Method and system for dynamic simulation test of response characteristics of temperature flow field disturbance of building cabin

Abstract

The application provides a building cabin temperature flow field disturbance response characteristic dynamic simulation test method and system, relates to the technical field of simulation test, and comprises the following steps: collecting temperature and humidity air flow field data through a multilevel sensor network, dividing a grid unit into multiple sub-regions to establish a data correlation model, calculating a disturbance response prediction error according to real-time data, adaptively adjusting a sampling frequency and triggering sensor recalibration, finally calculating space-time distribution data and a disturbance propagation path, generating a response characteristic evaluation report, and realizing precise dynamic simulation test on the building cabin temperature flow field disturbance response characteristic.

Description

Technical Field

[0001] This invention relates to simulation testing technology, and more particularly to a dynamic simulation testing method and system for the response characteristics of temperature and flow field disturbances in building cabins. Background Technology

[0002] With the rapid development of green and intelligent buildings, the accurate measurement and dynamic response analysis of the temperature and humidity airflow field inside building cabins has gradually become a key research area for building energy conservation optimization and comfort improvement. The temperature and humidity airflow field inside building cabins is affected by various factors, such as human activities, equipment operation, and external climate change. These factors can cause disturbances in the temperature and humidity airflow field inside the cabin, thereby affecting the building's energy consumption and indoor environmental quality. Traditional methods for testing the temperature and humidity airflow field in building cabins mainly rely on a small number of sensors at fixed locations to collect data and perform static or quasi-static analysis through simplified models. These methods cannot reflect the dynamic changes in the temperature and humidity airflow field inside the cabin in real time and comprehensively.

[0003] Traditional testing methods typically employ fixed-location, fixed-frequency sensor deployment, lacking the ability to adjust sampling strategies based on real-time data. This results in an inability to adaptively respond to the temperature and humidity airflow disturbance characteristics in different regions, leading to the omission of important disturbance information or severe data redundancy.

[0004] Existing testing systems mostly focus on the static distribution of macroscopic temperature and humidity data, lacking the ability to effectively track and analyze the propagation path and dynamic response characteristics of disturbances inside the chamber, making it difficult to accurately characterize the spatiotemporal evolution of temperature and flow field disturbances and the influence transmission mechanism between sub-regions.

[0005] The lack of prediction error assessment and sensor network self-calibration mechanisms based on real-time data means that when the measurement system drifts or environmental conditions change significantly, the measurement parameters cannot be adjusted and the system recalibrated in a timely manner, affecting the accuracy and reliability of the test data, and consequently impacting building environment optimization decisions based on the test data. Summary of the Invention

[0006] The present invention provides a dynamic simulation test method and system for the temperature and flow field disturbance response characteristics of building cabins, which can solve the problems in the prior art.

[0007] A first aspect of the present invention provides a dynamic simulation test method for the temperature and flow field disturbance response characteristics of a building cabin, comprising:

[0008] Obtain the geometric parameters of the building cabin, divide it into grid cells according to the geometric parameters, deploy a multi-level sensor network at the nodes of the grid cells, and collect temperature and humidity airflow field data inside the building cabin; divide the grid cells into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions;

[0009] Based on the real-time data collected by the multi-level sensor network, the disturbance response prediction error of each sub-region is calculated. The sampling frequency of the multi-level sensor network is adaptively adjusted based on the disturbance response prediction error. When the disturbance response prediction error is greater than a preset error threshold, the recalibration of the multi-level sensor network is triggered.

[0010] The data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions. Based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the temperature and flow field disturbance response characteristics of the building cabin is generated.

[0011] Dividing the grid cells into multiple sub-regions and establishing a data association model between the sub-regions using the temperature and humidity airflow field data includes:

[0012] The grid cells are divided according to the geometric parameters of the building cabin. The grid cells are then divided into multiple sub-regions based on the gradient variation characteristics of the temperature and humidity airflow field. Adaptive boundary optimization is performed on the multiple sub-regions. When the difference in temperature and humidity airflow field data between adjacent sub-regions is greater than a preset error threshold, the division boundary of the sub-regions is dynamically adjusted. Based on the optimized sub-region division results, a data association model between the sub-regions is established.

[0013] Based on the real-time data collected by the multi-level sensor network, the disturbance response prediction error of each sub-region is calculated, and the sampling frequency of the multi-level sensor network is adaptively adjusted based on the disturbance response prediction error, including:

[0014] The disturbance response prediction error of each sub-region is calculated based on the real-time data collected by the multi-level sensor network. Reference measurement points are set up in each sub-region of the multi-level sensor network. The time difference of temperature and humidity change between adjacent measurement points is calculated based on the reference measurement points. The disturbance propagation rate is calculated based on the time difference of temperature and humidity change and the spatial distance between adjacent measurement points.

[0015] The sub-region is divided into a fast response region and a slow response region according to the disturbance propagation rate. The disturbance propagation rate of the fast response region is greater than a preset rate threshold, and the disturbance propagation rate of the slow response region is less than or equal to the preset rate threshold.

[0016] A response link graph is constructed, which includes a set of sub-region nodes and a set of disturbance propagation paths. The importance of a node is calculated based on the edge set propagation rate and the path length of the sub-region node. The sampling frequency of the multi-level sensor network is adaptively adjusted according to the disturbance response prediction error and the node importance.

[0017] The time difference of temperature and humidity change between adjacent measuring points is calculated based on the reference measuring point, and the disturbance propagation rate is calculated based on the time difference of temperature and humidity change and the spatial distance between the adjacent measuring points, including:

[0018] Collect temperature and humidity time series data of reference measuring points, and pair the reference measuring points to form adjacent measuring point groups; calculate the mean of temperature and humidity time series data of two reference measuring points in the adjacent measuring point group, calculate the cross-correlation function between the two reference measuring points based on the difference between the temperature and humidity time series data and the mean, and obtain the time offset corresponding to the maximum value of the cross-correlation function as the temperature and humidity change time difference of the adjacent measuring point group.

