoring data into a global equivalent room temperature through nonlinear weighting, so as to more accurately depict the evolution of transient cooling load demand of the entire building, thereby providing a solid data foundation for high-precision prediction of cooling load and closed-loop feedback correction.

[0005] To achieve the above objectives, this invention provides a method for characterizing the global thermal state of a building based on an equivalent room temperature model, comprising the following steps: S1. Collect operational data from each zone of the building to construct a 5-dimensional feature vector. And standardize it; S2. Apply the K-means clustering algorithm to adaptively segment the operating data of building cooling conditions collected in S1 and assign a baseline weight to each operating segment. S3. Calculate the Pearson correlation coefficient between room temperature and cooling load by combining the zonal operation data collected in S1. Then, calculate the comprehensive weight coefficient by merging the Pearson correlation coefficient between room temperature and cooling load and the zonal area ratio through a weighted average method. S4. Based on the phase lag effect of the cooling system, a real-time dynamic response factor is introduced for adaptive adjustment. S5. The final weight coefficient is constructed by integrating the runtime baseline weight in S2, the comprehensive weight coefficient of the partition in S3, and the real-time dynamic response factor. S6. Using the final weighting coefficients from S5, perform a non-linear weighted fusion calculation on the room temperature of each zone to obtain the equivalent room temperature of the entire building.

[0006] Preferably, the specific process of S1 is as follows: S11. Collect operational data such as room temperature, chilled water inlet and outlet temperature difference, flow rate, and timestamp for each zone of the building. S12. Based on the operational data collected in S11, construct a 5-dimensional feature vector including cooling capacity, cooling capacity change rate, average room temperature, room temperature change rate, and normalized time. and ; S13, the 5-dimensional eigenvector of S12 Perform Z-score standardization.

[0007] Preferably, the specific process of S2 is as follows: S21. Input the standardized feature vectors into the k-means algorithm and compare different... k Silhouette coefficient, DB index, CH index under the specified values; S22. Select a profile coefficient with a small DB index and a large CH index. k The value is denoted as a As the best k The value, according to the k-means algorithm, is in the optimal... k The results of clustering under the value are divided into a Each time period is assigned an appropriate weight based on the actual application.

[0008] Preferably, based on the clustering results of S22, a baseline weight is set for each of the three runtime stages. The baseline weight is set according to the principle that the more stable the system operation and the more reliable the data quality, the higher the baseline weight; the greater the system fluctuation, the lower the baseline weight.

[0009] Preferably, the specific process of S3 is as follows: S31, Assume the building includes The functional partition is calculated. Average room temperature sequence of each zone Cooling load of the whole building In the runtime segment Pearson correlation coefficient on The specific formula is as follows: ; in, This is a general representation of the correlation coefficient; S32, Calculate the first The percentage of building area in each zone The specific formula is as follows: ; in, For the first The building area of ​​each zone This represents the total building area of ​​all zones. The total number of partitions; S33. By using a weighted average method, the correlation between room temperature and cooling load is integrated with the area ratio of each zone to obtain the... Weight coefficients of each partition The specific formula is as follows: ; in, For all zones, room temperature With the overall cooling load In the runtime segment The sum of the Pearson correlation coefficients within the range.

[0010] Preferably, the real-time dynamic response factor introduced in S4 The specific expression is as follows: ; in, runtime segment The weighting coefficients, for The cooling capacity of the entire building at all times. runtime segment Reference cooling load, For the first Temperature deviation in each zone and , For the first The actual temperature of each zone For the first The set temperature for each zone, The average temperature deviation across all zones and .

[0011] Preferably, the specific formula for the final weighting coefficients constructed in S5 is as follows: ; in, For the first Each partition The final weighting coefficient at each time step. The baseline weights for different runtime stages.

[0012] Preferably, the equivalent room temperature of the entire building in S6 The calculation formula is as follows: ; in, For the first Each partition at time The room temperature monitoring value.

[0013] Therefore, the present invention provides a method for characterizing the global thermal state of a building based on an equivalent room temperature model, which, compared with the prior art, has the following advantages: 1. This application organically combines the time period dimension (system stability), regional dimension (relative cooling load contribution), and operating condition dimension (real-time operating condition changes) into a three-layer weighting system, realizing a multi-dimensional characterization of the complex thermal characteristics of buildings; in the absence of zonal cooling load sensors, the cooling load contribution weight of each zone is inferred through correlation analysis, taking into account both physical rationality and engineering feasibility. 2. Compared with traditional methods, this application improves the correlation coefficient with cooling load by more than 20%, laying an accurate foundation for subsequent cooling load correction; by using time-segmented dynamically weighted equivalent room temperature, it can more accurately capture the actual thermal environment of the building, providing reliable input data for subsequent cooling load prediction and dynamic correction.

