Window and family environment linkage remote control system and control method thereof

Abstract

The invention discloses a remote control system for linkage of a window and a home environment and a control method of the remote control system, particularly relates to the field of smart home, is used for solving the problems of dynamic optimization of indoor air quality and efficient utilization of window resources, and comprises the following steps: dividing an indoor environment into a plurality of controllable subspaces and distributing window groups; a fluid dynamics air circulation model is constructed in combination with sensor data, and the contribution degree of the window to the air quality is quantified; a contribution degree decline curve is generated through marginal benefit analysis, a critical subspace which achieves expectation and is smooth in benefit is identified, and a window strategy is maintained; for the subspace with improvement potential, the opening number and the opening degree of the windows are accurately adjusted according to the curve, and efficient utilization of resources is ensured; and training historical data by using a machine learning model, optimizing prediction of an optimal window combination, and realizing adaptive and multi-window cooperative control optimization based on real-time data dynamic adjustment.

Description

Technical Field

[0001] The present invention relates to the field of smart home, and more specifically, to a remote control system for linking a window with a home environment and a control method thereof. Background Art

[0002] In modern smart home systems, it has become a common method to optimize indoor ventilation and air quality by controlling multiple windows at the same time. However, different windows contribute differently to improving indoor air quality, and after a certain level, the improvement in air quality brought about by opening multiple windows at the same time shows a diminishing marginal effect. For example, initially opening a window in a suitable position may significantly improve indoor air quality, but when the indoor air has reached a better state, opening more windows additionally will not only be difficult to obtain the same improvement gain, but may also lead to excessive changes in environmental factors such as indoor temperature, humidity and noise, and even increase energy loss and user discomfort. Therefore, how to optimize and replan the window opening strategy based on the current indoor and outdoor air quality data and window coordination model through a refined analysis model and dynamic decision-making mechanism, and accurately evaluate the marginal effect of the coordinated opening of multiple windows has become a technical difficulty and research hotspot. In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0003] In order to overcome the above-mentioned defects of the prior art, the embodiment of the present invention provides a remote control system and a control method for linking windows with home environments. The indoor environment is precisely divided into multiple controllable subspaces, and a corresponding window group is assigned to each subspace. In combination with high-frequency sensors to collect various data in real time, an air circulation model based on fluid dynamics is constructed to quantify the specific contribution of each window to the improvement of air quality. The contribution decreasing curve is generated by marginal benefit analysis to identify the critical subspaces that have reached the expected air quality and where the marginal benefit tends to be flat, and the window opening strategy is maintained unchanged; at the same time, for subspaces with the potential for improving air quality, the number and degree of window opening are accurately adjusted according to the decreasing curve to ensure efficient use of resources. Furthermore, through the accumulation and training of historical data by machine learning models, the model's prediction ability for the optimal window combination under different environmental conditions is optimized, and online analysis and prediction are performed based on real-time data to dynamically adjust the window opening combination of each subspace to achieve adaptive and refined multi-window collaborative control optimization. The system scientifically divides the environment, accurately collects and processes data, optimizes efficient strategies, and implements intelligent real-time control, which significantly improves the accuracy and efficiency of indoor air quality management, reduces resource waste, and improves the comfort and health level of the home environment, thereby solving the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A remote control method for linking a window with a home environment comprises the following steps:

[0006] S1: Divide the indoor environment into multiple subspaces according to the building layout and functional areas, and assign corresponding window groups to each subspace;

[0007] S2: Deploy multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and store these data in a structured manner in a unified cloud database;

[0008] S3: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is constructed for each subspace to quantify the specific contribution of each window to the improvement of air quality;

[0009] S4: Input the window contribution index output by the air circulation model into the marginal benefit analysis module, and generate the contribution reduction curve under the change of the number and opening degree of each window through historical data fitting, so as to clarify the marginal improvement effect of different opening combinations;

[0010] S5: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste;

[0011] S6: For subspaces in a critical state, the initial window opening strategy is maintained unchanged. At the same time, for subspaces whose air quality improvement potential reaches the set threshold, the number or degree of window opening is increased to the optimized parameter according to the decreasing curve to ensure efficient use of resources;

[0012] S7: Feed the dynamic window opening strategy and real-time environmental data of each subspace into the machine learning model to accumulate and train historical data and optimize the model's ability to predict the optimal window combination under different environmental conditions;

[0013] S8: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

[0014] In a preferred embodiment, step S1 includes the following contents:

[0015] Obtain a detailed floor plan of the building, and apply a graph-theory-based space partitioning algorithm based on the functional attributes of the room to divide the entire indoor environment into several subspaces with consistent functions and without interfering with each other. Then, for each divided subspace, identify its corresponding window group to ensure that each group of windows can effectively cover the air circulation area of ​​the corresponding subspace.

[0016] In a preferred embodiment, step S2 includes the following contents:

[0017] Various sensors are deployed in each subspace. The sampling frequency of all sensors is set, and the data collected by different sensors are structured and stored in the cloud database in a unified JSON format. In the cloud database, a special table structure is established to store the environmental parameters and window status data of each subspace in order according to timestamps, and an independent index is created for each type of parameter.

[0018] In a preferred embodiment, step S3 includes the following contents:

[0019] A fluid dynamics simulation method was adopted, and the architectural layout, window position and size, indoor and outdoor pressure difference and measured air flow velocity field of each subspace were combined as boundary conditions. The concentrations of PM2.5, carbon dioxide and volatile organic compounds were used as inputs. The indoor pollutant concentration distribution under different window opening combinations was simulated by numerically solving the Navier-Stokes equations and the diffusion equation. The equilibrium concentration of pollutants under the baseline state was defined, that is, when all windows were closed, the predicted equilibrium pollutant concentration was obtained by setting the window opening state parameter to closed and running the simulation. For each window in each subspace, the target window was opened separately and the other windows were kept closed, and the simulation was repeated to obtain the predicted equilibrium pollutant concentration when only the target window was opened. Based on the difference between the baseline concentration and the single window opening concentration, the air quality improvement contribution index of each window in the corresponding subspace was calculated. The contribution index represents the relative proportion of the reduction in pollutant concentration caused by window opening. The contribution index of all windows in each subspace was inserted into the model result table of the cloud database in the form of structured records.

