Intelligent whole house temperature control method and system based on AI

Abstract

The invention provides an AI-based intelligent whole-house temperature control method and system, and relates to the technical field of temperature control, and the method comprises the steps: obtaining regional physical data of a plurality of whole-house regional data of whole-house physical data, and generating a reference temperature control instruction for the regional physical data; carrying out regional environment acquisition and analysis on the whole house regional data, determining a temperature rise factor and a temperature drop factor which cause the temperature to deviate from the standard according to the acquired and analyzed data, and carrying out data fusion to obtain multi-factor fusion data; performing positive and negative influence analysis based on an AI model according to the multi-factor fusion data, performing temperature control offset analysis according to the positive and negative influence analysis data, generating regional temperature control adjustment data according to the temperature control offset analysis data, performing temperature control adjustment on the whole house regional data, and obtaining temperature optimization control data. According to the invention, global sensing, change prediction and collaborative optimization intelligent temperature control can be realized, so that the problem of thermal interference between regions can be solved fundamentally.

Description

Technical Field

[0001] This invention proposes an AI-based intelligent whole-house temperature control method and system, which relates to the field of temperature control technology, specifically the field of AI intelligent whole-house temperature control technology. Background Technology

[0002] Traditional whole-house temperature control systems often employ a zoned control strategy based on fixed temperature setpoints, with each zone's temperature control equipment operating independently. However, the internal thermal environment of a house is a dynamic whole. Temperature increases in one area due to factors such as sunlight or the gathering of people, or temperature drops due to factors such as opening windows, can affect adjacent areas through airflow and wall conduction, creating complex cross-zone thermal interference. This thermal coupling effect causes traditional systems to operate in isolation, with uncoordinated actions between devices, resulting not only in large indoor temperature fluctuations and reduced comfort, but also in significant energy waste. Existing technologies lack the ability to collaboratively perceive and intelligently decide on multi-dimensional environmental factors throughout the house, and cannot predict and proactively mitigate these cross-zone thermal effects. Summary of the Invention

[0003] This invention provides an AI-based intelligent whole-house temperature control method and system to solve the above-mentioned problems:

[0004] The present invention proposes an AI-based intelligent whole-house temperature control method and system, wherein the method includes:

[0005] S1. Obtain the regional physical data of multiple whole-house areas from the whole-house physical data, and generate a reference temperature control command from the regional physical data;

[0006] S2. Collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, determine the temperature rise and temperature drop factors that cause the temperature to deviate from the baseline, and perform data fusion to obtain multi-factor fused data. S3. Based on the multi-factor fusion data, perform positive and negative impact analysis using an AI model. Based on the positive and negative impact analysis data, perform temperature control offset analysis. Based on the temperature control offset analysis data, generate regional temperature control adjustment data and perform temperature control adjustment for the entire house area to obtain optimized temperature control data.

[0007] Further, S1 includes:

[0008] Physical data collection is performed on the house to obtain physical data for the entire house;

[0009] Physical features are extracted from the physical data of the whole house to obtain physical feature data of the whole house;

[0010] Based on the physical characteristic data of the whole house, the house is divided into whole house areas to obtain multiple whole house area data;

[0011] The physical data of the whole house is divided into areas based on the whole house area data to obtain the area physical data;

[0012] Regional reference temperature control data is generated based on regional physical data, and reference temperature control commands are generated based on regional reference temperature control data.

[0013] Further, S2 includes:

[0014] The regional data analysis module is triggered according to the reference control command, and the regional data analysis module sends data control signals to the multi-sensor nodes.

[0015] Multiple sensor nodes collect regional environmental information of the whole house area based on data control signals to obtain regional environmental data.

[0016] The multiple sensor nodes transmit the environmental data collected in the area to the area data analysis module;

[0017] The regional data analysis module performs environmental data analysis on the regional environmental data collected to obtain regional environmental analysis data.

[0018] The factors contributing to temperature rise and temperature fall were determined based on the regional environmental analysis data.

[0019] Data fusion instructions are generated for temperature rise and temperature fall factors. The data fusion module is triggered according to the data fusion instructions. The data fusion module performs weighted fusion processing on the factors to obtain multi-factor fused data.

[0020] Furthermore, the regional data analysis module performs environmental data analysis on the collected regional environmental data to obtain regional environmental analysis data, including:

[0021] The regional environmental data is classified into environmental types by the regional data analysis module to obtain regional environmental type data.