[0019] Calculate the spatial distance between two reference measuring points in the adjacent measuring point group, and divide the spatial distance of the adjacent measuring point group by the time difference of temperature and humidity changes in the adjacent measuring point group to obtain the disturbance propagation rate of the adjacent measuring point group.

[0020] The data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions, including:

[0021] A temporal attention model is constructed based on the calibrated temperature and humidity airflow field data. The temporal attention model calculates the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin by weighted superposition of the importance weight of each measuring point and the calibration data.

[0022] Based on the calibrated temperature and humidity airflow field data, the sensor measurement points in the multi-level sensor network are constructed as a measurement point association structure, the temperature and humidity airflow field data difference between adjacent sensor measurement points is calculated, and the temperature and humidity airflow field data difference is converted into the connection strength between measurement points through the time-series attention model.

[0023] Spatial features are extracted from the node features in the measurement point association structure to obtain the spatial dimension features of the nodes; the spatial dimension features are introduced into the temporal attention model, and the importance of each temporal feature is calculated based on the correlation of the feature vectors. The historical temporal features of the nodes are weighted and combined according to the importance to obtain the spatiotemporal dimension features of the nodes.

[0024] The propagation probability of disturbance between adjacent nodes is calculated based on the spatiotemporal dimension characteristics. Sensor measurement point pairs with a propagation probability greater than a preset probability threshold are selected as effective paths for disturbance propagation. The effective paths are connected into a propagation network based on the temporal attention model to track and record the propagation path of disturbance between sub-regions inside the building cabin.

[0025] Based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the disturbance response characteristics of the building cabin's temperature and flow field is generated, including:

[0026] Based on spatiotemporal distribution data, the real-time field values ​​of each sub-region along the propagation path are extracted. The deviation between the real-time field values ​​and the preset reference field values ​​is calculated to obtain the sub-region disturbance response intensity. Based on the sub-region disturbance response intensity, a sub-region response coupling matrix is ​​constructed. The sum of the squares of the differences between the sub-region disturbance response intensity and the predicted values ​​of the sub-region response coupling matrix is ​​used as the judgment criterion to optimize the correlation coefficient of the sub-region response coupling matrix and obtain the disturbance attenuation parameters of each sub-region along the propagation path.

[0027] The correlation of disturbance response intensities in adjacent sub-regions is calculated along the propagation path, and the sub-region response coupling matrix is ​​updated. The attenuation law of disturbances on the propagation path is analyzed based on the sub-region response coupling matrix. The response delay, attenuation rate, and residual intensity of each sub-region on the propagation path are calculated. An evaluation report on the disturbance response characteristics of the building cabin temperature flow field is generated based on the response delay, attenuation rate, residual intensity, and sub-region response coupling matrix.

[0028] A second aspect of the present invention provides a dynamic simulation test system for the temperature and flow field disturbance response characteristics of a building cabin, comprising:

[0029] The first unit is used to acquire the geometric parameters of the building cabin, divide the building cabin into grid units based on the geometric parameters, deploy a multi-level sensor network at the nodes of the grid units, and collect temperature and humidity airflow field data inside the building cabin; divide the grid units into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions.

[0030] The second unit is used to calculate the disturbance response prediction error of each sub-region based on the real-time data collected by the multi-level sensor network, adaptively adjust the sampling frequency of the multi-level sensor network based on the disturbance response prediction error, and trigger the recalibration of the multi-level sensor network when the disturbance response prediction error is greater than a preset error threshold.

[0031] The third unit is used to process and analyze the data collected by the calibrated multi-level sensor network, calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and track and record the propagation path of disturbances between sub-regions; based on the spatiotemporal distribution data and the propagation path, calculate the disturbance attenuation characteristics of each sub-region, and generate an evaluation report on the temperature and humidity airflow field disturbance response characteristics of the building cabin.

[0032] A third aspect of the present invention provides an electronic device, comprising:

[0033] processor;

[0034] Memory used to store processor-executable instructions;

[0035] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0036] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0037] The beneficial effects of this application are as follows:

[0038] The present invention provides a dynamic simulation test method for the temperature and humidity airflow field disturbance response characteristics of building cabins. By deploying a multi-level sensor network to collect temperature and humidity airflow field data inside the building cabin, and dividing the grid cells into multiple sub-regions to establish a data association model, it is possible to achieve high-precision monitoring and analysis of the temperature and humidity airflow field inside the building cabin, thereby improving the accuracy of disturbance response characteristic assessment.

[0039] This testing method adaptively adjusts the sampling frequency of the sensor network based on the disturbance response prediction error and triggers a recalibration mechanism when the error exceeds a threshold. This greatly enhances the system's adaptability to environmental changes, improves the real-time performance and reliability of data acquisition, reduces unnecessary data redundancy, and optimizes system resource utilization.

[0040] This invention calculates the spatiotemporal distribution data of temperature and humidity airflow fields inside a building cabin and tracks the disturbance propagation path, thereby achieving accurate calculation of the disturbance attenuation characteristics of each sub-region. Finally, it generates a comprehensive evaluation report, providing a scientific basis for the optimized design of building environmental control systems and the formulation of energy-saving strategies. It has significant practical value and economic benefits. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the dynamic simulation test method for the temperature and flow field disturbance response characteristics of a building cabin according to an embodiment of the present invention.

[0042] Figure 2 This is a flowchart illustrating the perturbation propagation rate calculation method according to an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0045] Figure 1 This is a flowchart illustrating the dynamic simulation test method for the thermal flow field disturbance response characteristics of a building cabin according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0046] Obtain the geometric parameters of the building cabin, divide it into grid cells according to the geometric parameters, deploy a multi-level sensor network at the nodes of the grid cells, and collect temperature and humidity airflow field data inside the building cabin; divide the grid cells into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions;

[0047] Based on the real-time data collected by the multi-level sensor network, the disturbance response prediction error of each sub-region is calculated. The sampling frequency of the multi-level sensor network is adaptively adjusted based on the disturbance response prediction error. When the disturbance response prediction error is greater than a preset error threshold, the recalibration of the multi-level sensor network is triggered.

[0048] The data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions. Based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the temperature and flow field disturbance response characteristics of the building cabin is generated.