[0014] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0015] Figure 1 This is an overall flowchart of a method for characterizing the global thermal state of a building based on an equivalent room temperature model, according to the present invention. Figure 2 This is a comparison chart of the three major evaluation indicators of a building global thermal state characterization method based on an equivalent room temperature model according to the present invention. Figure 3 This is a box plot distribution analysis diagram of a building global thermal state characterization method based on an equivalent room temperature model according to the present invention. Figure 4 This is a statistical chart showing the monthly evolution of Pearson correlation coefficients and data scale of each partition in a global building thermal state characterization method based on an equivalent room temperature model, according to the present invention. Figure 5 This invention provides a method for characterizing the global thermal state of a building based on an equivalent room temperature model, and includes a diagram showing the analysis of room temperature and cooling load in different building zones. Figure 6 This invention relates to a method for characterizing the global thermal state of a building based on an equivalent room temperature model, with time-period benchmark weights under different operating conditions. The dynamic evolution curve; Figure 7 This invention relates to a typical daily temperature deviation in a method for characterizing the global thermal state of a building based on an equivalent room temperature model. With time period weight The change characteristic diagram; Figure 8 This is a diagram showing the calculation results of dynamic factors of north-facing, south-facing, and public area rooms based on indoor temperature fluctuations, according to a global thermal state characterization method for buildings based on an equivalent room temperature model, as presented in this invention. Figure 9 This is a spatiotemporal evolution characteristic analysis diagram of the comprehensive weight coefficient of different regions in a building global thermal state characterization method based on an equivalent room temperature model according to the present invention. Figure 10 This invention presents a method for characterizing the global thermal state of a building based on an equivalent room temperature model, showing the daily temperature variation characteristics of different functional zones. Figure 11 This is a time series diagram of the equivalent room temperature deviation of a building global thermal state characterization method based on an equivalent room temperature model according to the present invention. Detailed Implementation

[0016] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product of this invention is usually placed when in use. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0017] Example like Figures 1-11 As shown, the present invention provides a method for characterizing the global thermal state of a building based on an equivalent room temperature model, comprising the following steps: S1. Collect operational data from each zone of the building to construct a 5-dimensional feature vector. And standardize it; S11. Collect operational data such as room temperature, chilled water inlet and outlet temperature difference, flow rate, and timestamp for each zone of the building. S12. Based on the operational data collected in S11, construct a 5-dimensional feature vector including cooling capacity, cooling capacity change rate, average room temperature, room temperature change rate, and normalized time. and As shown in Table 1; Table 1. Detailed definition of parameters for each dimension;

