Building ventilation thermal load regulation method and system based on thermal plume prediction

Abstract

The present application relates to the technical field of building thermal environment regulation, and discloses a building ventilation thermal load regulation method and system based on thermal plume prediction, which comprises the following steps: constructing a curtain wall surface three-dimensional coordinate system and a thermal state temperature atlas, and obtaining a current thermal plume trajectory line; combining a historical thermal plume trajectory database to perform similar trajectory matching and predict a thermal plume development path; based on the prediction result, establishing a floor load grading sequence and a neighboring layer feedforward temperature compensation relationship, and implementing time sequence misplacement dimming through spiral progressive control; and after time sequence misplacement dimming, performing thermal-light coupling feedback correction; the present application can effectively inhibit thermal air backflow, reduce high-rise room temperature and air conditioning load, protect indoor natural lighting, optimize building thermal environment, and realize efficient regulation of building ventilation thermal load.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of building thermal environment regulation, and more particularly, to a building ventilation thermal load regulation method and system based on thermal plume prediction. BACKGROUND

[0002] Glass curtain walls, with their transparency and light weight, have become the preferred solution for modern high-rise building envelope systems, and their facade form directly affects building ventilation paths and thermal environment balance. However, high-intensity solar radiation in summer directly acts on the outer glass surface of the curtain wall, inducing a transient temperature rise of over 60℃. Driven by buoyancy, these high-temperature areas form a "thermal plume" that rises rapidly along the curtain wall facade. This thermal plume is coupled with the thermal pressure channel formed by the open windows at the top of the building, changing the airflow direction of natural ventilation and creating a unique "curtain wall heat island microclimate" around the curtain wall. If the building ventilation thermal load regulation system still follows the traditional single-point light-temperature feedback approach, only the local temperature and light parameters are adjusted passively, and the thermal plume's rising path along the curtain wall and its temporal and spatial intensity cannot be predicted and intervened, leading to the continuous intrusion of high-temperature air carried by the thermal plume into the room, disrupting the thermal balance of the building ventilation system, and ultimately resulting in high air conditioning loads in high-rise buildings and difficulty in reducing the overall building ventilation thermal load energy consumption baseline.

[0003] Chinese patent application CN119041605A proposes an intelligent light-adjusting thermal insulation curtain wall and a light-adjusting method. The thermal insulation curtain wall is arranged with rotatable reflective prisms between the double-layer glass, which adjusts the visible light flux by controlling the angle through a program, and relies on the hollow layer to reduce the heat conduction between the outer and inner sheets. The control variable is only the preset illuminance-time curve, and real-time sensing and prediction of the external surface thermal plume trajectory are not involved. Chinese patent CN115478645B discloses a building curtain wall and a lighting adjustment method. The building curtain wall changes the light transmission area by rotating the entire separation layer to achieve uniform lighting, but still uses a static optical target as input, lacking an active inhibition mechanism for the thermal pressure transmission between floors induced by the thermal plume. Both of them stay at the "light-heat isolation" level, lacking a "trajectory prediction-load grading-neighbor layer feedforward" closed loop.

[0004] The existing blind area of the prior art is that when the surface of the outer glass is heated and forms an upward thermal plume due to solar radiation, if the upper window is in an open ventilation state, the thermal plume will directly "backflow" into the indoor space under the driving of the pressure difference, forming a "heat source-opening" transient coupling. At this time, even if the prism or the interlayer has reflected visible light to the comfort zone, due to the lack of real-time calculation of the thermal plume rising path and the thermal pressure gradient between the floors in the control variables, the system cannot reduce the light transmittance of the upper curtain wall in advance to weaken the intensity of the outer wall heat source, nor can it compensate for the adjacent layer feedforward for the high-temperature thermal plume that is about to rise in the lower layer, resulting in abnormal rise of the upper room temperature, long-time high-load operation of the air conditioner, and the energy consumption baseline is raised. Therefore, how to make the light adjustment ahead of the thermal plume and form a thermodynamic block between the floors has become a sub-allocation problem that needs to be broken through for the glass curtain wall heat island microclimate regulation. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the present application provides a building ventilation heat load regulation method and system based on thermal plume prediction, which captures the current thermal plume trajectory by constructing a curtain wall surface three-dimensional coordinate system, predicts the thermal plume development path in combination with the historical thermal plume trajectory database, and establishes a floor load classification sequence and an adjacent layer feedforward temperature compensation relationship based on the path prediction result, which can effectively inhibit the backflow of hot air and realize the reduction of high-rise room temperature and building comprehensive energy consumption. Through the spiral progressive control of the time sequence staggered light adjustment scheme and the thermal-optical coupling feedback correction, the accuracy and dynamic adaptability of the glass curtain wall "heat island microclimate" regulation are improved, while ensuring indoor natural lighting, the building thermal environment regulation effect is further optimized.

[0006] To achieve the above object, the present application provides the following technical scheme:

[0007] The building ventilation heat load regulation method based on thermal plume prediction comprises:

[0008] A curtain wall surface three-dimensional coordinate system is constructed, a thermal state temperature map is constructed in the curtain wall surface three-dimensional coordinate system, and a current thermal plume trajectory line is constructed based on the thermal state temperature map; a historical thermal plume trajectory database is constructed, a similar historical trajectory of the current thermal plume trajectory line is extracted from the historical thermal plume trajectory database, and a development path prediction result of the current thermal plume is predicted based on the similar historical trajectory;

[0009] A floor load classification sequence is established by using the development path prediction result of the current thermal plume; an adjacent layer feedforward temperature compensation relationship is constructed based on the floor load classification sequence, and time sequence staggered light adjustment is implemented through spiral progressive control based on the adjacent layer feedforward temperature compensation relationship; after the time sequence staggered light adjustment is executed, thermal-optical coupling feedback correction is performed.

[0010] The method for constructing the current thermal plume trajectory line based on the thermal state temperature map comprises:

[0011] calculating a temperature gradient field based on the thermal temperature map, identifying and marking a thermal plume source point;

[0012] connecting the thermal plume source points according to the space-time dual criteria, constructing a current thermal plume trajectory line, and associating a current time label with the current thermal plume trajectory line.

[0013] The method for constructing the thermal temperature map comprises:

[0014] marking grid nodes on the curtain wall surface, collecting the temperature of each grid node at a preset sampling period under sunlight conditions, and generating a thermal temperature map at each collection time.

[0015] The identification of the thermal plume source point comprises:

[0016] For the thermal temperature map at each collection time, calculating a vertical temperature gradient;

[0017] When the vertical temperature gradient is greater than or equal to a temperature gradient threshold α, marking the corresponding grid node as a thermal plume source point.

[0018] The method for implementing time sequence misplacement dimming based on the adjacent layer feedforward temperature compensation relationship comprises:

[0019] generating a floor light transmittance control instruction sequence based on the adjacent layer feedforward temperature compensation relationship;

[0020] numbering the grid nodes on the curtain wall surface in a spiral path, distributing the grid nodes to different dimming groups according to the node number, and forming a spiral progressive dimming node sequence;

[0021] determining the time sequence misplacement execution time of each dimming group according to the floor light transmittance control instruction sequence, and executing time sequence misplacement dimming according to the spiral progressive dimming node sequence and the time sequence misplacement execution time.

[0022] The method for establishing a floor load classification sequence comprises:

[0023] calculating the thermal intrusion risk index of each floor based on the development path prediction result of the current thermal plume;

[0024] classifying the thermal load of each floor according to the thermal intrusion risk index, recording the thermal load classification result in order of floor height, and generating a floor load classification sequence.

[0025] The development path prediction result of the current thermal plume at least comprises a future thermal plume trajectory line;

[0026] The method for calculating the thermal intrusion risk index of each floor comprises:

[0027] Based on the prediction result of the current thermal plume development path, the probability P(z') that the future thermal plume trajectory line passes through each floor is calculated, and the temperature peak coefficient C and the window opening rate W(z') of each floor are calculated, wherein z' is the floor identifier. temp

[0028] The thermal intrusion risk index R(z') of each floor is equal to the product of P(z'), C temp and W(z').

[0029] The method for classifying the thermal load of each floor comprises:

[0030] When R(z') < β1, the floor z' is classified as a low-load floor.

[0031] When β1≤R(z')≤β2, the floor z' is classified as a medium-load floor.

[0032] When R(z') > β2, the floor z' is classified as a high-load floor; wherein β1 is a first threshold value, and β2 is a second threshold value.

[0033] The historical thermal plume trajectory database comprises historical thermal plume trajectory lines of the past n days.

[0034] The method for extracting similar historical trajectories of the current thermal plume trajectory line comprises:

[0035] Extracting the trajectory features of the current thermal plume trajectory line and the historical trajectory features of the historical thermal plume trajectory lines.

[0036] Performing a space-temperature-time triple similarity determination on the trajectory features of the current thermal plume trajectory line and the historical trajectory features of the historical thermal plume trajectory lines, and taking the historical thermal plume trajectory lines that satisfy the space-temperature-time triple similarity determination as similar historical trajectories to obtain k similar historical trajectories, and adjusting the related parameters of the space-temperature-time triple similarity determination until k is greater than or equal to the sample quantity threshold value if k is less than the sample quantity threshold value.