[0022] The environmental type data of the region is compared with the preset environmental type range to obtain the environmental type comparison result;

[0023] Based on the comparison results of the environmental types, the regional environmental type data is determined to obtain regional type determination data;

[0024] Based on the regional classification data, obtain abnormal environment type data and normal environment type data;

[0025] The abnormal environment type data and the normal environment type data constitute the regional environmental analysis data.

[0026] Furthermore, the temperature rise and temperature fall factors generate data fusion instructions, which trigger a data fusion module. The data fusion module then performs weighted fusion processing on the factors to obtain multi-factor fused data, including:

[0027] The temperature rise and temperature fall factors are matched with preset influence coefficients to obtain the influence weight of each factor; the data of all factors are weighted and summed according to the influence weight to obtain multi-factor fusion data.

[0028] Furthermore, the multi-factor fusion data is spatiotemporally correlated and integrated to generate regional environmental situation data for AI analysis, including:

[0029] Obtain the physical location information of the sensors corresponding to each factor's data;

[0030] Based on the location information, a temperature field influence diagram for the entire house is constructed;

[0031] In this context, nodes are sensors, and the weights of edges represent the temperature field influence coefficients between the locations of two nodes.

[0032] Based on the temperature field influence relationship diagram, the data diffusion of a single factor is calculated as its influence value on adjacent areas;

[0033] Timestamp the multi-factor fusion data and real-time temperature data, and align them with the historical operating status data of the equipment in the same time period to form a time series data block;

[0034] Normalize the multi-source data after spatial and temporal correlation integration;

[0035] The normalized data is combined into a unified data structure, which includes at least a region identifier, timestamp, fusion factor strength, real-time temperature value, and device status identifier, to generate the regional environmental situation data.

[0036] Further, S3 includes:

[0037] Based on the fusion of multi-factor data and whole-house temperature data, a positive and negative impact analysis based on an AI model is conducted to obtain a predicted temperature change curve.

[0038] Based on the predicted temperature change curve, regional temperature control offset analysis is performed to obtain regional temperature control adjustment data.

[0039] Based on the regional temperature control adjustment data, the regional physical data is controlled and adjusted to obtain the regional baseline temperature control data to obtain temperature optimization control data.

[0040] Furthermore, the step of performing positive and negative impact analysis based on an AI model using multi-factor fusion data and whole-house temperature data to obtain a predicted temperature change curve includes:

[0041] Multi-factor fusion data and historical temperature data are used as inputs to a pre-trained temperature prediction neural network model; the model outputs predicted temperature change curves for each region within a preset future time period.

[0042] Furthermore, the step of performing regional temperature control offset analysis based on the predicted temperature change curve to obtain regional temperature control adjustment data includes:

[0043] Obtain predicted temperature change curves from physical data of adjacent areas;

[0044] Determine the reverse-direction change data of physical data in adjacent areas based on the predicted temperature change curves of physical data in adjacent areas.

[0045] Obtain the offset data of the reverse change data;

[0046] The reverse change data is adjusted based on the offset data to obtain change offset data; the optimal temperature setpoint is determined based on the change offset data; the difference between the optimal temperature setpoint and the current reference temperature control data is the regional temperature control adjustment data.

[0047] Furthermore, the system includes:

[0048] The zone division module is used to acquire zone physical data of multiple whole-house zones from the whole-house physical data, and generate baseline temperature control commands based on the zone physical data.

[0049] The environmental analysis module is used to collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, it determines the factors that cause the temperature to deviate from the baseline, including the factors that cause the temperature to rise and fall. It also performs data fusion to obtain multi-factor fused data. The collaborative control module is used to perform positive and negative impact analysis based on AI models using multi-factor fusion data, perform temperature control offset analysis based on the positive and negative impact analysis data, generate regional temperature control adjustment data based on the temperature control offset analysis data, and perform temperature control adjustment of the whole house area data to obtain temperature optimization control data.