[0018] S13, the 5-dimensional eigenvector of S12 Perform Z-score standardization; make the mean of all features 0 and the standard deviation 1, as shown in the following formula: ; in, These are the standardized eigenvalues. For eigenvalues, Features The mean on the training set, Features The standard deviation on the training set. This is the general formula for the above five-dimensional features; S2. Apply the K-means clustering algorithm to adaptively segment the operating data of building cooling conditions collected in S1 and assign a baseline weight to each operating segment. S21. Input the standardized feature vectors into the k-means algorithm and compare different... k Silhouette coefficient, DB index, CH index under the specified values; Silhouette coefficient: measures the similarity of a sample to its own cluster and its distance from other clusters. The value range is [-1, 1]. The closer to 1, the better the clustering effect. Generally, 0.7 or above is excellent, 0.5-0.7 is good, and 0.3-0.5 is average. The Davies-Bouldin (DB) index measures the ratio of intra-cluster compactness to inter-cluster separation. A smaller value indicates better performance; typically, DB < 1 indicates excellent performance, and DB < 1.5 indicates good performance. The CH (Calinski-Harabasz) index is a statistical measure to evaluate the quality of clustering results. The CH coefficient is based on the ratio of intra-cluster dispersion to inter-cluster dispersion. A higher CH value indicates a more obvious clustering structure with greater inter-cluster differences and smaller intra-cluster differences. Therefore, good clustering results should have a higher CH coefficient. S22. Select a profile coefficient with a small DB index and a large CH index. k The value is denoted as a As the best k The value, according to the k-means algorithm, is in the optimal... k The results of clustering under the value are divided into a Each time period is assigned an appropriate weight based on the actual application. K-means clustering analysis was performed on the operational data of a building from 7:00 to 18:00. To determine the optimal number of clusters, this paper comprehensively compared the changes of three evaluation indicators when k∈[2,7]. Figure 2As shown, the three indicators reached a consistent optimal level when k=3: the silhouette coefficient reached its maximum value of 0.521 when k=3, the Davies-Bouldin index achieved a relatively small value of 0.589, and the CH index reached as high as 7945.00. The common indication of these three indicators fully proves the rationality of dividing the data into 3 categories. At the same time, the clustering results of k=3 correspond exactly to the three real operating stages of the central air conditioning system in a day: "start-up, stabilization, and decline". This shows that the division has practical significance in both statistics and practical engineering. Therefore, this study finally determined k=3 as the optimal number of clusters. Combining clustering results with Figure 3 The box plot distribution can map these three data clusters to the actual physical operation stages of the central air conditioning system, as shown in Table 2: Start-up phase (41.90%): From Figure 3 It can be seen that the cooling load distribution is extremely wide, showing a trend of rapid increase from a low value; at this time, the chiller has just started from a stopped state and needs to output a large amount of cooling capacity to reduce the indoor temperature, so the room temperature drops significantly and the system is in a dynamic transient process. Stable phase (53.10%): The room temperature data distribution is the most concentrated, with the median closely following the set value; although the cooling load is in the high range during this phase, the box narrows, indicating that the system has entered the optimal thermodynamic steady state; the data quality is the highest during this phase, which is the core reliable data segment for extracting the subsequent equivalent room temperature correction coefficient; Attenuation phase (4.90%): Near the end of the workday, the cooling load drops sharply, the room temperature distribution in the enclosure widens again and shows an upward trend, reflecting the uncontrolled system response after the unit is unloaded; Table 2. Analysis of the characteristics of air conditioning load operation stages;

[0019] Based on the clustering results of S22, benchmark weights were set for the three runtime phases. The benchmark weights were set according to the principle that the more stable the system operation and the more reliable the data quality, the higher the benchmark weight; the greater the system fluctuation, the lower the benchmark weight. As shown in Table 3, the benchmark weight for the stable phase was set to 1.00 (the system is optimal, so the weight is the highest), the benchmark weight for the start-up phase was set to 0.80 (the system is unstable, so the weight is reduced accordingly), and the benchmark weight for the decay phase was set to 0.90 (between the two). This differentiated weight setting ensures that the equivalent room temperature can prioritize the most reliable stable phase data, while also taking into account the temperature information during the start-up and decay processes. Table 3. Characteristics and weighting of cooling load of air conditioning system in different time periods;