[0020] In a preferred embodiment, step S4 includes the following contents:

[0021] For each subspace, the contribution index of all windows is retrieved, and the corresponding historical time series data of the number of windows opened and the window opening degree are collected; the number of windows opened and the window opening degree are taken as independent variables, and the contribution index is taken as the dependent variable. The polynomial regression analysis method is used to establish a nonlinear relationship model between the contribution of windows in each subspace and the number of windows opened and the window opening degree; by fitting the polynomial regression model, the contribution decreasing curve of each window in each subspace is generated to describe the changing trend of the contribution index when the number of windows opened and the window opening degree increase; these decreasing curve data are structured and recorded and stored in the marginal benefit analysis table of the cloud database.

[0022] In a preferred embodiment, step S5 includes the following contents:

[0023] a. Perform curvature analysis on each decreasing curve and calculate the second-order derivative for each combination of window opening quantity and window opening degree to evaluate the smoothness of marginal benefit change;

[0024] b. For each subspace, the second-order derivative values ​​of all windows are integrated. If the absolute values ​​of the second-order derivatives of all windows are lower than the preset corresponding threshold, the corresponding subspace is determined to be a critical state subspace;

[0025] c. For the identified critical state subspace, maintain its current number of open windows and window openings unchanged to prevent increased resource consumption; mark the window opening strategy of the critical state subspace as stable and record it in the strategy adjustment table of the cloud database; store the identification results and strategy adjustment decisions in the critical state table of the cloud database in a structured data format.

[0026] In a preferred embodiment, step S6 includes the following contents:

[0027] For the subspace that is not identified as a critical state, retrieve the contribution reduction curve data of its window group, including the number of windows opened and the contribution corresponding to the window opening;

[0028] Defining air quality improvement potential indicators Based on the current contribution and the maximum contribution in the decreasing curve The calculation is as follows: in: For window W j In the subspace M i The maximum contribution that can be achieved within For the current window W j In the subspace M i If the air quality improvement potential index is greater than or equal to the preset index, the window group in the subspace is judged to have significant air quality improvement potential;

[0029] For subspaces with potential for improving air quality, the optimal number and degree of window opening are determined based on the contribution decreasing curve; the specific method is as follows:

[0030] Using the decreasing curve model, input the current number of windows opened and the window opening degree, and calculate the contribution changes under different increments of the number of windows opened and the window opening degree;

[0031] Select the combination of window opening quantity increment and window opening degree increment that maximizes the new contribution gain and meets the system resource limit to obtain the optimized window opening quantity Q opt and window opening D opt ;Q opt =Q current +ΔQ max;D opt =D current +ΔD max ; Where: ΔQ max ΔD is the maximum increment of the number of windows opened without exceeding the resource limit; max The maximum increment of window opening without exceeding the system control range;

[0032] The optimized window opening quantity and window opening degree are applied to the window group of the subspace, and the optimized strategy information is recorded in the strategy adjustment table of the cloud database.

[0033] In a preferred embodiment, step S7 includes the following contents:

[0034] For each subspace, first retrieve all its historical records from the cloud database; then preprocess the retrieved data; subsequently, perform feature engineering to extract feature vectors that can reflect the changing trend of air quality and the impact of window opening strategies; select a predetermined machine learning algorithm, and design the corresponding model architecture, defining the input layer as the extracted feature vector and the output layer as the predicted optimal window combination parameters; divide the preprocessed data set into a training set and a validation set, use the training set to train the model, and optimize the model parameters by minimizing the prediction error; finally, store the trained and optimized model parameters and performance indicators in a structured data format in the machine learning model table of the cloud database.

[0035] In a preferred embodiment, step S8 includes the following contents:

[0036] Based on the machine learning model that has been trained and deployed in the cloud, the real-time environmental parameters and window opening status data of each subspace are used as input to obtain real-time updated data from the cloud database; these data are structured and passed to the model input layer through a unified interface, and the model is called for online prediction to output the current optimal window opening combination for each subspace, including the optimal number of windows opened and the optimal window opening degree; based on the optimization results output by the model and combined with the change rate of the real-time environmental parameters of each subspace, the predicted value of the air quality improvement trend in the short term is calculated, and the formula is as follows: Where: T imp Indicates the predicted value of air quality improvement trend; I pred (j is the predicted optimized air quality contribution of the jth window; I current (j) is the actual air quality contribution of the current j-th window; I max (j) and I min (j) are the maximum and minimum contributions of the jth window in the historical data; w adj(j) is the dynamic adjustment factor output by the model, reflecting the adjustment priority of the jth window in the current environment;

[0037] According to the calculated air quality improvement trend value, the window opening strategy of each subspace is adjusted in real time, and the predicted optimal window opening combination parameters are passed to the window control module in the form of standardized instructions; for the subspace that fails to achieve the expected air quality improvement target after the adjustment, secondary prediction and dynamic adjustment are performed by re-calling the machine learning model and combining it with real-time feedback data; finally, all the results of the optimization and adjustment are output in the form of a data interface for model update and performance iteration.

[0038] A remote control system for linking windows with home environments, comprising: an environment division module, a data acquisition module, a model construction module, a benefit analysis module, a state recognition module, a strategy optimization module, a model training module, and a real-time adjustment module;

[0039] Environment division module: divides the indoor environment into multiple subspaces according to the building layout and functional areas, assigns corresponding window groups to each subspace, and transmits the divided subspace information to the data acquisition module;

[0040] Data acquisition module: deploys multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and stores these data in a structured manner in a unified cloud database. After the collected data is stored in the cloud database, it is passed to the model building module;

[0041] Model building module: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is built for each subspace to quantify the specific contribution of each window to the improvement of air quality;

[0042] Benefit analysis module: The window contribution index output by the air circulation model is input into the marginal benefit analysis module. The contribution decreasing curve under the change of the number and opening degree of each window is generated by fitting historical data, and the marginal improvement effect of different opening combinations is clarified. The results are then transmitted to the state recognition module.