[0050] The beneficial effects of this invention are as follows: This invention enables intelligent temperature control that senses the entire environment, predicts changes, and performs collaborative optimization, fundamentally solving the problem of inter-regional thermal interference and achieving a balance between comfort and energy efficiency. This invention represents a leap from independent control of a single region to whole-house collaborative intelligent control. It transforms the passive response temperature control mode into an active prediction and intervention mode, significantly improving control quality. Through system-level collaborative optimization, it effectively reduces total energy consumption while ensuring overall comfort throughout the house. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of an AI-based intelligent whole-house temperature control method and system. Detailed Implementation

[0052] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0053] In one embodiment of the present invention, the AI-based intelligent whole-house temperature control method and system proposed by the present invention includes:

[0054] S1. Obtain the regional physical data of multiple whole-house areas from the whole-house physical data, and generate a reference temperature control command from the regional physical data;

[0055] S2. Collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, determine the temperature rise and temperature drop factors that cause the temperature to deviate from the baseline, and perform data fusion to obtain multi-factor fused data. S3. Based on the multi-factor fusion data, perform positive and negative impact analysis using an AI model. Based on the positive and negative impact analysis data, perform temperature control offset analysis. Based on the temperature control offset analysis data, generate regional temperature control adjustment data and perform temperature control adjustment for the entire house area to obtain optimized temperature control data.

[0056] The working principle and technical effects of the above solution are as follows: Different temperature control zones are divided based on the physical structure of the building (e.g., area, layout), and a reference temperature command is initialized for each zone. Environmental data (e.g., temperature, humidity, occupancy) is collected in real time from each zone. Factors causing temperature rise (e.g., sunlight, large gatherings) and temperature drop (e.g., open windows, unoccupied status) that lead to temperature deviations from the reference temperature are identified and quantified. This multi-source data is then fused. An AI model is used to perform deep analysis on the fused data to predict future temperature trends. Based on the prediction results, temperature control offset analysis is performed. The system calculates how to coordinate the temperature control devices (e.g., air conditioners, underfloor heating) in each zone to ensure their actions complement each other and offset unnecessary temperature fluctuations. Based on the analysis results, optimized control commands are generated and sent to the actuators in each zone to achieve precise and energy-efficient temperature control. Figure 1 As shown.

[0057] This invention represents a leap from independent control of a single area to whole-house collaborative intelligent control. It transforms the passive, reactive temperature control mode into an active prediction and intervention mode, significantly improving control quality. Through system-level collaborative optimization, it effectively reduces total energy consumption while ensuring overall comfort throughout the house.

[0058] In one embodiment of the present invention, S1 includes:

[0059] Physical data collection is performed on the house to obtain physical data for the entire house;

[0060] Physical features are extracted from the physical data of the whole house to obtain physical feature data of the whole house;

[0061] Based on the physical characteristic data of the whole house, the house is divided into whole house areas to obtain multiple whole house area data;

[0062] The physical data of the whole house is divided into areas based on the whole house area data to obtain the area physical data;

[0063] Regional reference temperature control data is generated based on regional physical data, and reference temperature control commands are generated based on regional reference temperature control data.

[0064] The working principle and technical effects of the above solution are as follows: Physical structural data of the entire building is collected through methods such as LiDAR, architectural drawings, or manual input. Key features for temperature control are extracted from the data, such as room area, floor height, door and window locations, and wall insulation performance. Then, a clustering algorithm is used to automatically group spaces with similar thermodynamic properties (such as sunlight exposure and spatial connectivity) into the same control area. Based on the user-set global comfort temperature or learned historical preferences, the system generates an initial, static temperature setpoint for each zone, i.e., the zone's baseline temperature control data, and issues initial commands to the temperature control equipment in that zone accordingly. This method avoids the arbitrariness of manual zone division, making zone division more scientific and matching the characteristics of the thermal environment. It lays a solid foundation for refined and differentiated control, and improves the automation level of system deployment and user experience.

[0065] In one embodiment of the present invention, S2 includes:

[0066] The regional data analysis module is triggered according to the reference control command, and the regional data analysis module sends data control signals to the multi-sensor nodes.

[0067] Multiple sensor nodes collect regional environmental information of the whole house area based on data control signals to obtain regional environmental data.

[0068] The multiple sensor nodes transmit the environmental data collected in the area to the area data analysis module;

[0069] The regional data analysis module performs environmental data analysis on the regional environmental data collected to obtain regional environmental analysis data.

[0070] The factors contributing to temperature rise and temperature fall were determined based on the regional environmental analysis data.