[0037] The historical trajectory features of the historical thermal plume trajectory lines are associated with historical time labels.

[0038] The space-temperature-time triple similarity determination comprises a spatial position similarity determination, a temperature feature similarity determination, and a time similarity determination.

[0039] The condition for the time similarity determination is that the absolute value of the difference between the historical time label of the i'th historical thermal plume trajectory line and the current time label of the current thermal plume trajectory line is less than a time deviation threshold value.

[0040] ​The building ventilation thermal load regulation system based on thermal plume prediction is used for realizing the building ventilation thermal load regulation method based on thermal plume prediction, and comprises the following modules.

[0041] A thermal plume trajectory construction module is configured to construct a curtain wall surface three-dimensional coordinate system, construct a thermal state temperature atlas in the curtain wall surface three-dimensional coordinate system, and construct a current thermal plume trajectory line based on the thermal state temperature atlas.

[0042] A thermal plume trajectory prediction module is configured to construct a historical thermal plume trajectory database, extract a similar historical trajectory of the current thermal plume trajectory line from the historical thermal plume trajectory database, and predict a development path prediction result of the current thermal plume based on the similar historical trajectory.

[0043] A light regulation module is configured to establish a floor load hierarchical sequence by using the development path prediction result of the current thermal plume, construct a neighboring layer feedforward temperature compensation relationship based on the floor load hierarchical sequence, and implement time sequence staggered light regulation by using the spiral progressive control based on the neighboring layer feedforward temperature compensation relationship.

[0044] A feedback correction module is configured to perform thermal-light coupling feedback correction after the time sequence staggered light regulation is executed.

[0045] Compared with the prior art, the building ventilation thermal load regulation method based on thermal plume prediction has the following beneficial effects.

[0046] The closed-loop control of "trajectory prediction-hierarchical compensation-staggered execution-feedback correction" is adopted, so that the curtain wall light regulation is no longer only responsive to light, but the thermal plume rising channel is locked in advance, the heat is gradually reduced in stages before the heat invasion through the neighboring layer feedforward compensation, the optical and electrical impact caused by the synchronous action is eliminated by using the spiral progressive light regulation, and finally the high-rise room obtains lower surface temperature and less heat air backflow under the same sunlight condition, the air conditioning load is reduced, and the simultaneous optimization of light transmission, heat insulation and energy saving is realized. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0048] Figure 1 A principle flowchart of the building ventilation thermal load regulation method based on thermal plume prediction provided by the embodiments of the present application is shown in the following figure.

[0049] Figure 2 A method flowchart for establishing a floor load hierarchical sequence provided by the embodiments of the present application is shown in the following figure.

[0050] Figure 3 A method flowchart for generating a floor light transmittance control instruction sequence is provided for the embodiment of the present application.

[0051] Figure 4 A principle diagram of time sequence misplacement dimming is provided for the embodiment of the present application.

[0052] Figure 5 A method flowchart of thermal-optical coupling feedback correction is provided for the embodiment of the present application.

[0053] Figure 6 A functional module diagram of a building ventilation thermal load regulation system based on thermal plume prediction is provided for the embodiment of the present application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0055] Embodiment 1:

[0056] Please refer to Figure 1 The embodiment provides a building ventilation thermal load regulation method based on thermal plume prediction, which comprises the following steps:

[0057] In step S10, a curtain wall surface three-dimensional coordinate system is constructed, and a current thermal plume trajectory line is constructed in the curtain wall surface three-dimensional coordinate system; a historical thermal plume trajectory database is constructed, a similar historical trajectory of the current thermal plume trajectory line is extracted from the historical thermal plume trajectory database, a common development law of the similar historical trajectory is analyzed, and a development path prediction result of the current thermal plume is predicted.

[0058] Further, step S10 comprises the following steps:

[0059] In step S11, a curtain wall surface three-dimensional coordinate system is constructed with the center of the bottom of the curtain wall as the origin, and a grid node is marked on the curtain wall surface.

[0060] The origin of the curtain wall surface three-dimensional coordinate system is set at the center of the curtain wall bottom. The position is selected according to the starting rule of heat transfer of the curtain wall. The curtain wall bottom is usually the initial area of heat accumulation, and serves as the geometric reference point of the building facade, which facilitates the coordination with the building structure coordinate system and reduces the error of space conversion. The axes of the curtain wall surface three-dimensional coordinate system are strictly defined to match the monitoring requirements of the thermal plume: the X-axis extends along the horizontal direction of the curtain wall, which is used to represent the temperature difference of different horizontal areas of the curtain wall. This is because the thermal plume will be horizontally offset by wind force, curtain wall material joint gaps, etc. during its upward movement, and the X-axis can accurately capture this offset feature. The Z-axis is vertically upward, directly corresponding to the core movement direction of the upward movement of the thermal plume. The starting, development and termination height of the thermal plume all need to be quantified by the Z-axis coordinate. The Y-axis is perpendicular to the curtain wall and points to the outside, providing a spatial reference for measuring the temperature difference between the curtain wall surface and the outdoor environment. The temperature difference is the core driving force for the formation of the thermal plume. The setting of the Y-axis makes the temperature difference measurement have a clear spatial reference, avoiding errors caused by chaotic collection positions of environmental temperature.

[0061] The grid nodes are marked at a standard width a of the curtain wall unit block. This setting is directly adapted to the characteristics of the curtain wall industrial construction. The curtain wall is usually composed of standardized unit blocks. Marking the nodes at a interval of a can make each grid node correspond to the physical position of the unit block one by one, which facilitates the integration of temperature sensors in the pre-set installation position of the unit block, without the need for additional modification of the curtain wall structure, reducing the construction difficulty and cost. At the same time, the uniform interval a forms a fixed spatial resolution, ensuring that the temperature data collected in different areas and at different times have consistent spatial scales, providing standardized input for subsequent temperature gradient calculation and trajectory reconstruction.

[0062] In the prior art, the traditional dimming system lacks a unified coordinate reference matching the space features of the curtain wall, and the temperature data is mostly scattered single-point measurement values, which cannot establish spatial correlation, resulting in that the position, range and movement direction of the thermal plume cannot be accurately positioned. Step S11 gives all temperature data a clear spatial coordinate attribute by constructing a curtain wall surface three-dimensional coordinate system, so that the discrete temperature values are converted into a data set with spatial structure; the marking of the grid nodes further divides the curtain wall surface into standardized monitoring units, realizing full coverage and ordering of temperature monitoring. The adaptability design of the grid nodes and unit blocks takes into account the monitoring accuracy and engineering feasibility, avoiding data loss or excessive engineering cost caused by unreasonable arrangement of monitoring points; the standardized spatial resolution ensures the comparability and consistency of the data, making it possible to compare the thermal plume characteristics of different floors and different time periods. The historical thermal plume trajectory coordinates stored in the historical thermal plume trajectory database in subsequent step S13 are based on the same coordinate system as the current thermal plume trajectory, so that the similarity matching of the current thermal plume trajectory and the historical thermal plume trajectory has a unified spatial scale, without the need for additional coordinate conversion, significantly improving the matching efficiency and accuracy. If this step is missing, the thermal state temperature map in subsequent step S12 will be a disordered set of temperature values without spatial coordinates, which cannot calculate the temperature gradient to identify the thermal plume source, resulting in a lack of accuracy basis for thermal plume monitoring and regulation.

[0063] Step S12, constructing a thermal state temperature map in the curtain wall surface three-dimensional coordinate system, and constructing a current thermal plume trajectory line based on the thermal state temperature map;

[0064] In the prior art, the traditional dimming system relies on static single-point temperature or illumination data, which cannot capture the dynamic evolution process of the thermal plume, and cannot quantitatively describe its trajectory, resulting in a lack of accurate basis for dimming control to address the movement of the thermal plume. Step S12 addresses the core problem of real-time tracking and quantitative description of the thermal plume movement trajectory, providing key dynamic data support for subsequent trajectory prediction and coordinated dimming, focusing on "dynamicizing static data, continuousizing discrete data, and characterizing complex motion".