[0020] The weight of the stable period is a reference value of 1.00, because the system is most stable and has the highest weight during this period; the weights of the startup and decay periods are reduced accordingly to reflect the relative differences in data reliability. S3. Calculate the Pearson correlation coefficient between room temperature and cooling load by combining the zonal operation data collected in S1. Then, calculate the comprehensive weight coefficient by merging the Pearson correlation coefficient between room temperature and cooling load and the zonal area ratio through a weighted average method. S31, Assume the building includes The functional partition is calculated. Average room temperature sequence of each zone Cooling load of the whole building In the runtime segment Pearson correlation coefficient on The specific formula is as follows: ; in, This is a general representation of the correlation coefficient; S32, Calculate the first The percentage of building area in each zone The specific formula is as follows: ; in, For the first The building area of ​​each zone This represents the total building area of ​​all zones. The total number of partitions; S33. By using a weighted average method, the correlation between room temperature and cooling load is integrated with the area ratio of each zone to obtain the... Weight coefficients of each partition The specific formula is as follows: ; in, For all zones, room temperature With the overall cooling load In the runtime segment The sum of Pearson correlation coefficients within; like Figure 4 As shown, the correlation coefficients of each region evolve with the months: at the beginning and end of the cooling season (May and September), the correlation coefficients are low; while during the peak summer load period (July-August), the correlation coefficients converge significantly and reach their peak. This evolutionary characteristic proves that the mapping relationship between room temperature and cooling capacity is modulated by the time-varying seasonal meteorological boundary conditions. like Figure 5As shown, the south area has the highest correlation, while the north area has the lowest. Further weighting and merging of correlation coefficients and area proportions yielded the final weight allocation: north-facing rooms have a weight of 24.47%, public areas 23.50%, and south-facing rooms 52.02%. South-facing rooms, due to their largest area and the strongest influence from solar radiation, exhibit the highest sensitivity to cooling load demand. This fully reflects the importance of temperature control in south-facing rooms during this season, indicating that improving the accuracy of room temperature control and insulation performance in south-facing rooms will have the most significant impact on the building's energy consumption. Although north-facing rooms account for 30% of the area, their weight was adjusted to 24.47% due to their lowest cooling load sensitivity (correlation coefficient absolute value only 0.343), indicating that room temperature fluctuations in this area have a relatively weak driving effect on overall cooling load demand. The significant differences in weight distribution among the three zones highlight the different impacts of different orientations and functional characteristics on the building's cooling load. S4. Based on the phase lag effect of the cooling system, a real-time dynamic response factor is introduced for adaptive adjustment. The real-time dynamic response factor introduced here The specific expression is as follows: ; in, runtime segment The weighting coefficients, for The cooling capacity of the entire building at all times. runtime segment Reference cooling load, For the first Temperature deviation in each zone and , For the first The actual temperature of each zone For the first The set temperature for each zone, The average temperature deviation across all zones and Weighting coefficients are applied to the thermal response characteristics of different operating periods. Implement dynamic adaptive adjustment: assign a lower value during the highly unstable startup phase of the system ( =0.40) to reinforce the dominance of the temperature term; while in the stable operating period ( =0.65), then the load term dominates; this mechanism constitutes a feedforward compensation closed loop based on temperature difference. When a sudden overcooling / overheating occurs locally, this factor surges instantaneously, amplifying the representation weight of local anomalies in the global equivalent room temperature. S5. The final weight coefficient is constructed by integrating the runtime baseline weight in S2, the comprehensive weight coefficient of the partition in S3, and the real-time dynamic response factor. The specific formula for the final weighting coefficients is as follows: ; in, For the first Each partition The final weighting coefficient at each time step. The baseline weights for different runtime phases; S6. Use the final weighting coefficients from S5 to perform a non-linear weighted fusion calculation on the room temperature of each zone to obtain the equivalent room temperature of the entire building. The equivalent room temperature of the entire building The calculation formula is as follows: ; in, For the first Each partition at time The room temperature monitoring value.

[0021] In the specific implementation process, based on the aforementioned three-layer spatiotemporal coupled dynamic weighted framework, the global equivalent room temperature is... The complete calculation process can be divided into the following five steps: 1. Confirm Runtime Adaptive Recognition and Benchmark Weighting: The algorithm reads the current operating condition characteristics in real time and identifies the current time period. (Startup / Stable / Decrease), and map the corresponding time period baseline weights. ; Figure 6 and Figure 7 This indicates the dynamic evolution characteristics of the weight. In the non-steady-state stage where the system's thermal response is severe, the model reduces the weight to effectively suppress noise interference in the input data. 2. Local dynamic factor extraction based on steady-state deviation: For the three physical partitions, the system calculates the real-time dynamic response factor by substituting the formulas into each partition. ; Figure 8 This represents the time-domain non-stationary fluctuations of dynamic factors in the North Zone, South Zone, and Public Zone; 3. Final weighting coefficient aggregation: By integrating the three-dimensional features of time reference, spatial inherent contribution, and transient thermal response, the final weighting coefficients for each region are calculated. ; Figure 9 The spatiotemporal evolution characteristics of the final weight coefficients are shown, with the southern region having the highest weight baseline and exhibiting the most dramatic fluctuations. 4. Equivalent room temperature reconstruction: Outputs the nonlinear weighted equivalent room temperature of the entire building based on the equivalent room temperature calculation formula; Figure 10 The daily average variation patterns of each independent zone and the equivalent room temperature were compared; the equivalent room temperature smoothed out local extreme values ​​while keenly preserving the true evolution trend of the global cooling load. 5. Overall cooling capacity supply and demand matching assessment: Compare the temperature with the reference value of 26℃ to assess the overall cooling load requirement of the building. Figure 11 The statistical results of the entire cooling season (13,599 valid sample points) in Tables 4 and 5 show that there is a significant supply-demand mismatch in the system: the average deviation over the entire cycle is 0.694℃, the distribution shows a severe right skewness (the cooling shortage rate is as high as 65.4%), and the maximum transient positive deviation is 4.741℃.