[0071] Among them, factors that cause temperature rise include solar radiation intensity and the number of people exceeding a threshold, while factors that cause temperature drop include open doors and windows and the absence of people in specific areas;

[0072] Data fusion instructions are generated for temperature rise and temperature fall factors. The data fusion module is triggered according to the data fusion instructions. The data fusion module performs weighted fusion processing on the factors to obtain multi-factor fused data.

[0073] The working principle and technical effects of the above solution are as follows: After the system issues a data collection command, multiple sensor nodes deployed throughout the house (which may include temperature, humidity, light intensity, and human infrared sensors, etc.) are activated to collect real-time environmental information for their respective areas. All data is aggregated into the regional data analysis module. The module analyzes the data, comparing real-time data with preset thresholds (e.g., light intensity > preset light intensity threshold indicates sunshine, number of people > preset number of people threshold indicates large gatherings), thereby clearly identifying specific factors causing temperature rise and fall. The data fusion module is triggered, performing weighted summation on these identified factors and outputting a comprehensive multi-factor fusion data, which quantitatively reflects the overall influence of all environmental factors on temperature. This method achieves qualitative identification and quantitative assessment of complex environmental factors. It integrates scattered multi-dimensional sensor information into a comprehensive indicator with clear physical meaning, simplifying the input of the AI ​​model. It enhances the system's ability to perceive and understand dynamic environmental changes.

[0074] In one embodiment of the present invention, the regional data analysis module performs environmental data analysis on the regional environmental collection data to obtain regional environmental analysis data, including:

[0075] The regional environmental data is classified into environmental types by the regional data analysis module to obtain regional environmental type data.

[0076] The environmental type data of the region is compared with the preset environmental type range to obtain the environmental type comparison result;

[0077] Based on the comparison results of the environmental types, the regional environmental type data is determined to obtain regional type determination data;

[0078] Based on the regional classification data, obtain abnormal environment type data and normal environment type data;

[0079] The abnormal environment type data and the normal environment type data constitute the regional environmental analysis data.

[0080] The working principle and technical effects of the above solution are as follows: Based on the source and physical meaning of the data, environmental types are categorized, such as temperature, humidity, light intensity, and human activity. Each type of data is compared with preset normal ranges (e.g., determining whether the light intensity exceeds the no-sunlight range, or whether human activity signals change from absent to present). Based on the comparison results, the current environmental state is determined, ultimately outputting two types of data: one is abnormal environmental data (i.e., temperature rise / fall factors requiring attention), and the other is normal environmental data (i.e., a stable state requiring no special intervention). This method provides a clear and standardized data processing flow, reducing the complexity of system implementation. It can accurately identify key changing factors that significantly affect temperature control and filter out irrelevant interference. It provides reliable data for generating precise control strategies.

[0081] In one embodiment of the present invention, the temperature rise factor and temperature fall factor generate a data fusion instruction, the data fusion module is triggered according to the data fusion instruction, and the data fusion module performs weighted fusion processing on the factors to obtain multi-factor fused data, including:

[0082] The temperature rise and temperature fall factors are matched with preset influence coefficients to obtain the influence weight of each factor; the data of all factors are weighted and summed according to the influence weight to obtain multi-factor fusion data.

[0083] The working principle and technical effect of the above solution are as follows: An influence coefficient lookup table is preset, assigning a weight to each identified environmental factor. For example, the weight of western sun exposure might be 0.9, while the weight of individual activity might be 0.3. During fusion, the system multiplies the real-time measurement value of each factor by its corresponding influence weight, then sums all the products to obtain multi-factor fused data. This final data is a scalar value; its positive or negative sign represents whether the overall temperature is rising or falling, and its absolute value represents the intensity of the influence. This unifies environmental factors of different properties and dimensions into a comparable and calculable quantitative indicator. The method is simple and efficient, with low computational resource overhead, making it suitable for real-time control. Through weight configuration, the importance of different factors on temperature can be flexibly reflected.

[0084] One embodiment of the present invention integrates multi-factor fusion data through spatiotemporal correlation to generate regional environmental situation data for AI analysis, including:

[0085] Obtain the physical location information of the sensors corresponding to each factor's data;

[0086] Based on the location information, a temperature field influence diagram for the entire house is constructed;

[0087] In this context, nodes are sensors, and the weights of edges represent the temperature field influence coefficients between the locations of two nodes.