[0049] In one optional implementation, the grid cells are divided into multiple sub-regions, and a data association model between the sub-regions is established using the temperature and humidity airflow field data, including:

[0050] The grid cells are divided according to the geometric parameters of the building cabin. The grid cells are then divided into multiple sub-regions based on the gradient variation characteristics of the temperature and humidity airflow field. Adaptive boundary optimization is performed on the multiple sub-regions. When the difference in temperature and humidity airflow field data between adjacent sub-regions is greater than a preset error threshold, the division boundary of the sub-regions is dynamically adjusted. Based on the optimized sub-region division results, a data association model between the sub-regions is established.

[0051] In a specific implementation, the building's geometric parameters are used to create a grid, enabling precise monitoring of the temperature and humidity airflow field and acquiring the building's geometric parameters, including length, width, height, and internal structural layout. Using 3D modeling technology, the building is constructed as a digital model, and the entire space is divided into grid cells using finite element analysis. The size of the grid cells is determined based on the required monitoring accuracy, typically set to a 0.5m × 0.5m × 0.5m cube. For special areas, such as those with large temperature gradients, the grid size can be further refined to 0.2m × 0.2m × 0.2m.

[0052] Sub-region division based on the gradient variation characteristics of temperature and humidity airflow fields improves monitoring efficiency. In practical applications, temperature and humidity airflow fields are not uniformly distributed within building cabins but exhibit gradient variation characteristics. A temperature and humidity gradient clustering algorithm is used to calculate the temperature and humidity differences between adjacent grid cells. When the difference exceeds a preset threshold (e.g., a temperature difference greater than 1.5℃ or a humidity difference greater than 5%RH), it is marked as a region with significant gradient variation. Based on the marking results, the grid cells are divided into multiple sub-regions. For example, in a 200-square-meter office space, it would be divided into 5-7 sub-regions based on temperature and humidity gradient characteristics, including areas near windows, areas around air conditioning vents, densely populated areas, areas with concentrated equipment, and transitional areas.

[0053] Adaptive boundary optimization was performed on multiple sub-regions to ensure the scientific and reasonable division results. The initial sub-region boundaries were not precise enough and needed optimization and adjustment. A temperature and humidity sensor network was deployed, adding sampling points at the boundaries of each sub-region to collect temperature and humidity airflow data in real time. The collected data was compared with historical data to calculate the fluctuation of the temperature and humidity airflow field in the boundary region. When the temperature difference between the two sides of the boundary exceeded 2°C or the humidity difference exceeded 8%RH, the boundary division was considered unreasonable and needed adjustment. The boundary adjustment adopted an iterative optimization method, adjusting the boundary position by 0.5 meters each time, until the temperature and humidity difference between adjacent regions decreased to below the threshold. In a practical case, a data center was initially divided into 4 sub-regions. After 3 rounds of boundary optimization, it was finally determined to have 6 sub-regions, with the boundary position shifting by an average of 1.2 meters from the initial division.

[0054] When the difference in temperature and humidity airflow field data between adjacent sub-regions exceeds a preset error threshold, dynamically adjusting the sub-region boundaries reflects the system's adaptability. The temperature error threshold is set at 1.8℃, and the humidity error threshold at 7%RH. During system operation, boundary checks are performed every 10 minutes. When the temperature and humidity difference between the regions on either side of a boundary exceeds the threshold three times consecutively, the dynamic adjustment mechanism is triggered. Boundary adjustment uses a gradient descent method, moving the boundary along the temperature and humidity gradient direction, with each movement being 0.3 meters, allowing a maximum of 5 movements. If the error requirement cannot be met after 5 adjustments, a sub-region is further divided into two regions. In a factory workshop monitoring case, due to adjustments in the production equipment layout, the original 8 sub-regions were dynamically adjusted to 11 sub-regions after one week of system operation, and the average monitoring error decreased from the initial 2.3℃ to 0.9℃.

[0055] Based on the optimized sub-region division, establishing a data correlation model between sub-regions is the core of predicting the overall temperature and humidity airflow field. Sub-regions are physically interconnected, and their temperature and humidity airflow field changes are not independent. By analyzing historical data, the propagation patterns of temperature and humidity changes between sub-regions are identified. For each pair of adjacent sub-regions, a time-series correlation model is established to describe how temperature and humidity changes in one region affect adjacent regions. The model considers multiple factors, including distance between regions, air circulation conditions, and heat source distribution. A weighted average method is used to calculate the influence coefficient. For example, in two adjacent sub-regions, when the temperature in the upstream region rises by 1°C, the average temperature in the downstream region rises by 0.7°C after 15 minutes. After establishing a complete sub-region correlation network, the temperature and humidity distribution of the entire building can be extrapolated from a small amount of sensor data. In a laboratory environment, using 25 sensors covering 12 sub-regions, the correlation model can extrapolate the temperature and humidity status of more than 200 grid cells, with average errors controlled within 0.5°C and 3%RH.

[0056] By implementing the above technologies, efficient and accurate monitoring of the temperature and humidity airflow field inside building cabins can be achieved, significantly reducing sensor deployment costs, improving the coverage and accuracy of the monitoring system, and providing strong support for building environment optimization and energy conservation.

[0057] In one optional implementation, calculating the disturbance response prediction error for each sub-region based on real-time data collected by the multi-level sensor network, and adaptively adjusting the sampling frequency of the multi-level sensor network based on the disturbance response prediction error includes:

[0058] The disturbance response prediction error of each sub-region is calculated based on the real-time data collected by the multi-level sensor network. Reference measurement points are set up in each sub-region of the multi-level sensor network. The time difference of temperature and humidity change between adjacent measurement points is calculated based on the reference measurement points. The disturbance propagation rate is calculated based on the time difference of temperature and humidity change and the spatial distance between adjacent measurement points.

[0059] The sub-region is divided into a fast response region and a slow response region according to the disturbance propagation rate. The disturbance propagation rate of the fast response region is greater than a preset rate threshold, and the disturbance propagation rate of the slow response region is less than or equal to the preset rate threshold.