[0043] State recognition module: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste. The recognition result is transmitted to the strategy optimization module;

[0044] Strategy optimization module: maintains the initial window opening strategy for the critical subspaces, and accurately adjusts the number and degree of window opening for the subspaces with potential for air quality improvement based on the decreasing curve to ensure efficient use of resources. The adjusted strategy is fed back to the model training module;

[0045] Model training module: The dynamic window opening strategy and real-time environmental data feedback of each subspace are input into the machine learning model to accumulate and train historical data, optimize the model's prediction ability for the optimal window combination under different environmental conditions, and pass the optimized model to the real-time adjustment module;

[0046] Real-time adjustment module: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

[0047] The technical effects and advantages of a remote control system for linking a window with a home environment and a control method thereof of the present invention are as follows:

[0048] The present invention accurately divides the indoor environment into multiple controllable subspaces, assigns corresponding window groups to each subspace, and combines high-frequency sensors to collect various data in real time, constructs an air circulation model based on fluid dynamics, and quantifies the specific contribution of each window to the improvement of air quality. The marginal benefit analysis is used to generate a contribution decline curve, identify the critical subspace that has reached the expected air quality and the marginal benefit tends to be flat, and maintain its window opening strategy unchanged; at the same time, for the subspace with the potential for air quality improvement, the number and opening of windows are accurately adjusted according to the decline curve to ensure efficient use of resources. Furthermore, through the accumulation and training of historical data by machine learning models, the model's prediction ability for the optimal window combination under different environmental conditions is optimized, and online analysis and prediction are performed based on real-time data, and the window opening combination of each subspace is dynamically adjusted to achieve adaptive and refined multi-window collaborative control optimization. The system scientifically divides the environment, accurately collects and processes data, optimizes efficient strategies, and intelligently controls in real time, significantly improving the accuracy and efficiency of indoor air quality management, reducing resource waste, and improving the comfort and health level of the home environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flowchart of a remote control method for linking a window with a home environment according to the present invention;

[0050] Figure 2 The present invention is a schematic structural diagram of a remote control system for linking a window with a home environment. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Embodiment 1: Figure 1 The present invention provides a remote control method for linking a window with a home environment, comprising:

[0053] S1: Divide the indoor environment into multiple subspaces according to the building layout and functional areas, and assign corresponding window groups to each subspace;

[0054] S2: Deploy multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and store these data in a structured manner in a unified cloud database;

[0055] S3: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is constructed for each subspace to quantify the specific contribution of each window to the improvement of air quality;

[0056] S4: Input the window contribution index output by the air circulation model into the marginal benefit analysis module, and generate the contribution reduction curve under the change of the number and opening degree of each window through historical data fitting, so as to clarify the marginal improvement effect of different opening combinations;

[0057] S5: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste;

[0058] S6: For subspaces in a critical state, the initial window opening strategy is maintained unchanged. At the same time, for subspaces whose air quality improvement potential reaches the set threshold, the number or degree of window opening is increased to the optimized parameter according to the decreasing curve to ensure efficient use of resources;

[0059] S7: Feed the dynamic window opening strategy and real-time environmental data of each subspace into the machine learning model to accumulate and train historical data and optimize the model's ability to predict the optimal window combination under different environmental conditions;

[0060] S8: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

[0061] In smart home systems, multi-window coordinated control to optimize indoor air quality depends on the precise division and management of the indoor environment. In order to achieve efficient air quality improvement, the indoor space needs to be carefully divided according to the building layout and functional areas to ensure that the air circulation and environmental control of each subspace can be carried out independently and collaboratively. Therefore, the indoor environment must first be divided into multiple controllable subspaces and a corresponding window group must be assigned to each subspace to achieve targeted air quality management and optimized control.

[0062] Step S1 includes the following contents:

[0063] First, a detailed floor plan of the building is obtained, including the size, purpose, and specific location and number of windows of each room, to form an initial spatial data set. Then, according to the functional attributes of the room (such as bedroom, living room, kitchen, bathroom, etc.), a spatial partitioning algorithm based on graph theory is applied to divide the entire indoor environment into several subspaces with consistent functions and no interference with each other. Each subspace should have independent air circulation requirements and environmental control standards. Then, for each divided subspace, its corresponding window group is identified to ensure that each group of windows can effectively cover the air circulation area of ​​the corresponding subspace. In order to achieve accurate matching, the spatial relationship between the window position coordinates and the subspace boundary is used to accurately assign windows to the corresponding subspace groups through a geometric matching algorithm (such as a polygon covering algorithm). In this process, it is ensured that each subspace is equipped with at least one window group to meet the basic ventilation needs, and the number and distribution of window groups are dynamically adjusted according to the air quality improvement needs of the subspace. Finally, the partitioning results and window group allocation scheme are stored in a unified database in a structured data format (such as JSON or XML) to ensure that the subsequent steps can seamlessly call and reference these tag information, thereby laying a solid foundation for the further optimization of the multi-window collaborative control system.

[0064] Through the precise division and window group allocation in the above step S1, the detailed zoning of the indoor environment and the rational allocation of window resources are achieved, providing a structured and operational data basis for the subsequent optimization of air quality monitoring and control strategies, ensuring that the system can independently and efficiently manage air quality in different subspaces, thereby improving the comfort and health level of the overall home environment.

[0065] In order to ensure the accuracy of subsequent air circulation model construction and marginal benefit analysis, multiple high-precision sensors must be deployed in each controllable subspace to systematically collect key environmental parameters and store these data in a structured form in a unified cloud database. This process requires not only an efficient data collection and transmission mechanism, but also the integrity and consistency of the data to support accurate analysis and decision optimization in subsequent steps.

[0066] Step S2 includes the following contents:

[0067] Based on each subspace divided in step S1, various sensors are deployed respectively, including PM2.5 sensors, CO2 sensors, VOC sensors, temperature and humidity sensors, noise sensors and window status monitoring devices, to ensure that each sensor covers its corresponding environmental parameters; the sampling frequency of all sensors is set to once per minute to ensure high-frequency real-time monitoring of environmental parameters; the indoor and outdoor air quality parameters (PM2.5 concentration, carbon dioxide concentration, volatile organic compound concentration, temperature, humidity, noise level and window opening status data) collected by each sensor are stored in a cloud database in a unified JSON format through wired (such as Ethernet) or wireless (such as Wi-Fi, Zigbee) communication protocols; during data transmission , the AES-256 encryption algorithm is used to ensure the security of data transmission, and the data is initially filtered and denoised through edge computing devices to reduce the amount of redundant data; in the cloud database, a special table structure is established to store the environmental parameters and window status data of each subspace in order according to timestamps, and an independent index is created for each type of parameter to optimize data query efficiency; in addition, a data redundancy backup mechanism is implemented to ensure high availability and disaster recovery capabilities of data through distributed storage technology, and to ensure data integrity and continuity under different fault conditions; finally, all structured and highly reliable sensor data are seamlessly called and referenced by the air circulation model construction module in step S3 through a standardized interface, ensuring that subsequent analysis is based on accurate and consistent data sources.