[0088] Based on the temperature field influence relationship diagram, the data diffusion of a single factor is calculated as its influence value on adjacent areas;

[0089] Timestamp the multi-factor fusion data and real-time temperature data, and align them with the historical operating status data of the equipment in the same time period to form a time series data block;

[0090] Normalize the multi-source data after spatial and temporal correlation integration;

[0091] The normalized data is combined into a unified data structure, which includes at least a region identifier, timestamp, fusion factor strength, real-time temperature value, and device status identifier, to generate the regional environmental situation data.

[0092] The working principle and technical effects of the above solution are as follows: Spatially, the system constructs a temperature field influence relationship diagram based on the sensor installation location information. This diagram uses sensors as nodes, and the physical distance and obstruction conditions between nodes (such as the presence of doors) are used to calculate the temperature field influence coefficient as the edge weight. Using this diagram, the reading of a sensor can be diffused to its adjacent areas according to the weight ratio, calculating its spatial influence. Temporally, all data is accurately timestamped and aligned with historical operating status data of equipment such as air conditioners within the same time period, forming a regular time-series data block. This data from different spaces and sources is normalized and encapsulated into standardized regional environmental situation data containing information such as region, time, value, and status. This method breaks down data silos and constructs a unified, spatiotemporally synchronized environmental data view for the entire house. This enables the AI ​​model to learn the spatial transfer patterns of heat and the time lag effect of equipment actions. It provides high-quality, structured input data for accurate prediction and collaborative optimization.

[0093] In one embodiment of the present invention, S3 includes:

[0094] Based on the fusion of multi-factor data and whole-house temperature data, a positive and negative impact analysis based on an AI model is conducted to obtain a predicted temperature change curve.

[0095] Based on the predicted temperature change curve, regional temperature control offset analysis is performed to obtain regional temperature control adjustment data.

[0096] Based on the regional temperature control adjustment data, the regional physical data is controlled and adjusted to obtain the regional baseline temperature control data to obtain temperature optimization control data.

[0097] The working principle and technical effects of the above solution are as follows: Using regional environmental situation data as input, a pre-trained AI prediction model is invoked to output predicted temperature change curves for each region over a future period. The system then enters the regional temperature control offset analysis stage. Based on these predicted curves, an optimization algorithm calculates a new, dynamic set of regional temperature setpoints. These new setpoints aim to enable temperature control devices in each region to work collaboratively, mutually offsetting thermal interference between them. Finally, the system sends the calculated new instructions to the devices, completing closed-loop control. This method forms a complete intelligent control chain encompassing situational awareness, future prediction, collaborative decision-making, and precise execution. It significantly improves the system's foresight and overall coordination capabilities, and is a key link in achieving the dual goals of energy saving and comfort optimization.

[0098] In one embodiment of the present invention, the step of performing positive and negative influence analysis based on an AI model using multi-factor fusion data and whole-house temperature data to obtain a predicted temperature change curve includes:

[0099] Multi-factor fusion data and historical temperature data are used as inputs to a pre-trained temperature prediction neural network model; the model outputs predicted temperature change curves for each region within a preset future time period.

[0100] The working principle and technical effects of the above solution are as follows: Historical temperature data sequences and real-time generated multi-factor fusion data (or better regional environmental situation data) are input together into a pre-trained temperature prediction neural network model (e.g., a Long Short-Term Memory network, LSTM). The model has been trained on a large amount of historical data and can capture the temporal patterns of temperature changes and their complex nonlinear relationships with environmental factors. After the model runs, it directly outputs the predicted temperature change value for each region within a preset time period (e.g., the next 30 minutes), usually presented as a continuous curve. This achieves high-precision, advanced prediction of temperature changes, solving the problem of large lag in traditional control. It provides a scientific basis for taking control actions in advance and preventing problems before they occur. This is the core technical support for the entire system to achieve intelligence.

[0101] In one embodiment of the present invention, the step of performing regional temperature control offset analysis based on the predicted temperature change curve to obtain regional temperature control adjustment data includes:

[0102] Obtain predicted temperature change curves from physical data of adjacent areas;

[0103] Determine the reverse-direction change data of physical data in adjacent areas based on the predicted temperature change curves of physical data in adjacent areas.

[0104] Obtain the offset data of the reverse change data;

[0105] The reverse change data is adjusted based on the offset data to obtain change offset data; the optimal temperature setpoint is determined based on the change offset data; the difference between the optimal temperature setpoint and the current reference temperature control data is the regional temperature control adjustment data.