[0060] A response link graph is constructed, which includes a set of sub-region nodes and a set of disturbance propagation paths. The importance of a node is calculated based on the edge set propagation rate and the path length of the sub-region node. The sampling frequency of the multi-level sensor network is adaptively adjusted according to the disturbance response prediction error and the node importance.

[0061] Reference measurement points are deployed in each sub-region of the multi-level sensor network. These reference points are typically selected at key locations within the sub-regions, such as at region boundaries, near heat or humidity sources, and in densely populated areas. In practical applications, for a 1000-square-meter data center server room, one reference measurement point can be deployed per 100-square-meter area, for a total of 10 reference measurement points. The initial sampling frequency of the reference measurement points is set to once every 10 seconds to ensure that subtle fluctuations in environmental changes can be captured.

[0062] When the external environment changes, such as when the cooling system starts up or new equipment joins the network, the system records the time of temperature and humidity changes using reference measuring points. For example, if the cooling system starts up at point A, and the system records that the temperature at point A begins to drop at 13:45:30, while the temperature at the adjacent point B begins to drop at 13:45:42, and the distance between the two points is 12 meters, then the disturbance propagation rate is calculated to be 1 meter per second. The system calculates the disturbance propagation rate for all adjacent measuring point pairs in the network, forming a complete disturbance propagation rate matrix.

[0063] Based on the calculated disturbance propagation rate, the sub-region is divided into a fast response zone and a slow response zone. Assuming a preset rate threshold of 0.8 m / s, areas with a disturbance propagation rate greater than 0.8 m / s are marked as fast response zones, while areas with a disturbance propagation rate less than or equal to 0.8 m / s are marked as slow response zones. In actual deployments, cold aisle areas in data centers are typically identified as fast response zones, while densely populated equipment areas are often identified as slow response zones due to airflow obstruction.

[0064] A response link graph is constructed, consisting of a set of sub-region nodes and a set of disturbance propagation paths. Each sub-region corresponds to a node in the graph, and the connections between nodes represent disturbance propagation paths. The system assigns a weight to each edge, the weight being equal to the reciprocal of the disturbance propagation rate on the corresponding path, reflecting the time required for the disturbance to travel along that path. For example, if the disturbance propagation rate from point A to point B is 1 meter per second, and the path length is 12 meters, then the weight of that edge is 12 seconds.

[0065] Based on the constructed response link graph, the importance of each node is calculated. Node importance considers the shortest path length from that node to all other nodes and the number of shortest paths passing through that node. Specifically, the system first uses Dijkstra's algorithm to calculate the shortest path between any two nodes in the network, then counts the number of shortest paths passing through each node, and calculates the betweenness centrality of that node. The more shortest paths a node has, the higher its importance. For example, node C, located in the center of the data center, has a calculated importance value of 0.85, while node D, located on the edge, has an importance value of 0.32.

[0066] The disturbance response prediction error for each sub-region is calculated. A prediction model is trained based on historical data, taking into account spatial correlation and temporal continuity, to predict future temperature and humidity changes. The predicted values ​​are compared with the actual collected data to calculate the prediction error. For example, in a certain prediction, the system predicted a temperature change of 1.2℃ for the fast response zone E, while the actual measured value was 1.5℃. The prediction error is then 0.3℃, with a relative error of 20%.

[0067] The sampling frequency of a multi-level sensor network is adaptively adjusted based on the calculated node importance and prediction error. The adjustment strategy follows these rules: for nodes with high importance and large prediction errors, the sampling frequency is increased; for nodes with low importance and small prediction errors, the sampling frequency is decreased. The specific adjustment formula calculates the new sampling frequency based on the node importance, prediction error, and the current sampling frequency.

[0068] For example, for the fast response region E with a node importance of 0.85 and a prediction error of 20%, the initial sampling frequency was once every 10 seconds, which was adjusted to once every 5 seconds after calculation. For the slow response region F with a node importance of 0.32 and a prediction error of 5%, the initial sampling frequency was once every 10 seconds, which was adjusted to once every 30 seconds. This adjustment ensures that system resources are concentrated on monitoring critical and abnormally changing areas, while reducing the monitoring frequency of stable areas, thus optimizing the efficiency of system resource utilization.

[0069] The above steps enable adaptive adjustment of the sampling frequency of a multi-level sensor network based on disturbance response prediction error, improving overall monitoring efficiency and accuracy. In practical deployments, this method reduces energy consumption of the sensor network by 35% while improving monitoring accuracy in key areas by 25%, effectively balancing monitoring accuracy and resource consumption. The system can also periodically reassess node importance and prediction error based on environmental changes, continuously optimizing the sampling strategy to adapt to dynamically changing environmental conditions.

[0070] In one optional implementation, calculating the time difference of temperature and humidity changes between adjacent measuring points based on the reference measuring point, and calculating the disturbance propagation rate based on the time difference of temperature and humidity changes and the spatial distance between the adjacent measuring points includes:

[0071] Collect temperature and humidity time series data of reference measuring points, and pair the reference measuring points to form adjacent measuring point groups; calculate the mean of temperature and humidity time series data of two reference measuring points in the adjacent measuring point group, calculate the cross-correlation function between the two reference measuring points based on the difference between the temperature and humidity time series data and the mean, and obtain the time offset corresponding to the maximum value of the cross-correlation function as the temperature and humidity change time difference of the adjacent measuring point group.

[0072] Calculate the spatial distance between two reference measuring points in the adjacent measuring point group, and divide the spatial distance of the adjacent measuring point group by the time difference of temperature and humidity changes in the adjacent measuring point group to obtain the disturbance propagation rate of the adjacent measuring point group.

[0073] like Figure 2 As shown, the method includes:

[0074] In practice, reference measurement points can be multiple temperature and humidity sensors deployed in a specific environment. These sensors are connected to the data acquisition system via wired or wireless networks. The temperature and humidity data recorded by each reference measurement point includes a timestamp and the corresponding temperature and humidity values. For example, 10 reference measurement points can be evenly distributed in a 100-square-meter space, with each measurement point sampling once per minute for 24 hours, resulting in 1440 temperature and humidity data points for each measurement point.