[0065] Further, step S12 includes:

[0066] Step S121, under the condition of sunlight, collecting the temperature of each grid node at a preset sampling period to generate a thermal state temperature map at each collection time;

[0067] Under the condition of sunlight, the temperature of the curtain wall surface grid node is collected by the distributed temperature sensor network at a preset sampling period. For example, the sampling period can be set to 2-10 seconds. The sensor network adopts a composite monitoring scheme combining an infrared thermal imager and a contact thermocouple: the infrared thermal imager covers the whole area of the curtain wall, providing high-resolution temperature distribution (spatial resolution up to 0.1a); the contact thermocouple is arranged at key nodes, such as plate joints, to compensate for the accuracy loss of infrared monitoring in complex reflection environment. This multi-modal data fusion method ensures the integrity and accuracy of the temperature field data. The thermal temperature map T(x, z, t) is stored in a three-dimensional matrix form, where x represents the X-axis coordinate of the grid node, corresponding to the horizontal position of the curtain wall; z represents the Z-axis coordinate, corresponding to the floor height; t represents the collection time, and the three together constitute a "space-time" three-dimensional data structure, which records the dynamic distribution of the curtain wall surface temperature completely. The high temperature of 60-70°C on the outer surface of the curtain wall and the formation of the rising hot air flow (thermal plume) in summer are caused by the direct action of solar radiation. Solar radiation is the core energy source for the temperature rise of the curtain wall surface and the formation of temperature difference with the outdoor environment. Without sunlight, such as at night or on rainy days, there is no significant heat accumulation on the curtain wall surface, the temperature difference with the environment is very small, and the thermal plume cannot be formed or its intensity can be ignored. Therefore, limiting the sunlight condition is to accurately focus on the nature of the problem, ensuring that the monitoring data is directly related to the core cause of "heat island microclimate".

[0068] Step S122, calculating the temperature gradient field based on the thermal temperature map, identifying and marking the thermal plume source point;

[0069] The calculation of the temperature gradient field needs to be calculated for each collection time of the thermal temperature map, and the vertical temperature gradient ATz(x, z, t) and the horizontal temperature gradient ATx(x, z, t) are calculated respectively. The calculation logic of ATz(x, z, t) is the difference between the temperature of the upper node and the current node at the same X-axis coordinate, that is, ATz(x, z, t) = T(x, z+1, t)-T(x, z, t); ATx(x, z, t) is the difference between the temperature of the right node and the current node at the same Z-axis coordinate, that is, ATx(x, z, t) = T(x+1, z, t)-T(x, z, t). The basis of this calculation logic is the basic law of heat transfer. The core movement of the thermal plume is vertical upward, and its formation is usually accompanied by a significant vertical temperature difference. ATz(x, z, t) directly reflects the strength of the rising hot air; ATx(x, z, t) reflects the lateral diffusion trend of the thermal plume, which can assist in judging the influence range of the thermal plume.

[0070] When the vertical temperature gradient ATz(x, z, t) is greater than or equal to the temperature gradient threshold a, the corresponding grid node is marked as a hot plume source point. The setting of the temperature gradient threshold a is based on the type of curtain wall material, and the setting method is: by comparing the thermal response characteristics of different material curtain walls under the same sunlight conditions, the minimum vertical temperature difference when the observable hot plume is formed is counted. Dark materials such as dark gray fluorocarbon paint aluminum plates have high solar absorptivity, and the surface temperature is easy to rise, so a smaller vertical temperature difference can form a hot plume, so a is set to a lower value; light-colored materials such as beige ceramic plates have low absorptivity, and a larger temperature difference is needed to form a hot plume, so a is set to a higher value; reflective materials such as mirror glass have low overall absorptivity, but local sunlight focusing is easy to occur, and the temperature of the focused area rises sharply, forming a local strong hot plume, so a lower a needs to be set to capture the local features. For example, when the curtain wall material is dark gray aluminum plate, a is set to 3°C; when it is beige ceramic plate, a is set to 5°C; when it is mirror glass, a is set to 2°C. When the ATz(x, z, t) of a certain grid node is greater than or equal to a, the grid node is marked as a hot plume source point, and its coordinates and corresponding temperature value are recorded.

[0071] The defect of the prior art is that only the high and low temperature of a single point is used to judge the heat source, and it is impossible to distinguish between "static high temperature points" and "dynamic hot plume source points". Some nodes may form static high temperature due to poor heat dissipation caused by material aging, but no rising hot air flow is generated. If a static high temperature point is mistakenly judged as a hot plume source point, it will lead to inaccurate control. Step S122 identifies the source point by temperature gradient instead of single-point temperature, accurately captures the core feature of "heat transfer strength", and avoids the interference of static high temperature. The temperature gradient field converts the static temperature distribution into a dynamic heat transfer trend, realizing the upgrade from "temperature value monitoring" to "heat flow direction identification"; based on the material characteristics, the temperature gradient threshold a is set, so that the hot plume source point identification is adapted to different curtain wall types, and the universality of the system is improved;

[0072] In step S123, the hot plume source points are connected according to the space-time double criteria, the current hot plume trajectory line is constructed, and the trajectory features of the current hot plume trajectory line are extracted. The current time label is associated with the current hot plume trajectory line.

[0073] The connection of thermal plume source points follows a dual "space-time" principle: Spatially, the horizontal distance between two thermal plume source points is less than a preset horizontal connection threshold Dx, and the vertical distance is less than a preset vertical connection threshold Dz. Dx and Dz are set based on the diffusion characteristics of the thermal plume; typically, Dx is twice the spatial resolution parameter (where the spatial resolution is the standard width 'a' of the curtain wall unit panel), and Dz is three times the spatial resolution parameter. This is because the lateral diffusion range of the thermal plume is limited during its ascent, while its vertical continuity is stronger. Temporally, the time interval between the appearance of two thermal plume source points is less than or equal to one acquisition cycle, ensuring that the connected thermal plume source points belong to thermal plumes formed in the same time period. The connection process adopts the "nearest neighbor principle," prioritizing the connection of thermal plume source points with the shortest spatial distance and temporal synchronization, forming a continuous thermal plume trajectory line.

[0074] The trajectory features specifically include: starting height h1, where h1 is the Z-axis coordinate of the lowest point of the current thermal plume trajectory, reflecting the origin floor of the thermal plume; ending height h2, where h2 is the Z-axis coordinate of the highest point of the current thermal plume trajectory, reflecting the influence range of the thermal plume; trajectory offset angle θ, where θ is the angle between the current thermal plume trajectory and the Z-axis, calculated by the difference in X-axis coordinates Δx and Z-axis coordinates Δz between adjacent nodes on the trajectory, θ=arctan(Δx / Δz), reflecting the horizontal offset caused by factors such as wind force; and peak temperature Tmax, where Tmax is the highest temperature value of all nodes on the current thermal plume trajectory, reflecting the intensity of the thermal plume. The current time stamp is associated with the current thermal plume trajectory. These trajectory feature parameters collectively constitute the identity label of the thermal plume, and are stored in association with the coordinate data of the corresponding thermal plume trajectory, the acquisition time, and environmental parameters.

[0075] The limitation of existing technologies is that even if discrete thermal plume source points are identified, they cannot be correlated into continuous trajectories. This results in the loss of key information such as the movement path and impact height of the thermal plume, failing to provide precise targeting basis for upper-level lighting. Step S123 connects the source points using a dual spatial-temporal criterion, transforming discrete points into continuous trajectories, while simultaneously extracting feature parameters to quantify the motion patterns. The continuous trajectory line fully presents the spatial movement path of the thermal plume, visualizing the "start-development-termination" process of the thermal plume; the trajectory feature parameters simplify the complex thermal plume motion into quantifiable indicators, providing structured input for historical trajectory matching in the subsequent step S13; the extraction of the trajectory offset angle θ can be correlated with environmental factors such as wind speed, providing data support for analyzing the coupling relationship between the thermal plume and the environment.

[0076] Step S12 systematically solves the core problem of dynamic monitoring and quantification of thermal plumes through "map construction - gradient calculation - trajectory extraction", realizing the "visualization, quantification and traceability" of thermal plume movement, breaking the limitation of the traditional system's "fuzzy perception" of thermal plumes; the time-series trajectory data and feature parameters provide multi-dimensional input for subsequent prediction and control, enabling active regulation to have a data foundation.

[0077] Step S13: Construct a historical thermal plume trajectory database, extract similar historical trajectories to the current thermal plume trajectory from the historical thermal plume trajectory database, analyze the common development patterns of similar historical trajectories, and predict the development path of the current thermal plume.

[0078] Further, step S13 includes:

[0079] Step S131: Collect the historical temperatures of each grid node that have the same sampling period and the same collection time as the current monitoring over the past n days, generate the historical thermal plume trajectory lines for each collection time over the past n days, and construct a historical thermal plume trajectory database.

[0080] The historical temperature collection duration *n* must meet statistical significance requirements. This duration is based on the cyclical patterns of thermal plume evolution. The thermal response of building curtain walls is affected by solar cycles and weather changes. A sample size of 30 days or more typically covers typical environmental conditions such as different solar intensities, outdoor temperatures, and wind speeds, eliminating interference from short-term, accidental factors and ensuring the database contains sufficiently diverse thermal plume evolution patterns; therefore, *n* ≥ 30. Data acquisition must strictly maintain consistency with current monitoring. Data collected at the same sampling period and time ensures uniform temporal resolution, providing a direct time reference for comparing historical and current trajectories. The generation of historical thermal plume trajectories fully reuses the method in step S12: historical temperatures are collected to generate a thermal temperature map, the temperature gradient field is calculated to identify the source points of the thermal plume, and then, based on a dual spatial-temporal criterion, the source points are connected to form historical thermal plume trajectories, and trajectory features are extracted. These trajectory features are then stored in the historical thermal plume trajectory database.