[0068] Through the sensor deployment and data collection in step S2, the system realizes high-frequency, real-time and accurate monitoring of key environmental parameters and window status inside and outside each controllable subspace, and stores these data in a structured and encrypted form in a unified cloud database, ensuring data security, consistency and high availability, providing reliable basic data support for subsequent air circulation model construction (step S3) and marginal benefit analysis (step S4), and ensuring the overall optimization effect of the multi-window collaborative control system and the continuous comfort of the home environment.

[0069] This step S3 relies on the high-frequency environmental parameter data collected in step S2 and stored in the cloud database. Through fluid dynamics simulation and parameter fitting technology, an air circulation model for each controllable subspace is constructed, and the specific contribution index of each window to the improvement of air quality in different opening states is quantified, which provides accurate basic data for the marginal benefit analysis in step S4, ensuring that the subsequent window opening strategy optimization can be based on reliable contribution evaluation.

[0070] Step S3 includes the following contents:

[0071] Based on the structured data set stored in the cloud database in step S2, including PM2.5 concentration, carbon dioxide concentration, volatile organic compound concentration, temperature, humidity, noise level and window opening status collected at different time points in each controllable subspace; for each subspace divided in step S1, all corresponding historical time series data in the database are retrieved and arranged in chronological order to form a parameter time series matrix; a fluid dynamics simulation method is used, combined with the building layout, window position and size, indoor and outdoor air pressure difference and measured air flow velocity field of each subspace as boundary conditions, PM2.5, carbon dioxide and volatile organic compound concentrations are used as input, and the indoor pollutant concentration distribution under different window opening combinations is simulated by numerically solving the Navier-Stokes equations and diffusion equations; the pollutant equilibrium concentration under the baseline state is defined, that is, in all When the window is closed, the predicted equilibrium pollutant concentration is obtained by setting the window opening state parameter to closed and running the simulation; for each window in each subspace, the target window is opened separately and other windows are kept closed, and the simulation is repeated to obtain the predicted equilibrium pollutant concentration when only the target window is opened; based on the difference between the baseline concentration and the single window opening concentration, the air quality improvement contribution index of each window in the corresponding subspace is calculated, and the contribution index represents the relative proportion of the reduction in pollutant concentration caused by window opening; the contribution index of all windows in each subspace is inserted into the model result table of the cloud database in the form of structured records, and the record content includes the subspace identifier, window identifier, baseline concentration, single window opening concentration and contribution index, so as to ensure that the marginal benefit analysis module in step S4 can accurately retrieve the contribution index of each window through the database, and fit the marginal improvement curve accordingly.

[0072] Through the processing of step S3, the system successfully constructed the air circulation model of each subspace and quantified the air quality improvement contribution index of each window. These structured and accurate contribution data are stored in the cloud database, providing the necessary quantitative basis for the next marginal benefit analysis (step S4), ensuring that the optimization of the multi-window collaborative control strategy can be based on scientific and reliable data support, further improving the system's air quality management efficiency and the comfort of the home environment.

[0073] In order to further optimize the window opening strategy, it is necessary to analyze the marginal benefits of different window opening numbers and opening degrees on the overall air quality improvement. Existing methods mostly rely on linear regression or fixed models, which makes it difficult to capture the complex nonlinear relationship between window opening combinations and air quality improvement. Therefore, it is necessary to use advanced data fitting technology to generate a contribution reduction curve based on historical data for each window opening number and opening degree change, so as to clarify the marginal improvement effect of different opening combinations, thereby optimizing the control strategy, avoiding resource waste and improving air quality management efficiency.

[0074] Based on the window contribution index quantified in step S3, this step S4 generates the contribution reduction curve of each window at different opening numbers and opening degrees through the polynomial regression analysis method, and clarifies the marginal effect of window opening combination on air quality improvement. This provides a scientific basis for the critical state identification in step S5, enabling the system to accurately adjust the window opening strategy based on detailed marginal benefit analysis, avoid resource waste and optimize air quality management.

[0075] Step S4 includes the following contents:

[0076] Based on the air circulation model contribution index stored in the cloud database in step S3, for each subspace, the contribution index of all windows is retrieved, and the corresponding historical time series data of the number of windows opened and the window opening are collected; the number of windows opened and the window opening are used as independent variables, and the contribution index is used as the dependent variable. The polynomial regression analysis method is used to establish a nonlinear relationship model between the contribution of windows in each subspace and the number of windows opened and the window opening; the specific formula is as follows:

[0077]

[0078] in:

[0079] is the subspace M i Inner Window W j Air quality improvement contribution index;

[0080] is the subspace M i The number of windows opened at time t;

[0081] is the subspace M i The average opening degree of the window at time t (0 means closed, 1 means fully open);

[0082] and is the regression coefficient, which reflects the specific impact of the number and opening of windows on the contribution.

[0083] By fitting the polynomial regression model, a contribution reduction curve of each window in each subspace is generated to describe the change trend of the contribution index when the number of windows opened and the window opening increase. These reduction curve data are structured and recorded, including subspace identification, window identification, number of windows opened, window opening and corresponding contribution index, and stored in the marginal benefit analysis table of the cloud database to ensure that the critical state identification module in step S5 can accurately call these curve data to perform dynamic evaluation and strategy adjustment of the subspace air quality improvement status.

[0084] Through the marginal benefit analysis and diminishing curve generation in step S4, the system successfully established a contribution change model for each window in each subspace under different opening numbers and opening degrees, and stored these diminishing curve data in a structured manner in the cloud database. These precise marginal benefit data provide a reliable quantitative basis for critical state identification and window opening strategy adjustment in the subsequent step (step S5), ensuring that the multi-window collaborative control strategy can be based on scientific and reliable data support, dynamically optimize the window opening combination, maximize the air quality improvement effect and minimize resource waste, and further improve the overall comfort and health level of the home environment.