[0075] After data collection, these reference measuring points are paired to form adjacent measuring point groups. For 10 reference measuring points, 45 adjacent measuring point groups can be constructed. For each adjacent measuring point group, the system reads the temperature and humidity time series data of two reference measuring points. Assume that the temperature data collected by the first reference measuring point within 24 hours is T1=[20.1, 20.3, 20.5,..., 21.2]℃, and the humidity data is H1=[45.2, 45.5, 45.8, ..., 46.1]%; the temperature data collected by the second reference measuring point is T2=[20.2, 20.4, 20.6, ..., 21.3]℃, and the humidity data is H2=[45.3, 45.6, 45.9,..., 46.2]%.

[0076] For each pair of adjacent measuring points, the mean of the temperature and humidity time series data of the two reference measuring points is calculated. Specifically, for temperature data, the mean sequence is AvgT = [(T1[0]+T2[0]) / 2, (T1[1]+T2[1]) / 2, ..., (T1

[1439] +T2

[1439] ) / 2]; similarly, the mean sequence of humidity data is AvgH.

[0077] Based on the difference between the temperature and humidity time series data and the mean, the cross-correlation function between two reference measuring points is calculated. The specific steps are as follows: First, calculate the temperature difference sequence DiffT1 = [T1[0]-AvgT[0], T1[1]-AvgT[1], ..., T1

[1439] -AvgT

[1439] ] and the humidity difference sequence DiffH1 for the first reference measuring point; calculate the temperature difference sequence DiffT2 and the humidity difference sequence DiffH2 for the second reference measuring point in the same way.

[0078] The system calculates the cross-correlation function CorrT between temperature difference sequences DiffT1 and DiffT2, and the cross-correlation function CorrH between humidity difference sequences DiffH1 and DiffH2. In calculating the cross-correlation function, the system compares the two sequences at different time offsets to find the time offset that makes the two sequences most similar. For example, the temperature cross-correlation function CorrT reaches its maximum value of 0.87 at a time offset of 3 minutes, indicating that the temperature change at the second reference point lags behind the first point by 3 minutes; similarly, the humidity cross-correlation function CorrH reaches its maximum value of 0.92 at a time offset of 2.5 minutes.

[0079] Choose either the temperature cross-correlation function or the humidity cross-correlation function, or a combination of both, depending on the application scenario. In this example, assume that the time offset corresponding to the maximum value of the temperature cross-correlation function is selected as the time difference of temperature and humidity changes for the adjacent measurement point group, i.e., 3 minutes.

[0080] Calculate the spatial distance between two reference measuring points in an adjacent measuring point group. Assuming the coordinates of the first reference measuring point are (3.5m, 2.1m) and the coordinates of the second reference measuring point are (6.2m, 4.7m), the spatial distance between them is 3.93 meters (obtained by calculating the Euclidean distance between the two points).

[0081] Dividing the spatial distance between adjacent measuring point groups by the time difference of temperature and humidity changes yields the disturbance propagation rate of that adjacent measuring point group. In this example, the disturbance propagation rate is 3.93 meters ÷ 3 minutes = 1.31 meters / minute.

[0082] For all 45 adjacent measurement point groups, the system calculated the disturbance propagation rate for each group using the method described above. This rate data can be used to analyze the propagation characteristics of temperature and humidity disturbances in the environment, such as determining the presence of abnormal temperature and humidity areas, identifying heat source locations, or assessing airflow patterns.

[0083] In practical applications, the system can perform statistical analysis on the calculated disturbance propagation rate, such as calculating the average rate, rate distribution, or trend over time. If the disturbance propagation rate in a certain area is significantly higher or lower than that in other areas, it indicates that there is an air circulation problem or heat source in that area. The system can also combine disturbance propagation direction information to generate a temperature and humidity disturbance propagation vector field in the environment, visually displaying the flow of disturbances.

[0084] The disturbance propagation rate calculated by this method can provide data support for HVAC system adjustments, indoor environmental quality optimization, and energy efficiency improvement, and has broad application value.

[0085] In one optional implementation, the data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions, including:

[0086] A temporal attention model is constructed based on the calibrated temperature and humidity airflow field data. The temporal attention model calculates the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin by weighted superposition of the importance weight of each measuring point and the calibration data.

[0087] Based on the calibrated temperature and humidity airflow field data, the sensor measurement points in the multi-level sensor network are constructed as a measurement point association structure, the temperature and humidity airflow field data difference between adjacent sensor measurement points is calculated, and the temperature and humidity airflow field data difference is converted into the connection strength between measurement points through the time-series attention model.

[0088] Spatial features are extracted from the node features in the measurement point association structure to obtain the spatial dimension features of the nodes; the spatial dimension features are introduced into the temporal attention model, and the importance of each temporal feature is calculated based on the correlation of the feature vectors. The historical temporal features of the nodes are weighted and combined according to the importance to obtain the spatiotemporal dimension features of the nodes.

[0089] The propagation probability of disturbance between adjacent nodes is calculated based on the spatiotemporal dimension characteristics. Sensor measurement point pairs with a propagation probability greater than a preset probability threshold are selected as effective paths for disturbance propagation. The effective paths are connected into a propagation network based on the temporal attention model to track and record the propagation path of disturbance between sub-regions inside the building cabin.

[0090] When constructing a temporal attention model based on calibrated temperature and humidity airflow field data, a weighted combination of weights and data is used to calculate the spatiotemporal distribution of the temperature and humidity airflow field inside the building cabin. Specifically, by analyzing the historical time-series data of each measuring point in a multi-level sensor network, the importance weight of each measuring point at different times is calculated. For example, for a network with 100 sensor measuring points, analyzing the temperature data of the past 24 hours reveals that the weight value of sensor measuring points near doors and windows reaches 0.15, while the weight of measuring points in the internal area is 0.08. These weights are then weighted and superimposed with the calibrated temperature and humidity airflow field data of each measuring point to obtain the spatiotemporal distribution data of the temperature and humidity airflow field inside the entire building cabin. This method is particularly suitable for capturing the airflow field distribution characteristics in dynamically changing environments.