[0081] This reuse ensures that historical and current thermal plume trajectories are completely consistent in terms of feature dimensions and generation standards, avoiding incomparability issues caused by differences in technical logic. The storage structure of the historical thermal plume trajectory database needs to achieve a three-dimensional association between "trajectory-feature-environment." Each historical thermal plume trajectory is associated with its complete coordinate data, extracted historical starting and ending heights, historical trajectory offset angles, historical temperature peaks, and other historical trajectory feature parameters, as well as corresponding historical environmental parameters (solar intensity, outdoor temperature, wind speed) and historical time tags, such as date and collection time. Storage is performed with a dual index based on "time dimension + feature dimension." The time index facilitates quick location of historical data within the same time period, while the feature index provides an efficient retrieval path for subsequent similarity matching. A unified trajectory generation logic ensures the homogeneity and comparability of historical and current data, providing a prerequisite for accurate similarity matching. The three-dimensional associated storage structure achieves deep binding between trajectory features and environmental conditions, enabling subsequent predictions to simultaneously consider the coupling effects of the thermal plume's own characteristics and the external environment. The dual index design improves data retrieval efficiency and meets the real-time requirements of dimming control.

[0082] Step S132: Perform similarity matching between the current thermal plume trajectory and the historical thermal plume trajectory database to extract similar historical trajectories;

[0083] The trajectory features of the current thermal plume trajectory are compared with the historical trajectory features of historical thermal plume trajectories in the historical thermal plume trajectory database to determine spatial-temperature-temporal triple similarity. Historical thermal plume trajectories that meet the spatial-temperature-temporal triple similarity determination are considered as similar historical trajectories, resulting in a total of k similar historical trajectories. If k is less than the sample size threshold, the relevant parameters of the spatial-temperature-temporal triple similarity determination are adjusted until k is greater than or equal to the sample size threshold.

[0084] The spatial-temperature-temporal triple similarity determination includes spatial location similarity determination, temperature feature similarity determination, and temporal similarity determination. The conditions for spatial location similarity determination are: the absolute value of the difference between the historical starting height of the i'th historical plume trajectory and the starting height of the current plume trajectory is less than a spatial distance threshold, and the absolute value of the difference between the historical trajectory offset angle of the i'th historical plume trajectory and the trajectory offset angle of the current plume trajectory is less than an angle deviation threshold. The spatial distance threshold is set to an integer multiple of 'a', based on the vertical spacing of the grid nodes corresponding to the smallest unit of change in starting height. In other words, an integer multiple of 'a' ensures that the spatial accuracy of the difference judgment matches the monitoring. For example, it can be set to 2a. The angle deviation threshold is set based on the influence of wind force on plume offset. By statistically analyzing the fluctuation range of plume offset angles under different wind speeds, the minimum angle difference that can distinguish the influence of different wind forces is determined. When the wind speed is low, the offset angle fluctuation is small, and the angle deviation threshold can be set to a smaller value; when the wind speed fluctuation is large, the angle deviation threshold is appropriately widened to ensure the capture of trajectories under the influence of similar wind forces. For example, the angle deviation threshold is set to 10 degrees. i' is the index of the historical thermal plume trajectory line in the historical thermal plume trajectory database.

[0085] The condition for determining temperature feature similarity is as follows: the absolute value of the difference between the historical temperature peak value of the i'th historical thermal plume trajectory and the temperature peak value of the current thermal plume trajectory is less than the temperature deviation threshold. The temperature deviation threshold is based on the distinguishable range of thermal plume intensity. It analyzes the temperature peak fluctuations of similar thermal plume trajectories (i.e., those with the same starting height and trajectory offset angle) in the historical thermal plume trajectory database to determine the maximum temperature deviation that will not lead to significant differences in evolutionary patterns. Thermal plumes with higher intensity have slightly higher tolerance for temperature fluctuations, while those with lower intensity require strict control. For example, if the temperature deviation threshold is set to 5℃, and the current thermal plume trajectory's temperature peak value is 65℃, then the temperature feature similarity is satisfied if the historical thermal plume trajectory's historical temperature peak value is within the range of 60-70℃.

[0086] The condition for determining temporal similarity is that the absolute value of the difference between the historical time label of the i'th historical plume trajectory and the current time label of the current plume trajectory is less than the time deviation threshold. The time deviation threshold is based on the rate of change of the solar angle. If the solar angle changes slowly in a short period of time, the differences in solar conditions, ambient temperature, etc., at similar times are small, and the similarity of the plume evolution patterns is high. By statistically analyzing the impact of solar angle changes on plume characteristics, the time deviation threshold is determined to ensure environmental consistency at similar times. For example, the time deviation threshold can be set to 30 minutes. If the current time label of the current plume trajectory is 14:00, then the historical time labels of the historical plume trajectories within the range of 13:30-14:30 satisfy the time similarity requirement.

[0087] The sample size threshold is set to ensure the accuracy of thermal plume development path prediction. It requires a sufficient number of similar samples to support pattern analysis; for example, the sample size threshold is set to 3. When k < 3, the relevant parameters for spatial-temperature-temporal triple similarity determination are adjusted. These parameters refer to the core threshold parameters under the triple similarity determination dimensions, including the spatial distance threshold and angle deviation threshold for spatial location similarity determination, the temperature deviation threshold for temperature characteristic similarity determination, and the time deviation threshold for temporal similarity determination. Specifically, the relevant parameters for spatial-temperature-temporal triple similarity determination are adjusted by relaxing the core threshold parameters in the order of "space → angle → time," prioritizing dimensions with less impact on prediction accuracy. For example, the spatial distance threshold is relaxed from 2a to 3a, and the angle deviation threshold is relaxed from 10 degrees to 15 degrees, balancing sample size and accuracy. The reason for setting the sample size threshold is that the formation and development of thermal plumes are affected by multiple factors such as solar radiation intensity, material properties, and wind speed. Single or small historical trajectories are easily interfered with by accidental factors (such as instantaneous gusts or abnormal local material heat dissipation), failing to reflect the "typical evolutionary pattern under similar initial conditions." Only when the sample size reaches a certain scale can the common characteristics of multiple trajectories offset random fluctuations and present a stable evolutionary pattern. For example, for thermal plumes with the same initial height, trajectory offset angle, and peak temperature, a single historical trajectory may have a lower termination height due to instantaneous cloud cover. However, the common termination height of three or more similar trajectories can filter out such anomalies, forming a stable "initiation-development-termination" correlation logic, providing a reliable basis for prediction. The core reason for relaxing the threshold according to "space → angle → time" is that each dimension has a different impact on prediction accuracy. Prioritizing relaxation of the less influential dimensions is to ensure accuracy and supplement the sample size. Among them, the spatial dimension has the least impact: the core of the thermal plume rises vertically, the initial height deviation is 1-2 grid spacings, and the upper development path still converges, so relaxation has little impact; the angular dimension has a moderate impact: the offset angle is related to lateral diffusion, but the core predictions such as arrival time and temperature decay are stable, so relaxation can accommodate more samples; the temporal dimension has the greatest impact: it is directly related to core driving factors such as sunshine, and relaxation too early will cause prediction failure due to environmental differences, so it is relaxed last.

[0088] Existing technologies cannot achieve accurate similarity matching of thermal plume trajectories. Prediction bias often occurs due to a single matching dimension or insufficient sample size. Step S132 solves the problem of insufficient matching accuracy by comprehensively considering the influence of spatial location, temperature intensity, and temporal environment through a three-dimensional similarity judgment. The dynamic sample size adjustment mechanism balances the representativeness of the samples and the accuracy of the matching. Without this step, it would be impossible to select effective reference trajectories from historical data, and subsequent predictions would have to rely on real-time data, failing to reflect the guiding value of historical experience.

[0089] Step S133: Analyze the common development patterns of similar historical trajectories and predict the development path of the current thermal plume.

[0090] Based on the k similar historical trajectories matched in step S132, common features of their subsequent development paths are extracted to determine the evolution trajectory of the current thermal plume in the next m minutes. The prediction includes the predicted termination height h2. pre Maximum diffusion width w max Time t to reach each floor arrival (z'), predicted trajectory offset angle θ pre And the future thermal plume trajectory line, where z' is the floor identifier, these predictions provide core parameters for the risk assessment in step S21. h2 pre The arithmetic mean of the historical termination heights of k similar historical trajectories is used; the maximum diffusion width of the k similar historical trajectories is extracted, and the median is taken as w. max ;θ pre Let t be the arithmetic mean of the historical trajectory offset angles of k similar historical trajectories; calculate the average time for the thermal plume to reach each floor from the initial height in the similar historical trajectories, as t. arrival (z').