[0085] Through the window contribution decreasing curve generated in step S4, we can understand in detail the marginal effect of different window opening numbers and opening degrees on the improvement of air quality. However, the system needs to further identify which subspaces have achieved the expected air quality improvement effect and whose marginal benefits tend to be flat, so as to avoid continuing to waste resources in these subspaces. The existing technology often lacks a systematic identification method, and it is difficult to efficiently adjust the window opening strategy in a dynamic environment. Therefore, there is an urgent need for an accurate identification and strategy adjustment method based on the contribution decreasing curve to ensure efficient use of resources and continuous comfort of the indoor environment.

[0086] Based on the window contribution reduction curve generated in step S4, this step S5 uses the curvature analysis method to accurately identify the critical state subspaces where the expected air quality has been achieved and the marginal benefit has tended to be flat. By maintaining the window opening strategy of these subspaces, the system effectively avoids excessive consumption of resources and excessive fluctuations in environmental parameters, providing a clear goal and basis for the optimal allocation of resources and adjustment of window opening strategies in the subsequent step S6.

[0087] Step S5 includes the following contents:

[0088] Based on the marginal benefit analysis table stored in the cloud database in step S4, including subspaces, window identifiers, number of windows opened, window openings and corresponding contributions, for each subspace, the following logical steps are performed:

[0089] a. Analysis of contribution decreasing curve:

[0090] The contribution reduction curve data of all windows in the subspace are retrieved from the cloud database, including the contribution under different window opening numbers and window opening degrees.

[0091] Perform curvature analysis on each decreasing curve and calculate the second-order derivative for each combination of window opening quantity and window opening degree to evaluate the smoothness of marginal benefit changes. The specific formula is as follows: and

[0092] in, represents the curvature of the contribution index of the number of windows opened Q, It represents the curvature of the contribution index of window opening D. When the second-order derivative is close to zero, it means that the marginal benefit tends to be flat.

[0093] b. Critical state subspace identification:

[0094] For each subspace, the second-order derivative values ​​of all windows are integrated. If the absolute values ​​of the second-order derivatives of all windows are lower than the preset corresponding threshold θ (i.e. and ), then determine the corresponding subspace M i is the critical state subspace.

[0095] Among them, the threshold θ is set according to historical data and actual environmental requirements to ensure the accuracy of recognition and the response sensitivity of the system.

[0096] c. Strategy adjustment decision:

[0097] For the identified critical state subspace, the current number of open windows and window opening degree are maintained unchanged to prevent further increase in resource consumption.

[0098] The window opening strategy of the critical state subspace is marked as stable and recorded in the strategy adjustment table of the cloud database.

[0099] The identification results and strategy adjustment decisions are stored in a critical status table of a cloud database in a structured data format (such as JSON or XML).

[0100] Through the critical state subspace identification and strategy adjustment in step S5, the system can accurately identify the subspace that has achieved the expected air quality and has a flat marginal benefit, and maintain its window opening strategy accordingly, avoiding resource waste and environmental instability. This process achieves refined control of subspace air quality management through rigorous curvature analysis and threshold setting, laying a solid foundation for subsequent resource optimization and further window opening strategy adjustment, ensuring the efficiency and intelligence level of the entire multi-window collaborative control system, and significantly improving the overall comfort and health level of the home environment.

[0101] Step S5 successfully identifies the critical state subspace that has achieved the expected air quality and where the marginal benefit tends to be flat by analyzing the window contribution decreasing curve. In order to further optimize the utilization of system resources, it is necessary to implement differentiated window opening strategies for subspaces in different states in step S6: maintain the existing window opening strategy for the critical state subspace unchanged, and at the same time, for subspaces with air quality improvement potential, accurately adjust the number of windows opened or the opening degree to the optimized parameters according to the decreasing curve to ensure efficient utilization of resources and continuous improvement of air quality. The existing technology lacks a targeted and dynamic strategy adjustment method, making it difficult to achieve optimal allocation of resources under changing environmental conditions. Therefore, there is an urgent need for a precise strategy adjustment method based on the contribution decreasing curve to achieve reasonable allocation of resources and dynamic optimization of air quality.

[0102] This step S6 relies on the critical state subspace and improvement potential subspace identified by the contribution decreasing curve in step S5, and adopts precise strategy maintenance and optimization methods to implement differentiated window opening strategy adjustments for different subspaces. By maintaining the existing strategy of the critical state subspace, resource waste is avoided; at the same time, by optimizing the number and degree of window openings in the subspace with improvement potential according to the decreasing curve, efficient resource utilization and continuous improvement of air quality are ensured. This process provides data support for the optimization strategy for the machine learning model training in the subsequent step S7, further improving the intelligent control capability of the system.

[0103] Step S6 includes the following contents:

[0104] a. Critical state subspace strategy maintenance:

[0105] All subspace identifiers defined as critical state subspaces and their corresponding window groups, current window opening numbers and window opening degrees are retrieved from the cloud database.

[0106] For each critical state subspace, the number and degree of window opening of its window group are kept unchanged to ensure that the air quality is maintained at the expected level while avoiding additional consumption of resources.

[0107] The information on maintaining the strategy is recorded in the strategy adjustment table of the cloud database to ensure the traceability of the strategy adjustment process.

[0108] b. Improve potential subspace identification and evaluation:

[0109] For the subspace that is not identified as a critical state, the contribution reduction curve data of its window group is retrieved, including the contribution corresponding to the number of open windows and the window opening degree.

[0110] Defining air quality improvement potential indicators Based on the current contribution and the maximum contribution in the decreasing curve The calculation is as follows:

[0111] in:

[0112] For window W j In the subspace M i The maximum contribution that can be achieved within

[0113] For the current window W j In the subspace M i The actual contribution within

[0114] If the air quality improvement potential index is greater than or equal to the preset index, it is determined that the window group of the subspace has significant air quality improvement potential.

[0115] c. Window opening strategy optimization:

[0116] For the subspace with potential for improving air quality, the optimal number of windows to be opened and the window opening degree are determined according to the contribution decreasing curve generated in step S4. The specific method is as follows:

[0117] Using the decreasing curve model, the current number of windows opened and the window opening degree are input, and the contribution changes under different increments of the number of windows opened and the window opening degree are calculated.

[0118] Select the combination of window opening quantity increment and window opening degree increment that maximizes the new contribution gain and meets the system resource limit to obtain the optimized window opening quantity Q opt and window opening D opt .

[0119] Q opt =Q current +ΔQ max ;

[0120] D opt =D current +ΔD max ;

[0121] in:

[0122] ΔQ max The maximum increment of the number of windows that can be opened without exceeding the resource limit;

[0123] ΔD max It is the maximum increment of window opening without exceeding the control range of the system.