[0091] When constructing the measurement point association structure, the difference in temperature and humidity airflow field data between adjacent sensor measurement points is calculated. For example, for two adjacent sensors A and B, their temperature difference sequence [0.5, 0.8, 1.2, 0.7, 0.3] and humidity difference sequence [2.1, 2.5, 3.0, 2.8, 2.0] over a 10-minute period are calculated, and the overall difference is calculated to be 5.68 using the Euclidean distance of the feature vectors. This difference value is then converted into the connection strength between measurement points using a temporal attention model. When the difference is below a preset threshold of 4.0, a strong connection is considered to exist between the two points, and a connection strength of 0.85 is assigned; when the difference is between 4.0 and 8.0, a medium connection strength of 0.45 is assigned; and when the difference is above 8.0, a weak connection strength of 0.15 is assigned. This connection strength characterizes the possibility of disturbances propagating between measurement points.

[0092] When extracting spatial features from the node features in the sensor point association structure, graph convolutional network technology is used to process the sensor point association structure. For each sensor point, its geographical coordinates, surrounding building structure features, and airflow channel information are fused to extract the spatial dimension features of the node. For example, if a sensor point is located in a corner of a room and is 3 meters away from the nearest window, the extracted spatial features include a location feature vector [0.2, 0.8, 0.3] and a structural feature vector [0.6, 0.1, 0.9]. These spatial features can characterize the relative position of the node in the building space and the degree to which it is affected by the surrounding environment.

[0093] When introducing the temporal attention model, the historical temporal data of each node is analyzed to calculate the importance of data at different time points. For example, for the temperature data of a certain measuring point over the past 30 minutes [22.5, 22.8, 23.2, 24.1, 24.5, 24.3], based on the data change trend, the importance weight of each time point is calculated through feature vector correlation analysis as [0.05, 0.08, 0.15, 0.25, 0.30, 0.17]. This indicates that the temperature change from 24.1 to 24.5 has the greatest impact on the current state. Based on these weights, the historical temporal features of the node are weighted and combined to obtain the spatiotemporal dimension features that comprehensively characterize the spatiotemporal properties of the node.

[0094] When calculating the propagation probability of disturbances between adjacent nodes based on spatiotemporal characteristics, the connection strength, spatial distance, and historical propagation patterns between nodes are comprehensively considered. For example, for two adjacent sensor measurement points C and D, with a connection strength of 0.78, a spatial distance of 2.5 meters, and a historical propagation consistency of 0.65, the calculated probability of a disturbance propagating from C to D is 0.72. A propagation probability threshold of 0.60 is set, and measurement point pairs with propagation probabilities greater than this threshold are selected as valid paths for disturbance propagation. Connecting all valid paths into a propagation network allows for tracking and recording the propagation paths of disturbances between different sub-regions within the building module.

[0095] In practical applications, such as the temperature regulation scenario of an office building's air conditioning system, the propagation network analyzed using the above method shows that temperature disturbances originating in the first-floor lobby will propagate upwards along the elevator shaft, affecting the second-floor corridor after 5 minutes and the third-floor office area after 8 minutes. This propagation path analysis can be used to optimize air conditioning control strategies, adjust equipment parameters in affected areas in advance, and achieve precise regulation.

[0096] When dealing with a sudden indoor pollutant diffusion scenario, if a sudden increase in formaldehyde concentration is detected in a certain area, the system can predict, through the analysis of the constructed propagation network, that the pollutant will spread to three adjacent office areas along a specific path within 7 minutes. Based on this, the system can automatically adjust the fresh air volume in the relevant areas to block the pollutant diffusion path in advance.

[0097] Through the above technologies, a complete spatiotemporal analysis and disturbance propagation tracking system for the temperature and humidity airflow field inside a building cabin based on a multi-level sensor network has been established, enabling precise perception and proactive control of indoor environmental quality.

[0098] In one optional implementation, based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the disturbance response characteristics of the building cabin's temperature flow field is generated, including:

[0099] Based on spatiotemporal distribution data, the real-time field values ​​of each sub-region along the propagation path are extracted. The deviation between the real-time field values ​​and the preset reference field values ​​is calculated to obtain the sub-region disturbance response intensity. Based on the sub-region disturbance response intensity, a sub-region response coupling matrix is ​​constructed. The sum of the squares of the differences between the sub-region disturbance response intensity and the predicted values ​​of the sub-region response coupling matrix is ​​used as the judgment criterion to optimize the correlation coefficient of the sub-region response coupling matrix and obtain the disturbance attenuation parameters of each sub-region along the propagation path.

[0100] The correlation of disturbance response intensities in adjacent sub-regions is calculated along the propagation path, and the sub-region response coupling matrix is ​​updated. The attenuation law of disturbances on the propagation path is analyzed based on the sub-region response coupling matrix. The response delay, attenuation rate, and residual intensity of each sub-region on the propagation path are calculated. An evaluation report on the disturbance response characteristics of the building cabin temperature flow field is generated based on the response delay, attenuation rate, residual intensity, and sub-region response coupling matrix.

[0101] The system acquires spatiotemporal distribution data and propagation path information within the building's cabin. The spatiotemporal distribution data includes temperature and airflow field data at different times and spatial locations, which can be collected through a distributed sensor network. The propagation path is predetermined based on the building's structural characteristics and airflow patterns, representing the direction of disturbance propagation within the building's cabin.

[0102] The acquired spatiotemporal distribution data is propagated along a preset path, and real-time field values ​​for each sub-region are extracted. These real-time field values ​​include key parameters such as temperature, airflow velocity, and airflow direction. For example, if sub-region A has a measured temperature of 24.5℃ at time t, and the preset reference temperature is 23.0℃, then the temperature disturbance response intensity for that sub-region is 1.5℃. Similarly, the disturbance response intensity for parameters such as airflow velocity and airflow direction can be calculated.

[0103] Based on the calculated disturbance response intensity of each sub-region, a sub-region response coupling matrix is ​​constructed. Taking the temperature field as an example, if the building cabin is divided into 5 sub-regions, the response coupling matrix is ​​a 5×5 square matrix, and the matrix elements represent the disturbance transmission relationship between different sub-regions. Initially, initial values ​​can be set according to the physical model. For example, the coupling coefficient of adjacent sub-regions can be set to 0.8, the coupling coefficient of one sub-region apart can be set to 0.5, and the coupling coefficients of other more distant sub-regions can be set to 0.3 and 0.1.