[0091] The generation of future thermal plume trajectories needs to be based on "real-time trajectories, reference to historical commonalities, and constraints by quantitative parameters." This multi-dimensional collaborative construction ensures consistency with the actual evolution of thermal plumes. First, the starting point is strictly anchored to the endpoint of the current thermal plume trajectory, which is the current actual spatial location of the plume. Using this as the starting point avoids "starting point misalignment" with historical trajectories, ensuring spatial continuity between the future trajectory and the currently monitored trajectory, and preventing the disruption of real-time movement trends. Secondly, it is necessary to rely on the matched k similar historical trajectories to extract their common characteristics of subsequent development as path references. These include the overall path trend extending from the "equivalent position to the current trajectory endpoint" to its own termination height, such as a unified direction of mainly vertical ascent accompanied by small horizontal offsets, the distribution range of historical trajectory offset angles, and the diffusion rhythm with increasing floor height. The distribution range of historical trajectory offset angles provides a stability reference for the horizontal offset of future trajectories. For example, if similar trajectories all show an offset of 5°-8°, the future trajectory will follow this range. The diffusion rhythm with increasing floor height can be, for example, the common pattern of slow horizontal diffusion at lower floors and faster diffusion at middle and higher floors. These characteristics ensure that the movement pattern of the future trajectory conforms to the historical evolution logic, rather than being an isolated coordinate splicing. Finally, the quantified prediction parameters need to be used as hard constraints: the vertical endpoint of the trajectory line must be aligned with h2. pre Precise alignment is required; the maximum horizontal extension range must be in the range of w. max As the boundary, the overall horizontal offset direction must fit θ. pre Through a triple collaboration of "starting point + common reference + parameter constraints", a trajectory extending from the current endpoint of the thermal plume to h2 is ultimately constructed. pre The continuous trajectory forms a complete trajectory line of the future thermal plume.

[0092] The evolution of thermal plumes exhibits regularity; similar initial conditions, such as starting height, trajectory deviation angle, temperature peak, and temporal environment, often lead to similar development paths. The subsequent evolution of historical trajectories provides a reliable reference for the current development of thermal plumes. Existing technologies can only monitor thermal plumes in real time and cannot predict their future trajectory and impact range, resulting in dimming control remaining in a passive response state. Step S133 achieves predictive functionality through historical trajectory pattern analysis, transforming passive response into proactive prevention. Without this step, the subsequent adjacent-floor coordinated dimming in step S20 would fail to timely block the thermal plume transmission due to a lack of lead time, significantly reducing the control effect. The prediction results of step S13 provide precise temporal and spatial basis for the control strategy formulation in step S20, enabling operations such as floor load classification and adjacent-floor compensation to be carried out in advance based on the expected path of the thermal plume. After coordination, a closed-loop connection of "monitoring-prediction-control" is achieved, significantly improving the system's response efficiency and control accuracy compared to single monitoring or single control.

[0093] Step S13 primarily addresses the technical challenge of lacking predictive capabilities for thermal plume development, specifically including two core issues: First, discrete heat source points cannot reflect the continuous upward path of thermal airflow. Even if existing technologies identify discrete thermal plume source points, they cannot correlate them into a continuous trajectory, resulting in the loss of crucial information such as the plume's movement path and impact height. Second, real-time detected thermal plumes lack historical references, making it impossible to predict their future development trends. Existing technologies rely solely on real-time data for passive responses, failing to utilize historical evolution patterns for guidance and control. By constructing a historical thermal plume trajectory database, employing triple similarity matching, and analyzing patterns, step S13 precisely solves these problems, achieving intelligent prediction of thermal plume development paths. The triple similarity determination, through multi-dimensional constraints of space, temperature, and time, ensures that the selected similar historical trajectories have high reference value. Predictions based on the common patterns of similar historical trajectories allow for precise prediction of the current thermal plume's future development path, such as termination height, diffusion width, and arrival time, breaking the limitation of traditional technologies that "can only monitor the present and cannot predict the future."

[0094] Step S20: Using the current thermal plume development path prediction results, establish a floor load classification sequence; construct the adjacent floor feedforward temperature compensation relationship based on the floor load classification sequence, and generate a floor transmittance control command sequence; convert the floor transmittance control command sequence into a time-series staggered dimming scheme, and implement time-series staggered dimming through spiral progressive control; after the time-series staggered dimming is executed, perform thermal-optical coupling feedback correction.

[0095] Further, step S20 includes:

[0096] Step S21: Using the current thermal plume development path prediction results, calculate the thermal intrusion risk index of each floor and establish a floor load classification sequence.

[0097] Please see Figure 2 As shown, step S21 further includes:

[0098] Step S211: Calculate the heat intrusion risk index for each floor based on the current heat plume development path prediction results;

[0099] The thermal intrusion risk index is a quantitative indicator that comprehensively reflects the degree of influence of thermal plumes on each floor. Its core function is to transform the dynamic characteristics of thermal plumes into a stratified load basis that can be used for dimming control. The predicted development path of thermal plumes includes parameters such as termination height, maximum diffusion width, and arrival time at each floor. These parameters provide the basic spatial and temporal inputs for index calculation.

[0100] Step S211 calculates the heat intrusion risk index for each floor based on the current predicted development path of the heat plume. First, the probability P(z') of the future heat plume trajectory passing through each floor is determined. This probability P(z') is obtained by relying on the motion data of the future heat plume trajectory within the predicted time window. The duration for which the future heat plume trajectory traverses the height range corresponding to floor z' is statistically analyzed, and the ratio of this duration to the predicted time window is determined as the probability P(z'). P(z') directly reflects the likelihood of a floor being covered by the heat plume; the longer the coverage time, the higher the basic probability of being affected. Then, the peak temperature coefficient C is calculated. temp Peak temperature coefficient C temp The calculation requires first obtaining the predicted maximum temperature of the thermal plume near floor z' in the forecast results, and then obtaining the outdoor ambient temperature at the same time. The ratio of the two is the temperature peak coefficient C. temp Ambient temperature is collected by distributed temperature and humidity sensors pre-installed on the building's exterior walls. The sensors are positioned away from direct sunlight and airflow dead zones to ensure data representativeness. The predicted maximum temperature from the thermal plume is extracted from the predicted development path, specifying the temperature parameters corresponding to the floor height. temp The thermal intensity of the heat plume itself is quantified; the higher the ratio, the greater the potential for raising indoor temperature after heat intrusion. Finally, the window opening ratio W(z') for each floor is obtained. This is achieved by real-time monitoring of window opening status using position sensors installed on windows on each floor, calculating the total area of ​​all open windows on that floor, and then comparing it to the total designed area of ​​the curtain wall windows on that floor. Windows are the direct channel for hot air to enter the room; the higher the window opening ratio, the smoother the actual path of heat intrusion, and the higher the risk transmission efficiency. The formula for calculating the heat intrusion risk index R(z') is R(z') = P(z') × C. tempThe design logic of the formula ×W(z') is based on the three elements of heat intrusion: contact probability, energy intensity, and transfer channel. The product of the three elements enables the index to comprehensively and accurately map the actual heat load state. Changes in any element will be directly reflected in the index value, avoiding the deviation caused by single-factor evaluation.

[0101] Step S212: Classify the heat load of each floor according to the heat intrusion risk index; record the heat load classification results in order of floor height to generate a floor load classification sequence.

[0102] Based on the magnitude of R(z'), the floors are dynamically divided into three heat load levels: First, a first threshold β1 and a second threshold β2 are set. The setting of β1 and β2 needs to take into account the usage characteristics and heat demand differences of building types: Office buildings have stable personnel density and long stays, and moderate sensitivity to temperature fluctuations; the threshold setting needs to balance energy saving and office comfort. Residential buildings have flexible personnel activity times, and high requirements for temperature stability during nighttime rest; the threshold setting needs to be more biased towards low load control. Commercial buildings have high personnel flow, high spatial openness, and frequent heat exchange; the threshold setting needs to adapt to high load tolerance. The setting method is to monitor the heat load of similar buildings for no less than 30 days, record the changes in indoor temperature and personnel comfort feedback under different R(z'), determine the maximum R(z') that does not affect comfort as β2, and determine the minimum R(z') that can maintain comfort without additional dimming as β1. Based on this, when R(z') < β1, floor z' is classified as a low-load floor; when β1 ≤ R(z') ≤ β2, floor z' is classified as a medium-load floor; and when R(z') > β2, floor z' is classified as a high-load floor. After classification, the classification results are arranged in order of floors from low to high or from high to low to form a floor load classification sequence, where each element in the sequence corresponds to a floor and its heat load level.