[0124] d. Strategy implementation and recording:

[0125] Apply the optimized number and degree of window opening to the window groups in the subspaces to ensure continuous improvement in air quality and efficient use of resources.

[0126] The optimized strategy information is recorded in the strategy adjustment table of the cloud database to ensure the traceability of the strategy adjustment process and the consistency of the data.

[0127] All decision information for strategy maintenance and optimization is stored in the strategy adjustment table of the cloud database in a structured data format (such as JSON or XML) to ensure that the machine learning model in step S7 can seamlessly call these strategy adjustment data for further historical data accumulation and model training.

[0128] Through the critical state subspace strategy maintenance and improvement potential subspace strategy optimization in step S6, the system achieves accurate response to the air quality management needs of different subspaces. The window opening strategy of the critical state subspace is stably maintained to prevent unnecessary consumption of resources; and the subspace with improvement potential is optimized and adjusted based on the contribution decreasing curve to ensure the scientificity and efficiency of the window opening strategy. All strategy adjustment results are stored in the cloud database in the form of structured data, providing a solid data foundation for subsequent machine learning model training and dynamic control strategy optimization, further improving the overall efficiency and intelligence level of the multi-window collaborative control system, and ensuring the continuous comfort and health of the home environment.

[0129] This step S7 is based on the window opening strategy adjustment implemented for different subspaces in step S6. By collecting and preprocessing dynamic window opening strategies and real-time environmental data, and using advanced machine learning algorithms for model training and optimization, an intelligent model that can accurately predict the optimal window combination is constructed. This process not only improves the system's ability to adapt to complex environmental changes, but also provides strong data support and model foundation for real-time online analysis and prediction in step S8.

[0130] Step S7 includes the following contents:

[0131] Based on the dynamic window opening strategy data and real-time environmental parameter data stored in the cloud database in step S6, for each subspace, first retrieve all its historical records from the cloud database; then preprocess the retrieved data, including cleaning, removing outliers and missing data to ensure the integrity and accuracy of the data; then, standardize the environmental parameters and window opening strategy data to make them within a unified dimension and range for easy model training; then, perform feature engineering to extract feature vectors that can reflect the trend of air quality changes and the impact of window opening strategies, such as derived features such as moving averages and rates of change; in the feature selection stage, use correlation analysis methods to evaluate the correlation between each feature and the optimal window combination, and screen out features with significant predictive power; select a machine learning algorithm suitable for processing time series and multivariate inputs, such as a long short-term memory network or a gradient boosting decision tree, and design a corresponding model architecture, define the input layer as the extracted feature vector, and the output layer as the predicted optimal window combination parameters; The data set is divided into a training set and a validation set, the training set is used to train the model, and the model parameters are optimized by minimizing the prediction error; the model performance is evaluated on the validation set, and indicators such as root mean square error and mean absolute error are used to ensure that the model has stable prediction capabilities under different subspaces and environmental conditions; the generalization ability of the model is further evaluated using cross-validation technology, and the model performance is optimized by adjusting the model hyperparameters (such as learning rate and number of network layers); an integrated learning method, such as random forest or XGBoost, is used to improve the prediction accuracy and robustness of the model; after the model training is completed, the optimized machine learning model is deployed to the cloud server, and integrated with the control system management module through the API interface to realize the automated process of real-time data input and optimal window combination prediction output; finally, the trained and optimized model parameters and performance indicators are stored in the machine learning model table of the cloud database in a structured data format to ensure that the online analysis and prediction module in step S8 can seamlessly call and reference these model information, further improving the intelligent control capability of the system.

[0132] Through the training and optimization of the machine learning model in step S7, the system successfully built an intelligent prediction model for each controllable subspace, which can accurately predict the optimal window combination (number of windows opened and window opening degree) based on real-time environmental parameters and historical window opening strategy data. These optimized prediction models are structured and stored in the cloud database to ensure that the online analysis and dynamic adjustment in step S8 can rely on high-precision prediction results, further improve the intelligence level of the multi-window collaborative control system and the efficiency of air quality management, and ensure the continuous comfort and health of the home environment.

[0133] This step S8 relies on the machine learning model deployed in step S7 and the environmental data collected in real time. Through online analysis and dynamic prediction, it outputs the current optimal window opening combination and makes real-time adjustments based on the predicted value of the air quality improvement trend. At the same time, it ensures that the data recording and instruction transmission in the adjustment process are fully structured, providing accurate historical basis for model performance monitoring and updating.

[0134] Step S8 includes the following contents:

[0135] Based on the machine learning model trained in step S7 and deployed in the cloud, the real-time environmental parameters and window opening status data of each subspace are used as input, and real-time updated data [indoor and outdoor air quality parameters (PM2.5 concentration, CO2 concentration, VOC concentration), temperature, humidity, noise level, and the current number of windows opened and window opening] are obtained from the cloud database; these data are structured and transmitted to the model input layer through a unified interface, and the model is called for online prediction, and the current optimal window opening combination of each subspace is output, including the optimal number of windows opened and the optimal window opening; based on the optimization results output by the model, combined with the change rate of the real-time environmental parameters of each subspace, the predicted value of the air quality improvement trend in the short term is calculated, and the formula is as follows:

[0136] in:

[0137] T imp Indicates the predicted value of air quality improvement trend;

[0138] I pred (j) predict the optimized air quality contribution for the jth window;

[0139] I current (j) is the actual air quality contribution of the current j-th window;

[0140] I max (j) and I min (j) are the maximum and minimum contributions of the jth window in the historical data;

[0141] w adj (j) is the dynamic adjustment factor output by the model, reflecting the adjustment priority of the jth window in the current environment.

[0142] According to the calculated air quality improvement trend value, the window opening strategy of each subspace is adjusted in real time, and the predicted optimal window opening combination parameters (including the optimal number of windows opened and the optimal window opening) are passed to the window control module in standardized instructions to ensure that the adjustment process is consistent with the current environmental changes; after the adjustment is completed, all adjustment instructions, the actual environmental parameters after adjustment and the air quality improvement effect are recorded in the adjustment log table in the cloud database. The structured storage fields include the subspace identifier, the number of windows opened before and after the adjustment, the window opening before and after the adjustment, the real-time environmental parameters and the predicted value of the air quality improvement trend to ensure data consistency and traceability; for the subspace that does not achieve the expected air quality improvement target after the adjustment is completed, the machine learning model is re-called and combined with the real-time feedback data for secondary prediction and dynamic adjustment; finally, all the results of the optimization adjustment are output in the form of a data interface for model update and performance iteration.