[0106] The working principle and technical effects of the above solution are as follows: A predicted temperature change in one area is considered a virtual control quantity that can be utilized by its neighboring areas. The theoretical adjustment required to offset its own predicted change is calculated, as are the theoretical adjustment required for neighboring areas to offset their own predicted changes. Through optimization algorithms (such as linear programming and Model Predictive Control, MPC), an optimal set of global setpoints is found under the dual constraints of whole-house energy consumption and comfort. This set of setpoints redistributes the adjustment tasks of each area, so that the adjustment effects between areas thermodynamically cancel each other out (for example, using the slight cooling of the living room air conditioner to help offset the predicted temperature rise in the study), thereby avoiding all devices independently and intensively resisting local environmental changes. The difference between the solved optimal setpoint and the initial baseline value is the final regional temperature control adjustment data. This achieves true system-level energy saving, reducing ineffective counteraction and redundant work by devices through thermal synergy. It greatly improves the stability and uniformity of the whole-house thermal environment, avoiding localized overheating and cooling. Upgrading traditional temperature control to heat flow management represents the future development direction of intelligent temperature control.

[0107] According to one embodiment of the present invention, the system includes:

[0108] The zone division module is used to acquire zone physical data of multiple whole-house zones from the whole-house physical data, and generate baseline temperature control commands based on the zone physical data.

[0109] The environmental analysis module is used to collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, it determines the factors that cause the temperature to deviate from the baseline, including the factors that cause the temperature to rise and fall. It also performs data fusion to obtain multi-factor fused data. The collaborative control module is used to perform positive and negative impact analysis based on AI models using multi-factor fusion data, perform temperature control offset analysis based on the positive and negative impact analysis data, generate regional temperature control adjustment data based on the temperature control offset analysis data, and perform temperature control adjustment of the whole house area data to obtain temperature optimization control data.

[0110] The working principle and technical effects of the above solution are as follows: Different temperature control zones are divided based on the physical structure of the building (e.g., area, layout), and a baseline temperature command is initialized for each zone. Environmental data (e.g., temperature, humidity, occupancy) is collected in real time from each zone. Factors causing temperature rise (e.g., sunlight, large gatherings) and temperature drop (e.g., open windows, unoccupied status) that lead to temperature deviations from the baseline are identified and quantified. This multi-source data is then fused. An AI model is used to perform in-depth analysis of the fused data to predict future temperature trends. Based on the prediction results, temperature control offset analysis is performed. The system calculates how to coordinate the temperature control devices (e.g., air conditioners, underfloor heating) in each zone to ensure their actions complement each other and offset unnecessary temperature fluctuations. Optimized control commands are generated based on the analysis results and sent to the actuators in each zone to achieve precise and energy-efficient temperature control.

[0111] This invention represents a leap from independent control of a single area to whole-house collaborative intelligent control. It transforms the passive, reactive temperature control mode into an active prediction and intervention mode, significantly improving control quality. Through system-level collaborative optimization, it effectively reduces total energy consumption while ensuring overall comfort throughout the house.

[0112] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An AI-based intelligent whole-house temperature control method, characterized in that, The method includes: S1. Obtain the regional physical data of multiple whole-house areas from the whole-house physical data, and generate a reference temperature control command from the regional physical data; S2. Collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, determine the temperature rise and temperature drop factors that cause the temperature to deviate from the baseline, and perform data fusion to obtain multi-factor fused data. S3. Based on the multi-factor fusion data, perform positive and negative impact analysis using an AI model. Based on the positive and negative impact analysis data, perform temperature control offset analysis. Based on the temperature control offset analysis data, generate regional temperature control adjustment data and perform temperature control adjustment for the entire house area to obtain optimized temperature control data.

2. The AI-based intelligent whole-house temperature control method according to claim 1, characterized in that, S1 includes: Physical data collection is performed on the house to obtain physical data for the entire house; Physical features are extracted from the physical data of the whole house to obtain physical feature data of the whole house; Based on the physical characteristic data of the whole house, the house is divided into whole house areas to obtain multiple whole house area data; The physical data of the whole house is divided into areas based on the whole house area data to obtain the area physical data; Regional reference temperature control data is generated based on regional physical data, and reference temperature control commands are generated based on regional reference temperature control data.