[0103] Step S21 addresses the problem that traditional static floor division cannot quantify dynamic heat load differences. Existing technologies often divide load zones based on fixed floor height or orientation, ignoring the trajectory and intensity fluctuations of heat plumes over time, leading to a mismatch between dimming control and actual heat load. By calculating the three elements of the heat intrusion risk index, a shift from "static zone division" to "dynamic risk quantification" is achieved. The probability P(z') captures the persistence of the time dimension's impact, and the peak temperature coefficient C... tempBy capturing the intensity of the energy dimension's influence and the window opening rate W(z') capturing the channel state of the transmission dimension, the synergy of these three factors ensures the accuracy and real-time performance of the heat load assessment for each floor. Step S21 provides the core basis for the subsequent neighboring floor collaborative control in step S22. The floor load classification sequence directly determines the basic dimming strategy and neighboring floor compensation intensity for each floor. Without this step, step S22 would lose the quantitative standard for tiered control, and could only adopt a uniform dimming mode, making it impossible to implement precise control for high-load floors. Step S21 achieves differentiated identification of floor heat load, shifting dimming control from "average distribution" to "on-demand distribution," while providing a tiered effect evaluation benchmark for subsequent feedback correction. By comparing the actual cooling amplitude of floors with different load levels, weak links in the control strategy can be quickly identified. In addition, the calculation of the heat intrusion risk index incorporates the human factor of window opening behavior into the heat load assessment, solving the problem of uncontrolled heat intrusion caused by human window opening in traditional systems, and enabling dimming control to dynamically adapt to load changes caused by user habits.

[0104] Step S22: Construct the adjacent floor feedforward temperature compensation relationship based on the floor load classification sequence, and generate the floor transmittance control command sequence based on the adjacent floor feedforward temperature compensation relationship.

[0105] Please see Figure 3 As shown, step S22 further includes:

[0106] Step S221: Based on the floor load classification sequence, construct the feedforward temperature compensation relationship between adjacent floors;

[0107] Step S222: Based on the floor's own heat load level and the feedforward temperature compensation relationship with adjacent floors, calculate the light transmittance adjustment amount for each floor to form a floor light transmittance control command sequence.

[0108] The adjacent-floor feedforward temperature compensation relationship is the core mechanism for establishing the correlation between dimming strategies on different floors. Its design logic stems from the vertical transfer characteristics of heat plumes. Heat plumes rise from their source point on lower floors and continuously conduct heat to the upper floors they pass through; dimming on a single floor cannot break this transmission chain. The floor load classification sequence provides load level input for this relationship, with different levels corresponding to different basic control requirements and compensation priorities. Based on the floor load classification sequence, an adjacent-floor temperature compensation relationship is established for each floor. The compensation rules for the adjacent floor temperature compensation relationship are set based on the physical laws of heat plume transfer: When the lower floor z'-1 is a high-load floor, it indicates that a large amount of high-temperature hot air is moving upward. As the next station in the transfer path, floor z' needs to reduce its light transmittance in advance to reduce its own heat absorption and avoid superposition effect with the rising hot air. Therefore, the adjacent floor compensation amount δ1 is set. When the upper floor z'+1 is a high-load floor, it indicates that the hot plume has reached this height and may form a local high-pressure zone. The air on floor z' is easily drawn upward to form a "chimney effect" and accelerate heat transfer. Therefore, the adjacent floor compensation amount δ2 is set to reduce its own light transmittance and slow down the upward speed of the airflow. The settings of δ1 and δ2 need to be determined through heat plume transfer experiments. In these experiments, the transfer rate of the heat plume between floors is monitored under different load levels. The minimum transmittance adjustment range that reduces the transfer rate is statistically analyzed and used as the base compensation amount. This is then adjusted based on the building's floor height. The greater the floor height, the longer the heat plume diffuses between floors, and the compensation amount can be appropriately reduced; conversely, the smaller the floor height, the faster the transfer, and the compensation amount needs to be appropriately increased. When floor z' faces high loads from adjacent floors above and below, δ1 and δ2 are superimposed to form a dual compensation to cope with the bidirectional heat transfer pressure.

[0109] For floors with different load levels, different basic light transmittance adjustment values ​​δ are set. base The low-load layer is minimally affected by the thermal plume and requires no additional adjustment to its transmittance; therefore, δ base Set to 0 to maintain the standard light transmittance required for indoor lighting; the medium-load layer is moderately affected and needs to appropriately reduce light transmittance to suppress heat absorption, δ base The settings are based on the average temperature rise under medium load, ensuring that adjustments can offset most of the temperature increase; the high-load layer is most significantly affected, requiring a substantial reduction in transmittance, δ base The setting aims to block the main heat from entering. The δ values ​​corresponding to each load level... base Long-term field testing and calibration are required to record different δ values. base For each floor with different load levels, the minimum adjustment range that maintains the temperature within the comfortable range is selected as the final value. Based on the basic adjustment, compensation amounts δ1 or δ2 from adjacent floors are added to obtain the light transmittance adjustment. The calculation of the light transmittance adjustment must follow the principle of "basic adjustment first, compensation added later," first determining the δ corresponding to its own load. baseThen, based on the status of adjacent floors, it is determined whether to add δ1, δ2, or the sum of both. The floor transmittance control command sequence consists of transmittance adjustment amount for each floor, floor identification, execution priority, and other information. The execution priority is set according to the load level, with high-load floors having higher command priority than medium- and low-load floors, ensuring that key areas are regulated first.

[0110] Step S22 addresses the problem that single-floor dimming cannot interrupt the heat plume transfer chain. In existing technologies, each floor's dimming system operates independently, adjusting only based on its own temperature or illumination data. This causes the heat plume generated in the lower floors to continuously transfer upwards, making the upper floors a heat "collection zone," where even dimming in the upper floors cannot offset the heat input from the lower floors. By using a neighboring-floor feedforward temperature compensation relationship, the dispersed floor dimming is transformed into a collaborative system. When the lower floors are under high load, the upper floors intervene in advance; when the upper floors are under high load, the middle floors actively buffer, forming a vertical heat transfer blocking zone. The floor load grading sequence provides clear triggering conditions for the neighboring-floor feedforward temperature compensation relationship. Compensation is only activated when the neighboring floor is at a high load level, avoiding insufficient illumination caused by indiscriminate compensation. The neighboring-floor feedforward temperature compensation relationship translates the grading results into specific control actions, ensuring that the value of the grading is reflected in the dimming effect. Without step S22, the grading results of step S21 can only be used for single-floor dimming, failing to leverage the advantages of layered collaboration. The problem of vertical heat plume transfer remains unresolved, and the phenomenon of abnormally high room temperatures in high-rise buildings persists. Step S22 achieves "joint prevention and control" between floors, breaking the limitations of traditional systems that operate independently. Through feedforward compensation, the control action is made earlier than the arrival time of the heat plume, shifting from passive response to active prevention. This compensation mechanism can indirectly suppress the intensification of the "chimney effect." When the light transmittance of the middle floors is reduced, the temperature gradient on the curtain wall surface tends to be gentler, and the upward airflow power is weakened. This not only reduces the heat intrusion from the upper floors but also reduces the overall heat exchange intensity of the building, further optimizing energy consumption.

[0111] Step S23: Convert the floor transmittance control command sequence into a time-staggered dimming scheme, and implement time-staggered dimming through spiral progressive control;

[0112] The core of step S23 is to solve the dual problems of optical shock and electrical power peak caused by large-area synchronous dimming. Optical shock manifests as the visual stimulation of indoor and outdoor personnel caused by instantaneous changes in brightness, while power peak originates from the sudden increase in circuit load caused by the simultaneous start and stop of a large number of dimming units. Both reduce the practicality and safety of the system. The combination of spiral progressive control and timing misalignment strategy achieves the smoothness of the dimming process and the balance of electrical load.

[0113] Further, step S23 includes:

[0114] Step S231: Number the grid nodes on the curtain wall surface according to the spiral path, and assign the grid nodes to different dimming groups according to the node numbers to form a spiral progressive dimming node sequence.

[0115] The implementation of node numbering along a spiral path requires using the three-dimensional coordinate system of the curtain wall surface constructed in step S11 as a reference. The starting point for numbering is determined to be the center of the bottom of the curtain wall corresponding to the origin of the coordinate system. This starting point is chosen based on the laws of visual perception; the brightness change in the central area has a relatively gentle impact on the human eye, and expanding outward from the center can reduce the perceived intensity of visual abrupt changes. The numbering direction is clockwise, based on the fact that most human eyes are more adaptable to dynamic changes in a clockwise direction, and it matches the common installation sequence of curtain wall unit panels, reducing spatial misalignment between the numbering and the physical structure. The numbering process starts from the center point and extends outward in a clockwise direction, with each circle covering adjacent grid nodes, ensuring that the numbering sequence and spatial position form a continuous spiral trajectory, and each grid node corresponds to a unique number value. The grouping rule is based on modulo operation of the node numbers. A group number n1 is set, and all nodes are divided into n1 dimming groups according to the remainder when the node number is divided by n1. The setting of the group number n1 needs to balance dimming smoothness and execution efficiency: if n1 is too small, adjacent nodes are easily grouped together, failing to eliminate optical impact; if n1 is too large, the total dimming time is too long, affecting the real-time performance of thermal control. The setting method is to record the subjective evaluation of the dimming process by personnel when n1 is at different values ​​through visual comfort testing. Combined with the total dimming time requirement, the minimum n1 that can eliminate visual impact and meet real-time requirements is determined. Typically, n1 is set to 4. For example, with n1=4, nodes whose numbers remainder 0 modulo 4 are assigned to group 1, nodes whose numbers remainder 1 modulo 4 are assigned to group 2, nodes whose numbers remainder 2 modulo 4 are assigned to group 3, and nodes whose numbers remainder 3 modulo 4 are assigned to group 4, forming a spiral progressive dimming node sequence consisting of 4 dimming groups.