[0143] Through the online analysis and real-time adjustment of step S8, the system successfully achieved dynamic optimization of the optimal window combination based on machine learning prediction, ensuring that the window opening strategy of each subspace can adapt to environmental changes in real time and maximize the air quality improvement effect; all adjusted data and operation instructions are stored in a structured form in the cloud database, which lays a reliable foundation for model performance monitoring and further optimization, and further improves the real-time, intelligence and efficiency of the multi-window collaborative control system, ensuring the continuous comfort of the home environment and the dynamic optimization of air quality.

[0144] Embodiment 2: Figure 2 The invention provides a remote control system for linking windows with home environments, including: an environment division module, a data acquisition module, a model construction module, a benefit analysis module, a state recognition module, a strategy optimization module, a model training module, and a real-time adjustment module;

[0145] Environment division module: divides the indoor environment into multiple subspaces according to the building layout and functional areas, assigns corresponding window groups to each subspace, and transmits the divided subspace information to the data acquisition module;

[0146] Data acquisition module: deploys multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and stores these data in a structured manner in a unified cloud database. After the collected data is stored in the cloud database, it is passed to the model building module;

[0147] Model building module: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is built for each subspace to quantify the specific contribution of each window to the improvement of air quality;

[0148] Benefit analysis module: The window contribution index output by the air circulation model is input into the marginal benefit analysis module. The contribution decreasing curve under the change of the number and opening degree of each window is generated by fitting historical data, and the marginal improvement effect of different opening combinations is clarified. The results are then transmitted to the state recognition module.

[0149] State recognition module: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste. The recognition result is transmitted to the strategy optimization module;

[0150] Strategy optimization module: maintains the initial window opening strategy for the critical subspaces, and accurately adjusts the number and degree of window opening for the subspaces with potential for air quality improvement based on the decreasing curve to ensure efficient use of resources. The adjusted strategy is fed back to the model training module;

[0151] Model training module: The dynamic window opening strategy and real-time environmental data feedback of each subspace are input into the machine learning model to accumulate and train historical data, optimize the model's prediction ability for the optimal window combination under different environmental conditions, and pass the optimized model to the real-time adjustment module;

[0152] Real-time adjustment module: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

[0153] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0154] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0155] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0156] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A remote control method for linking a window with a home environment, characterized in that: Includes steps: S1: Divide the indoor environment into multiple subspaces according to the building layout and functional areas, and assign corresponding window groups to each subspace; S2: Deploy multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and store these data in a structured manner in a unified cloud database; S3: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is constructed for each subspace to quantify the specific contribution of each window to the improvement of air quality; S4: Input the window contribution index output by the air circulation model into the marginal benefit analysis module, and generate the contribution reduction curve under the change of the number and opening degree of each window through historical data fitting, so as to clarify the marginal improvement effect of different opening combinations; S5: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste; S6: For subspaces in a critical state, the initial window opening strategy is maintained unchanged. At the same time, for subspaces whose air quality improvement potential reaches the set threshold, the number or degree of window opening is increased to the optimized parameter according to the decreasing curve to ensure efficient use of resources; S7: Feed the dynamic window opening strategy and real-time environmental data of each subspace into the machine learning model to accumulate and train historical data and optimize the model's ability to predict the optimal window combination under different environmental conditions; S8: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

2. A remote control method for linking a window with a home environment according to claim 1, characterized in that: Step S1 includes the following contents: Obtain a detailed floor plan of the building, and apply a graph-theory-based space partitioning algorithm based on the functional attributes of the room to divide the entire indoor environment into several subspaces with consistent functions and without interfering with each other. Then, for each divided subspace, identify its corresponding window group to ensure that each group of windows can effectively cover the air circulation area of ​​the corresponding subspace.

3. A remote control method for linking a window with a home environment according to claim 2, characterized in that: Step S2 includes the following contents: Various sensors are deployed in each subspace. The sampling frequency of all sensors is set, and the data collected by different sensors are structured and stored in the cloud database in a unified JSON format. In the cloud database, a special table structure is established to store the environmental parameters and window status data of each subspace in order according to timestamps, and an independent index is created for each type of parameter.

4. A remote control method for linking a window with a home environment according to claim 3, characterized in that: Step S3 includes the following contents: A fluid dynamics simulation method was adopted, and the architectural layout, window position and size, indoor and outdoor pressure difference and measured air flow velocity field of each subspace were combined as boundary conditions. The concentrations of PM2.5, carbon dioxide and volatile organic compounds were used as inputs. The indoor pollutant concentration distribution under different window opening combinations was simulated by numerically solving the Navier-Stokes equations and the diffusion equation. The equilibrium concentration of pollutants under the baseline state was defined, that is, when all windows were closed, the predicted equilibrium pollutant concentration was obtained by setting the window opening state parameter to closed and running the simulation. For each window in each subspace, the target window was opened separately and the other windows were kept closed, and the simulation was repeated to obtain the predicted equilibrium pollutant concentration when only the target window was opened. Based on the difference between the baseline concentration and the single window opening concentration, the air quality improvement contribution index of each window in the corresponding subspace was calculated. The contribution index represents the relative proportion of the reduction in pollutant concentration caused by window opening. The contribution index of all windows in each subspace was inserted into the model result table of the cloud database in the form of structured records.

5. A remote control method for linking a window with a home environment according to claim 4, characterized in that: Step S4 includes the following contents: For each subspace, the contribution index of all windows is retrieved, and the corresponding historical time series data of the number of windows opened and the window opening degree are collected; the number of windows opened and the window opening degree are taken as independent variables, and the contribution index is taken as the dependent variable. The polynomial regression analysis method is used to establish a nonlinear relationship model between the contribution of windows in each subspace and the number of windows opened and the window opening degree; by fitting the polynomial regression model, the contribution decreasing curve of each window in each subspace is generated to describe the changing trend of the contribution index when the number of windows opened and the window opening degree increase; these decreasing curve data are structured and recorded and stored in the marginal benefit analysis table of the cloud database.