[0104] To optimize the correlation coefficients of the sub-region response coupling matrix, iterative calculations are performed based on measured data. The sum of squares of the differences between the sub-region disturbance response intensity and the predicted value of the response coupling matrix is ​​calculated and used as the optimization objective function. Gradient descent is employed to optimize the coupling coefficients, iterating repeatedly until the objective function converges to below a preset threshold, such as 0.01. This process yields a more accurate understanding of the disturbance propagation relationship between sub-regions and simultaneously calculates the disturbance attenuation parameters along the propagation path for each sub-region. For example, for temperature disturbances, the attenuation parameter from sub-region A to sub-region B is 0.75, indicating that 75% of the temperature disturbance will propagate from A to B.

[0105] The correlation of disturbance response intensities between adjacent sub-regions is calculated along the propagation path using a sliding window method. A 10-minute time window is used to calculate the Pearson correlation coefficient of temperature changes in adjacent sub-regions within that time window. For example, a correlation coefficient of 0.87 between the temperature changes of sub-regions A and B indicates a high correlation between the temperature changes in the two regions. Based on the calculated correlation data, the sub-region response coupling matrix is ​​updated to more accurately reflect the disturbance propagation characteristics.

[0106] Based on the updated response coupling matrix analysis, the attenuation law of the disturbance along the propagation path was found to be as follows: for temperature disturbances, the attenuation follows an exponential decay pattern, and the attenuation coefficient is related to factors such as the distance between sub-regions and the barrier structure. For example, in an open space, the attenuation coefficient is 0.15 / m, while in a space with partitions, the attenuation coefficient increases to 0.30 / m.

[0107] Further calculations are performed on key disturbance characteristics along the propagation path for each sub-region, including response delay, decay rate, and residual strength. Response delay refers to the time difference between the generation of a disturbance in the source region and its detection in the target region; for example, the response delay for a temperature disturbance from sub-region A to sub-region C is 78 seconds. The decay rate represents the rate at which the disturbance strength decreases over time; for example, the decay rate for a temperature disturbance is 0.05℃ / minute. Residual strength represents the proportion of the disturbance's remaining strength at the endpoint relative to its initial strength; for example, a residual strength of 0.35 for a temperature disturbance propagating from sub-region A to sub-region E indicates that 35% of the disturbance strength is retained at the endpoint.

[0108] Based on the above analysis results, the system generates an evaluation report on the disturbance response characteristics of the building cabin's temperature and flow field. The evaluation report includes the following: a data table of disturbance response intensity for each sub-region, showing the disturbance intensity values ​​of each sub-region at different time points; a visualization diagram of the sub-region response coupling matrix, clearly showing the disturbance transmission relationship between each sub-region; a disturbance propagation path analysis diagram, marking the response delay, decay rate, and residual intensity of key nodes; a disturbance response risk assessment, identifying sub-regions that are particularly sensitive to disturbances or propagation path segments with abnormal disturbance decay; and optimization suggestions, proposing specific solutions for the building cabin's temperature and flow field control system to address the identified problems.

[0109] In practical applications, this method can be used to analyze the impact of air conditioning system startup and shutdown on the temperature of different areas of a building. For example, when the central air conditioning system starts up, if the temperature in the source area drops from 26°C to 23°C, the above method can be used to analyze how this temperature disturbance propagates to other areas of the building and the response characteristics of each area. The results show that the adjacent office area takes 120 seconds to sense the temperature change, and the temperature eventually drops by 2.1°C, while the distant conference room takes 310 seconds to sense the temperature change, and the temperature eventually drops by only 0.9°C. These data provide important basis for optimizing air conditioning system operation strategies.

[0110] A second aspect of the present invention provides a dynamic simulation test system for the temperature and flow field disturbance response characteristics of a building cabin, comprising:

[0111] The first unit is used to acquire the geometric parameters of the building cabin, divide the building cabin into grid units based on the geometric parameters, deploy a multi-level sensor network at the nodes of the grid units, and collect temperature and humidity airflow field data inside the building cabin; divide the grid units into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions.

[0112] The second unit is used to calculate the disturbance response prediction error of each sub-region based on the real-time data collected by the multi-level sensor network, adaptively adjust the sampling frequency of the multi-level sensor network based on the disturbance response prediction error, and trigger the recalibration of the multi-level sensor network when the disturbance response prediction error is greater than a preset error threshold.

[0113] The third unit is used to process and analyze the data collected by the calibrated multi-level sensor network, calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and track and record the propagation path of disturbances between sub-regions; based on the spatiotemporal distribution data and the propagation path, calculate the disturbance attenuation characteristics of each sub-region, and generate an evaluation report on the temperature and humidity airflow field disturbance response characteristics of the building cabin.

[0114] A third aspect of the present invention provides an electronic device, comprising:

[0115] processor;

[0116] Memory used to store processor-executable instructions;

[0117] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0118] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0119] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0120] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dynamic simulation test method for the temperature and flow field disturbance response characteristics of a building cabin, characterized in that, include: Obtain the geometric parameters of the building cabin, divide it into grid cells according to the geometric parameters, deploy a multi-level sensor network at the nodes of the grid cells, and collect temperature and humidity airflow field data inside the building cabin; divide the grid cells into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions; Based on the real-time data collected by the multi-level sensor network, the disturbance response prediction error of each sub-region is calculated, and the sampling frequency of the multi-level sensor network is adaptively adjusted based on the disturbance response prediction error, including: The disturbance response prediction error of each sub-region is calculated based on the real-time data collected by the multi-level sensor network. Reference measurement points are set up in each sub-region of the multi-level sensor network. The time difference of temperature and humidity change between adjacent measurement points is calculated based on the reference measurement points. The disturbance propagation rate is calculated based on the time difference of temperature and humidity change and the spatial distance between adjacent measurement points. The sub-region is divided into a fast response region and a slow response region according to the disturbance propagation rate. The disturbance propagation rate of the fast response region is greater than a preset rate threshold, and the disturbance propagation rate of the slow response region is less than or equal to the preset rate threshold. A response link graph is constructed, which includes a set of sub-region nodes and a set of disturbance propagation paths. The importance of a node is calculated based on the edge set propagation rate and the path length of the sub-region node. The sampling frequency of the multi-level sensor network is adaptively adjusted according to the disturbance response prediction error and the node importance. When the disturbance response prediction error is greater than a preset error threshold, the recalibration of the multi-level sensor network is triggered. The data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions. Based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the temperature and flow field disturbance response characteristics of the building cabin is generated.