[0116] Step S232: Determine the timing of each dimming group's timing misalignment execution time based on the floor transmittance control command sequence; execute timing misalignment dimming based on the spiral progressive dimming node sequence and the timing misalignment execution time.

[0117] The timing of the staggered execution is determined starting from the reference time t1, which is the starting point of the first control cycle after the generation of the floor transmittance control command sequence, ensuring minimal time delay between command generation and execution. The time interval Δt between each dimming group needs to be related to the transmittance adjustment amount. The larger the adjustment amount, the more significant the brightness change of the dimming unit, and the higher the sensitivity of the human eye to this change, requiring a longer interval to adapt; the smaller the adjustment amount, the smoother the change, and the shorter the interval can be to improve efficiency. The method for setting Δt is to establish the correspondence between transmittance adjustment amount and adaptation time by measuring the visual adaptation time under different transmittance adjustment amounts, and to use 70% of the adaptation time as Δt, ensuring both comfort and avoiding excessively low efficiency. The execution time of n1 dimming groups is: the first group executes at time t1, and the n1 group executes at time t1 + (n1-1)Δt. For example, Figure 4 As shown, when n1=4, the four groups execute dimming sequentially according to the time axis: Group 1 executes at time t1, Group 2 executes at time t1+Δt, Group 3 executes at time t1+2Δt, and Group 4 executes at time t1+3Δt. During execution, the grid nodes within each dimming group synchronously receive transmittance adjustment commands and complete the brightness change according to the preset adjustment rate. The adjustment rate matches Δt, ensuring that the nodes within the group complete dimming before the adjacent group starts, avoiding cross-group time overlap.

[0118] Step S23 addresses the optical impact and power peak issues of traditional synchronous dimming. In existing technologies, all dimming units operate simultaneously, causing a sudden and dramatic change in the brightness of the curtain wall surface, creating a strong contrast between light and dark. Furthermore, the simultaneous power consumption of numerous units leads to a surge in circuit current, potentially triggering overload protection. By numbering nodes along a spiral path, adjacent grid nodes are assigned to different dimming groups, avoiding synchronous changes in the same area and visually presenting a gradual change effect spreading outwards from the center, reducing the perception of sudden changes in brightness. The staggered execution further distributes the electrical load across multiple time segments, with only 1 / n1 of the dimming units operating at any given moment, keeping the circuit load within a safe range. Step S23 transforms the abstract instruction sequence generated in step S22 into executable physical actions. Without this step, the collaborative dimming strategy of step S22 will fail due to improper execution, failing to achieve the expected heat protection effect and potentially even causing system failure. Step S23 achieves a dual improvement in visual comfort and electrical safety. Simultaneously, the spatial distribution of the spiral path complements the vertical upward trajectory of the thermal plume. Dimming the central area first preferentially suppresses heat absorption by the core thermal plume, further enhancing the thermal control effect. The staggered execution timing enhances the system's adaptability to grid voltage fluctuations; dispersed load changes are less likely to cause voltage drops, improving the system's reliability in unstable power supply environments. Node numbering is based on standardized grid nodes, ensuring that the numbering rules are consistent with the spatial resolution of the monitoring units. This allows the division of dimming groups to accurately correspond to the thermal plume's influence area, avoiding misalignment between dimming and thermal load areas.

[0119] Step S24: After the timing-displacement dimming is executed, the temperature spectrum after dimming is collected, and the temperature spectrum after dimming is compared with the thermal temperature spectrum before dimming to perform thermal-optical coupling feedback correction.

[0120] Thermal-optical coupling feedback correction refers to the reverse correction of key parameters affecting thermal plume monitoring and dimming control by monitoring the actual thermal effect after dimming, so that the system can adapt to performance deviations caused by environmental changes and its own aging. The core is to establish a closed-loop logic of "control-monitoring-correction".

[0121] Please see Figure 5 As shown, step S24 further includes:

[0122] Step S241: After the timing-shifted dimming is executed, the temperature spectrum after dimming is collected and compared with the thermal temperature spectrum before dimming to calculate the actual cooling amplitude of each grid node.

[0123] The acquisition of the post-dimming temperature spectrum needs to reuse the monitoring system of S121. It should be initiated within the first sampling cycle after the dimming is completed to ensure the shortest time interval between acquisition and dimming completion, minimizing environmental interference. The acquisition method is consistent with S121, using a combined monitoring scheme of infrared thermal imager and contact thermocouples to obtain the temperature of each grid node and generate the post-dimming temperature spectrum. The actual cooling amplitude of each grid node is calculated by subtracting the temperature value of the corresponding grid node in the post-dimming temperature spectrum from the temperature value of each grid node in the pre-dimming thermal temperature spectrum. A positive actual cooling amplitude indicates a cooling effect from dimming, while a negative amplitude indicates an abnormal temperature rise.

[0124] Step S242: Calculate the expected cooling range for each grid node, compare the actual cooling range of each grid node with the calculated expected cooling range, and establish a performance deviation record table.

[0125] The calculation of the expected cooling range is based on the predicted thermal plume development path in step S133 and the transmittance adjustment in step S222. The core logic is that there is a correlation between the transmittance adjustment and the cooling effect; the same transmittance adjustment produces different cooling effects depending on the thermal plume intensity. Specifically, the calculation process first extracts similar historical trajectories from the historical thermal plume trajectory database that satisfy the spatial-temperature-temporal similarity criteria with the current thermal plume trajectory. The historical cooling range of these similar historical trajectories under the same transmittance adjustment is obtained, and the statistical mean of the historical cooling range is calculated as the basic expected value. Then, it is corrected by combining the current thermal plume's peak temperature Tmax with the average peak temperature Tavg of similar historical trajectories. The correction coefficient Kcorr = Tmax / Tavg. The final expected cooling range = basic expected value × Kcorr. This calculation method is based on the fact that thermal plume intensity directly affects the cooling demand and effect; the higher the peak temperature, the greater the theoretically achievable cooling potential under the same transmittance adjustment. The correction coefficient makes the expected value more closely match the actual characteristics of the current thermal plume. The calculation method for performance deviation is: expected cooling rate minus actual cooling rate. A positive performance deviation indicates that the actual cooling rate did not meet expectations, indicating insufficient thermal control; a negative performance deviation indicates that the cooling rate exceeded expectations, which may lead to insufficient sunlight. The performance deviation record table should include the time of deviation occurrence, location of deviation, value of performance deviation, and environmental parameters at the time (outdoor temperature, wind speed, solar radiation intensity). This information provides multi-dimensional input for subsequent deviation trend analysis, and environmental parameters can help distinguish whether the deviation is caused by system parameter issues or sudden changes in the external environment.

[0126] Step S243: Analyze the deviation trend in the effect deviation record table and perform adaptive adjustments to the key parameters.

[0127] Key parameters include the temperature gradient threshold α, neighboring layer compensation amounts δ1 and δ2, and the prediction time window m. These parameters affect the sensitivity of thermal plume source identification, the intensity of neighboring layer coordination, and the time lead of trajectory prediction, respectively, and are all core factors determining the system's control accuracy. Deviation trend analysis requires setting statistical standards: a single analysis cycle consists of k' consecutive dimming cycles, with k' set based on the adjustment frequency to ensure sufficient deviation samples are included within the cycle to exclude random factors; typically, k' is set to 3. A deviation coverage threshold of 60% of the grid nodes is used to ensure that the deviation is regionally universal rather than a local anomaly. When more than 60% of the grid node regions show positive deviations in k' consecutive dimming cycles, it indicates that the system's response to the thermal plume is lagging or the control intensity is insufficient, requiring increased sensitivity and control intensity: reducing α allows the system to identify thermal plume sources earlier and initiate control earlier; increasing δ1 and δ2 strengthens the blocking effect of neighboring layer coordination; shortening m makes the prediction results closer to real-time changes and reduces time lag. When negative deviations occur in more than 60% of the area, it indicates that the system response is too aggressive, and the sensitivity and control intensity need to be reduced: increasing α can avoid misjudging static high temperature as the source of the thermal plume; decreasing δ1 and δ2 can prevent excessive dimming leading to insufficient illumination; extending m can make the prediction more stable and avoid frequent adjustments. The magnitude of parameter adjustments needs to be determined based on the magnitude of the deviation; the larger the absolute value of the deviation, the larger the adjustment magnitude. The setting method is to establish a linear relationship between the mean effect deviation and the adjustment magnitude to ensure that the adjustment amount of key parameters matches the degree of deviation. The adjusted key parameters need to be updated in the system configuration, and the adjustment event and adjustment magnitude should be marked in the effect deviation record table to provide a reference for deviation analysis in subsequent cycles.