6. A remote control method for linking a window with a home environment according to claim 5, characterized in that: Step S5 includes the following contents: a. Perform curvature analysis on each decreasing curve and calculate the second-order derivative for each combination of window opening quantity and window opening degree to evaluate the smoothness of marginal benefit change; b. For each subspace, the second-order derivative values ​​of all windows are integrated. If the absolute values ​​of the second-order derivatives of all windows are lower than the preset corresponding threshold, the corresponding subspace is determined to be a critical state subspace; c. For the identified critical state subspace, maintain its current number of open windows and window openings unchanged to prevent increased resource consumption; mark the window opening strategy of the critical state subspace as stable and record it in the strategy adjustment table of the cloud database; store the identification results and strategy adjustment decisions in the critical state table of the cloud database in a structured data format.

7. A remote control method for linking a window with a home environment according to claim 6, characterized in that: Step S6 includes the following contents: For the subspace that is not identified as a critical state, retrieve the contribution reduction curve data of its window group, including the number of windows opened and the contribution corresponding to the window opening; Defining air quality improvement potential indicators Based on the current contribution and the maximum contribution in the decreasing curve The calculation is as follows: in: For window W j In the subspace M i The maximum contribution that can be achieved within For the current window W j In the subspace M i If the air quality improvement potential index is greater than or equal to the preset index, the window group in the subspace is judged to have significant air quality improvement potential; For subspaces with potential for improving air quality, the optimal number and degree of window opening are determined based on the contribution decreasing curve; the specific method is as follows: Using the decreasing curve model, input the current number of windows opened and the window opening degree, and calculate the contribution changes under different increments of the number of windows opened and the window opening degree; Select the combination of window opening quantity increment and window opening degree increment that maximizes the new contribution gain and meets the system resource limit to obtain the optimized window opening quantity Q opt and window opening D opt ;Q opt =Q current +ΔQ max ;D opt =D current +ΔD max ; Where: ΔQ max ΔD is the maximum increment of the number of windows opened without exceeding the resource limit; max The maximum increment of window opening without exceeding the system control range; The optimized window opening quantity and window opening degree are applied to the window group of the subspace, and the optimized strategy information is recorded in the strategy adjustment table of the cloud database.

8. A remote control method for linking a window with a home environment according to claim 7, characterized in that: Step S7 includes the following contents: For each subspace, first retrieve all its historical records from the cloud database; then preprocess the retrieved data; subsequently, perform feature engineering to extract feature vectors that can reflect the changing trend of air quality and the impact of window opening strategies; select a predetermined machine learning algorithm, and design the corresponding model architecture, defining the input layer as the extracted feature vector and the output layer as the predicted optimal window combination parameters; divide the preprocessed data set into a training set and a validation set, use the training set to train the model, and optimize the model parameters by minimizing the prediction error; finally, store the trained and optimized model parameters and performance indicators in a structured data format in the machine learning model table of the cloud database.

9. A remote control method for linking a window with a home environment according to claim 8, characterized in that: Step S8 includes the following contents: Based on the machine learning model that has been trained and deployed in the cloud, the real-time environmental parameters and window opening status data of each subspace are used as input to obtain real-time updated data from the cloud database; these data are structured and passed to the model input layer through a unified interface, and the model is called for online prediction to output the current optimal window opening combination for each subspace, including the optimal number of windows opened and the optimal window opening degree; based on the optimization results output by the model and combined with the change rate of the real-time environmental parameters of each subspace, the predicted value of the air quality improvement trend in the short term is calculated, and the formula is as follows: Where: T imp Indicates the predicted value of air quality improvement trend; I pred (j) predicts the optimized air quality contribution of the jth window; I current (j) is the actual air quality contribution of the current j-th window; I max (j) and I min (j) are the maximum and minimum contributions of the jth window in the historical data; w adj (j) is the dynamic adjustment factor output by the model, reflecting the adjustment priority of the jth window in the current environment; According to the calculated air quality improvement trend value, the window opening strategy of each subspace is adjusted in real time, and the predicted optimal window opening combination parameters are passed to the window control module in the form of standardized instructions; for the subspace that fails to achieve the expected air quality improvement target after the adjustment, secondary prediction and dynamic adjustment are performed by re-calling the machine learning model and combining it with real-time feedback data; finally, all the results of the optimization and adjustment are output in the form of a data interface for model update and performance iteration.

10. A remote control system for linking a window with a home environment, used to implement a remote control method for linking a window with a home environment as claimed in any one of claims 1 to 9, characterized in that: include: Environment division module, data collection module, model building module, benefit analysis module, state identification module, strategy optimization module, model training module, real-time adjustment module; Environment division module: divides the indoor environment into multiple subspaces according to the building layout and functional areas, assigns a corresponding window group to each subspace, and transmits the divided subspace information to the data acquisition module; Data acquisition module: deploys multiple sensors in each subspace to collect various data in real time at a preset sampling frequency, and stores these data in a structured manner in a unified cloud database. After the collected data is stored in the cloud database, it is passed to the model building module; Model building module: Based on the structured data in the cloud database, using fluid dynamics simulation and parameter fitting technology, an air circulation model is built for each subspace to quantify the specific contribution of each window to the improvement of air quality; Benefit analysis module: The window contribution index output by the air circulation model is input into the marginal benefit analysis module. The contribution decreasing curve under the change of the number and opening degree of each window is generated by fitting historical data, and the marginal improvement effect of different opening combinations is clarified. The results are then transmitted to the state recognition module. State recognition module: Based on the contribution decreasing curve, identify the subspace that has achieved the expected air quality and where the marginal benefit tends to be flat, define it as the critical state subspace, and adjust its window opening strategy to reduce resource waste. The recognition result is passed to the strategy optimization module; Strategy optimization module: maintains the initial window opening strategy for the subspace in a critical state, and accurately adjusts the number and degree of window opening for the subspace with potential for air quality improvement according to the decreasing curve to ensure efficient use of resources. The adjusted strategy is fed back to the model training module; Model training module: The dynamic window opening strategy and real-time environmental data feedback of each subspace are input into the machine learning model to accumulate and train historical data, optimize the model's prediction ability for the optimal window combination under different environmental conditions, and pass the optimized model to the real-time adjustment module; Real-time adjustment module: Use the trained machine learning model to perform online analysis and prediction on the real-time collected environmental data, dynamically adjust the window opening combination of each subspace, and realize multi-window collaborative control optimization.

Images