[0022] Table 4. Statistical analysis of temperature deviation;

[0023] Table 5 Monthly Year Summary Table;

[0024] Therefore, the present invention adopts a building global thermal state characterization method based on an equivalent room temperature model, which aims to integrate spatially dispersed multi-dimensional room temperature monitoring data into a global equivalent room temperature through nonlinear weighting, so as to more accurately depict the transient cooling load demand evolution of the entire building, thereby providing a solid data foundation for high-precision prediction of cooling load and closed-loop feedback correction.

[0025] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for characterizing the global thermal state of a building based on an equivalent room temperature model, characterized in that, Includes the following steps: S1. Collect operational data from each zone of the building to construct a 5-dimensional feature vector. And standardize it; S2. Apply the K-means clustering algorithm to adaptively segment the operating data of building cooling conditions collected in S1 and assign a baseline weight to each operating segment. S3. Calculate the Pearson correlation coefficient between room temperature and cooling load by combining the zonal operation data collected in S1. Then, calculate the comprehensive weight coefficient by merging the Pearson correlation coefficient between room temperature and cooling load and the zonal area ratio through a weighted average method. S4. Based on the phase lag effect of the cooling system, a real-time dynamic response factor is introduced for adaptive adjustment. S5. The final weight coefficient is constructed by integrating the runtime baseline weight in S2, the comprehensive weight coefficient of the partition in S3, and the real-time dynamic response factor. S6. Using the final weighting coefficients from S5, perform a non-linear weighted fusion calculation on the room temperature of each zone to obtain the equivalent room temperature of the entire building.

2. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 1, characterized in that: The specific process of S1 is as follows: S11. Collect operational data such as room temperature, chilled water inlet and outlet temperature difference, flow rate, and timestamp for each zone of the building. S12. Based on the operational data collected in S11, construct a 5-dimensional feature vector including cooling capacity, cooling capacity change rate, average room temperature, room temperature change rate, and normalized time. and ; S13, the 5-dimensional eigenvector of S12 Perform Z-score standardization.

3. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 2, characterized in that: The specific process of S2 is as follows: S21. Input the standardized feature vectors into the k-means algorithm and compare different... k Silhouette coefficient, DB index, CH index under the specified values; S22. Select a profile coefficient with a small DB index and a large CH index. k The value is denoted as a As the best k The value, according to the k-means algorithm, is in the optimal... k The results of clustering under the value are divided into a Each time period is assigned an appropriate weight based on the actual application.

4. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 3, characterized in that: Based on the clustering results of S22, benchmark weights are set for the three runtime stages. The benchmark weights are set according to the principle that the more stable the system operation and the more reliable the data quality, the higher the benchmark weights will be. The greater the system fluctuation, the lower the benchmark weight.

5. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 4, characterized in that: The specific process of S3 is as follows: S31, Assume the building includes The functional partition is calculated. Average room temperature sequence of each zone Cooling load of the whole building In the runtime segment Pearson correlation coefficient on The specific formula is as follows: ； in, This is a general representation of the correlation coefficient; S32, Calculate the first The percentage of building area in each zone The specific formula is as follows: ； in, For the first The building area of ​​each zone This represents the total building area of ​​all zones. The total number of partitions; S33. By using a weighted average method, the correlation between room temperature and cooling load is integrated with the area ratio of each zone to obtain the... Weight coefficients of each partition The specific formula is as follows: ； in, For all zones, room temperature With the overall cooling load In the runtime segment The sum of the Pearson correlation coefficients within the range.

6. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 5, characterized in that: Real-time dynamic response factor introduced in S4 The specific expression is as follows: ； in, runtime segment The weighting coefficients, for The cooling capacity of the entire building at all times. runtime segment Reference cooling load, For the first Temperature deviation in each zone and , For the first The actual temperature of each zone For the first The set temperature for each zone, The average temperature deviation across all zones and .

7. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 6, characterized in that: The specific formula for the final weighting coefficients constructed in S5 is as follows: ； in, For the first Each partition The final weighting coefficient at each time step. The baseline weights for different runtime stages.

8. The method for characterizing the global thermal state of a building based on an equivalent room temperature model according to claim 7, characterized in that: Equivalent room temperature of the entire building in S6 The calculation formula is as follows: ； in, For the first Each partition at time The room temperature monitoring value.

Images