2. The method according to claim 1, characterized in that, Dividing the grid cells into multiple sub-regions and establishing a data association model between the sub-regions using the temperature and humidity airflow field data includes: The grid cells are divided according to the geometric parameters of the building cabin. The grid cells are then divided into multiple sub-regions based on the gradient variation characteristics of the temperature and humidity airflow field. Adaptive boundary optimization is performed on the multiple sub-regions. When the difference in temperature and humidity airflow field data between adjacent sub-regions is greater than a preset error threshold, the division boundary of the sub-regions is dynamically adjusted. Based on the optimized sub-region division results, a data association model between the sub-regions is established.

3. The method according to claim 1, characterized in that, Calculating the time difference of temperature and humidity changes between adjacent measuring points based on the reference measuring point, and calculating the disturbance propagation rate based on the time difference of temperature and humidity changes and the spatial distance between the adjacent measuring points, includes: Collect temperature and humidity time series data of reference measuring points, and pair the reference measuring points to form adjacent measuring point groups; calculate the mean of temperature and humidity time series data of two reference measuring points in the adjacent measuring point group, calculate the cross-correlation function between the two reference measuring points based on the difference between the temperature and humidity time series data and the mean, and obtain the time offset corresponding to the maximum value of the cross-correlation function as the temperature and humidity change time difference of the adjacent measuring point group. Calculate the spatial distance between two reference measuring points in the adjacent measuring point group, and divide the spatial distance of the adjacent measuring point group by the time difference of temperature and humidity changes in the adjacent measuring point group to obtain the disturbance propagation rate of the adjacent measuring point group.

4. The method according to claim 1, characterized in that, The data collected by the calibrated multi-level sensor network is processed and analyzed to calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and to track and record the propagation path of disturbances between sub-regions, including: A temporal attention model is constructed based on the calibrated temperature and humidity airflow field data. The temporal attention model calculates the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin by weighted superposition of the importance weight of each measuring point and the calibration data. Based on the calibrated temperature and humidity airflow field data, the sensor measurement points in the multi-level sensor network are constructed as a measurement point association structure, the temperature and humidity airflow field data difference between adjacent sensor measurement points is calculated, and the temperature and humidity airflow field data difference is converted into the connection strength between measurement points through the time-series attention model. Spatial features are extracted from the node features in the measurement point association structure to obtain the spatial dimension features of the nodes; the spatial dimension features are introduced into the temporal attention model, and the importance of each temporal feature is calculated based on the correlation of the feature vectors. The historical temporal features of the nodes are weighted and combined according to the importance to obtain the spatiotemporal dimension features of the nodes. The propagation probability of disturbance between adjacent nodes is calculated based on the spatiotemporal dimension characteristics. Sensor measurement point pairs with a propagation probability greater than a preset probability threshold are selected as effective paths for disturbance propagation. The effective paths are connected into a propagation network based on the temporal attention model to track and record the propagation path of disturbance between sub-regions inside the building cabin.

5. The method according to claim 1, characterized in that, Based on the spatiotemporal distribution data and the propagation path, the disturbance attenuation characteristics of each sub-region are calculated, and an evaluation report on the disturbance response characteristics of the building cabin's temperature and flow field is generated, including: Based on spatiotemporal distribution data, the real-time field values ​​of each sub-region along the propagation path are extracted. The deviation between the real-time field values ​​and the preset reference field values ​​is calculated to obtain the sub-region disturbance response intensity. Based on the sub-region disturbance response intensity, a sub-region response coupling matrix is ​​constructed. The sum of the squares of the differences between the sub-region disturbance response intensity and the predicted values ​​of the sub-region response coupling matrix is ​​used as the judgment criterion to optimize the correlation coefficient of the sub-region response coupling matrix and obtain the disturbance attenuation parameters of each sub-region along the propagation path. The correlation of disturbance response intensities in adjacent sub-regions is calculated along the propagation path, and the sub-region response coupling matrix is ​​updated. The attenuation law of disturbances on the propagation path is analyzed based on the sub-region response coupling matrix. The response delay, attenuation rate, and residual intensity of each sub-region on the propagation path are calculated. An evaluation report on the disturbance response characteristics of the building cabin temperature flow field is generated based on the response delay, attenuation rate, residual intensity, and sub-region response coupling matrix.

6. A dynamic simulation testing system for the temperature and flow field disturbance response characteristics of a building cabin, used to implement the method described in any one of claims 1-5, characterized in that, include: The first unit is used to acquire the geometric parameters of the building cabin, divide the building cabin into grid units based on the geometric parameters, deploy a multi-level sensor network at the nodes of the grid units, and collect temperature and humidity airflow field data inside the building cabin; divide the grid units into multiple sub-regions, and use the temperature and humidity airflow field data to establish a data association model between the sub-regions. The second unit is used to calculate the disturbance response prediction error of each sub-region based on the real-time data collected by the multi-level sensor network, adaptively adjust the sampling frequency of the multi-level sensor network based on the disturbance response prediction error, and trigger the recalibration of the multi-level sensor network when the disturbance response prediction error is greater than a preset error threshold. The third unit is used to process and analyze the data collected by the calibrated multi-level sensor network, calculate the spatiotemporal distribution data of the temperature and humidity airflow field inside the building cabin, and track and record the propagation path of disturbances between sub-regions; based on the spatiotemporal distribution data and the propagation path, calculate the disturbance attenuation characteristics of each sub-region, and generate an evaluation report on the temperature and humidity airflow field disturbance response characteristics of the building cabin.

7. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 5.

Images