[0128] Step S24 addresses the problem of fixed parameters failing to adapt to dynamic environments. In existing technologies, system parameters are mostly fixed values ​​preset at the factory. With seasonal changes, aging of curtain wall materials, and changes in environmental conditions, the deviation between fixed parameters and actual needs gradually increases, leading to a decline in control effectiveness. Through thermal-optical coupling feedback correction, dynamic adaptive parameter control is achieved, enabling the system to continuously match the evolution of thermal plumes and dimming requirements. Step S24 provides the system with self-learning capabilities and is a key link in connecting the "monitoring-prediction-control" closed loop. Without step S24, the system control accuracy will continuously decline over time, making it impossible to maintain the effects of reducing high-rise room temperature and optimizing energy consumption in the long term. Adaptive adjustment can improve the long-term stability of system performance. Parameter adjustment can offset the effects of long-term factors such as material aging and seasonal changes. At the same time, the correlation of environmental parameters in the deviation record table can help identify the impact of special weather (such as sudden gusts or cloudy skies turning sunny) on the control effect, providing data support for subsequent optimization.

[0129] Example 2:

[0130] This embodiment, based on Embodiment 1, provides a building ventilation heat load control system based on thermal plume prediction, such as...Figure 6 As shown, it includes:

[0131] Thermal plume trajectory construction module: used to construct a three-dimensional coordinate system on the curtain wall surface, and to construct the current thermal plume trajectory line in the three-dimensional coordinate system on the curtain wall surface;

[0132] Thermal plume trajectory prediction module: used to build a historical thermal plume trajectory database, extract similar historical trajectories of the current thermal plume trajectory line from the historical thermal plume trajectory database, analyze the common development patterns of similar historical trajectories, and predict the development path of the current thermal plume.

[0133] Dimming Module: Utilizes the current thermal plume development path prediction results to establish a floor load classification sequence; constructs a feedforward temperature compensation relationship between adjacent floors based on the floor load classification sequence to generate a floor transmittance control command sequence; converts the floor transmittance control command sequence into a time-shifted dimming scheme, and implements time-shifted dimming through spiral progressive control;

[0134] Feedback correction module: After timing misalignment dimming is executed, it is used to perform thermal-optical coupling feedback correction.

[0135] Furthermore, in the dimming module, the method for establishing the floor load classification sequence includes:

[0136] Step S211: Calculate the heat intrusion risk index for each floor based on the current heat plume development path prediction results;

[0137] Step S212: Classify the heat load of each floor according to the heat intrusion risk index; record the heat load classification results in order of floor height to generate a floor load classification sequence.

[0138] Furthermore, in the dimming module, the method for implementing timing-shifted dimming through spiral progressive control includes:

[0139] Step S231: Number the grid nodes on the curtain wall surface according to the spiral path, and assign the grid nodes to different dimming groups according to the node numbers to form a spiral progressive dimming node sequence.

[0140] Step S232: Determine the timing of each dimming group's timing misalignment execution time based on the floor transmittance control command sequence; execute timing misalignment dimming based on the spiral progressive dimming node sequence and the timing misalignment execution time.

[0141] The methods and systems of this application may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the method is for illustrative purposes only, and the steps of the method of this application are not limited to the order specifically described above, unless otherwise specifically stated.

[0142] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0143] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for regulating building ventilation heat load based on thermal plume prediction, characterized in that, The method includes: A three-dimensional coordinate system for the curtain wall surface is constructed, and a thermal temperature map is built in the three-dimensional coordinate system. Based on the thermal temperature map, the current thermal plume trajectory line is constructed. A historical thermal plume trajectory database is constructed, and similar historical trajectories to the current thermal plume trajectory line are extracted from the historical thermal plume trajectory database. Based on the similar historical trajectories, the development path prediction result of the current thermal plume is predicted. The method for constructing the current hot plume trajectory line based on the thermal temperature map includes: calculating the temperature gradient field based on the thermal temperature map, identifying and marking the hot plume source points; connecting the hot plume source points according to the space-time dual criterion to construct the current hot plume trajectory line, and associating the current hot plume trajectory line with the current time label; Using the current thermal plume development path prediction results, a floor load classification sequence is established; based on the floor load classification sequence, an adjacent floor feedforward temperature compensation relationship is constructed; based on the adjacent floor feedforward temperature compensation relationship, a time-series staggered dimming is implemented through spiral progressive control; after the time-series staggered dimming is executed, thermal-optical coupling feedback correction is performed. The method for establishing a floor load classification sequence includes: calculating the heat intrusion risk index for each floor based on the current heat plume development path prediction results; classifying the heat load for each floor according to the heat intrusion risk index; recording the heat load classification results in order of floor height to generate a floor load classification sequence. The method for implementing time-shifted dimming based on the adjacent-layer feedforward temperature compensation relationship and through spiral progressive control includes: generating a floor transmittance control command sequence based on the adjacent-layer feedforward temperature compensation relationship; marking grid nodes on the curtain wall surface, numbering the grid nodes on the curtain wall surface according to a spiral path, and assigning the grid nodes to different dimming groups according to the node numbers to form a spiral progressive dimming node sequence; determining the time-shifted execution time of each dimming group according to the floor transmittance control command sequence; and executing time-shifted dimming according to the spiral progressive dimming node sequence and the time-shifted execution time.

2. The building ventilation heat load control method based on thermal plume prediction according to claim 1, characterized in that, The method for constructing the thermal temperature map includes: Under sunlight conditions, the temperature of each grid node is collected according to a preset sampling period to generate a thermal temperature map at each sampling moment.

3. The building ventilation heat load control method based on thermal plume prediction according to claim 2, characterized in that, The identification of the thermal plume source point includes: For the thermal temperature spectrum at each acquisition time, calculate the vertical temperature gradient; When the vertical temperature gradient is greater than or equal to the temperature gradient threshold α, the corresponding grid node is marked as a thermal plume source point.

4. The building ventilation heat load control method based on thermal plume prediction according to claim 3, characterized in that, The current thermal plume development path prediction results include at least the future thermal plume trajectory line; The methods for calculating the heat intrusion risk index for each floor include: Based on the current prediction of the thermal plume's development path, the probability P(z') of the future thermal plume trajectory passing through each floor is calculated, and the peak temperature coefficient C is also calculated. temp And the window opening ratio W(z') for each floor, where z' is the floor identifier; The heat intrusion risk index R(z') for each floor is equal to P(z') and C. temp The product of W(z').

5. The building ventilation heat load control method based on thermal plume prediction according to claim 4, characterized in that, The method for classifying the heat load of each floor includes: When R(z') < β1, floor z' is classified as a low-load floor; When β1≤R(z')≤β2, floor z' is classified as a medium load floor; When R(z') > β2, floor z' is classified as a high-load floor; where β1 is the first threshold and β2 is the second threshold.

6. The building ventilation heat load control method based on thermal plume prediction according to claim 5, characterized in that, The historical thermal plume trajectory database includes historical thermal plume trajectory lines from the past n days; The method for extracting similar historical trajectories of the current thermal plume trajectory includes: Extract the trajectory features of the current thermal plume trajectory and the historical trajectory features of historical thermal plumes; The trajectory characteristics of the current thermal plume trajectory and the historical trajectory characteristics of historical thermal plume trajectories are used to determine spatial-temperature-temporal triple similarity. Historical thermal plume trajectories that meet the spatial-temperature-temporal triple similarity determination are taken as similar historical trajectories, resulting in k similar historical trajectories. If k is less than the sample number threshold, the relevant parameters of the spatial-temperature-temporal triple similarity determination are adjusted until k is greater than or equal to the sample number threshold.

7. The building ventilation heat load control method based on thermal plume prediction according to claim 6, characterized in that, The historical trajectory characteristics of the historical thermal plume trajectory lines are associated with historical time labels; The spatial-temperature-temporal triple similarity determination includes spatial location similarity determination, temperature feature similarity determination, and temporal similarity determination; The condition for determining time similarity is that the absolute value of the difference between the historical time tag of the i'th historical thermal plume trajectory and the current time tag of the current thermal plume trajectory is less than the time deviation threshold.

8. A building ventilation heat load control system based on thermal plume prediction, used to implement the building ventilation heat load control method based on thermal plume prediction as described in any one of claims 1-7, characterized in that, The system includes: Thermal plume trajectory construction module: used to construct a three-dimensional coordinate system for the curtain wall surface, construct a thermal temperature map in the three-dimensional coordinate system for the curtain wall surface, and construct the current thermal plume trajectory line based on the thermal temperature map; Thermal plume trajectory prediction module: used to build a historical thermal plume trajectory database, extract similar historical trajectories of the current thermal plume trajectory line from the historical thermal plume trajectory database, and predict the development path of the current thermal plume based on similar historical trajectories; Dimming module: Utilizes the current thermal plume development path prediction results to establish a floor load classification sequence; Based on the floor load classification sequence, constructs an adjacent floor feedforward temperature compensation relationship; Based on the adjacent floor feedforward temperature compensation relationship, implements time-staggered dimming through spiral progressive control; Feedback correction module: After timing misalignment dimming is executed, it is used to perform thermal-optical coupling feedback correction.

Images