3. The AI-based intelligent whole-house temperature control method according to claim 1, characterized in that, S2 includes: The regional data analysis module is triggered according to the reference control command, and the regional data analysis module sends data control signals to the multi-sensor nodes. Multiple sensor nodes collect regional environmental information of the whole house area based on data control signals to obtain regional environmental data. The multiple sensor nodes transmit the environmental data collected in the area to the area data analysis module; The regional data analysis module performs environmental data analysis on the regional environmental data collected to obtain regional environmental analysis data. The factors contributing to temperature rise and temperature fall were determined based on the regional environmental analysis data. Data fusion instructions are generated for temperature rise and temperature fall factors. The data fusion module is triggered according to the data fusion instructions. The data fusion module performs weighted fusion processing on the factors to obtain multi-factor fused data.

4. The AI-based intelligent whole-house temperature control method according to claim 3, characterized in that, The regional data analysis module performs environmental data analysis on the collected regional environmental data to obtain regional environmental analysis data, including: The regional environmental data is classified into environmental types by the regional data analysis module to obtain regional environmental type data. The environmental type data of the region is compared with the preset environmental type range to obtain the environmental type comparison result; Based on the comparison results of the environmental types, the regional environmental type data is determined to obtain regional type determination data; Based on the regional classification data, obtain abnormal environment type data and normal environment type data; The abnormal environment type data and the normal environment type data constitute the regional environmental analysis data.

5. The AI-based intelligent whole-house temperature control method according to claim 3, characterized in that, The temperature rise and temperature fall factors generate data fusion instructions. Based on these instructions, a data fusion module is triggered. The data fusion module performs weighted fusion processing on the factors to obtain multi-factor fused data, including: The temperature rise and temperature fall factors are matched with preset influence coefficients to obtain the influence weight of each factor; the data of all factors are weighted and summed according to the influence weight to obtain multi-factor fusion data.

6. The AI-based intelligent whole-house temperature control method according to claim 5, characterized in that, By integrating multi-factor fusion data spatiotemporally and correlated, regional environmental situation data for AI analysis is generated, including: Obtain the physical location information of the sensors corresponding to each factor's data; Based on the location information, a temperature field influence diagram for the entire house is constructed; In this context, nodes are sensors, and the weights of edges represent the temperature field influence coefficients between the locations of two nodes. Based on the temperature field influence relationship diagram, the data diffusion of a single factor is calculated as its influence value on adjacent areas; Timestamp the multi-factor fusion data and real-time temperature data, and align them with the historical operating status data of the equipment in the same time period to form a time series data block; Normalize the multi-source data after spatial and temporal correlation integration; The normalized data is combined into a unified data structure to generate regional environmental situation data.

7. The AI-based intelligent whole-house temperature control method according to claim 1, characterized in that, S3 includes: Based on the fusion of multi-factor data and whole-house temperature data, a positive and negative impact analysis based on an AI model is conducted to obtain a predicted temperature change curve. Based on the predicted temperature change curve, regional temperature control offset analysis is performed to obtain regional temperature control adjustment data. Based on the regional temperature control adjustment data, the regional physical data is controlled and adjusted to obtain the regional baseline temperature control data to obtain temperature optimization control data.

8. The AI-based intelligent whole-house temperature control method according to claim 7, characterized in that, The step involves performing a positive and negative impact analysis based on an AI model using multi-factor fusion data and whole-house temperature data to obtain a predicted temperature change curve, including: Multi-factor fusion data and historical temperature data are used as inputs to a pre-trained temperature prediction neural network model; the model outputs predicted temperature change curves for each region within a preset future time period.

9. The AI-based intelligent whole-house temperature control method according to claim 7, characterized in that, The step of performing regional temperature control offset analysis based on the predicted temperature change curve to obtain regional temperature control adjustment data includes: Obtain predicted temperature change curves from physical data of adjacent areas; Determine the reverse-direction change data of physical data in adjacent areas based on the predicted temperature change curves of physical data in adjacent areas. Obtain the offset data of the reverse change data; The reverse change data is adjusted based on the offset data to obtain change offset data; the optimal temperature setpoint is determined based on the change offset data; the difference between the optimal temperature setpoint and the current reference temperature control data is the regional temperature control adjustment data.

10. An AI-based intelligent whole-house temperature control system, characterized in that: The system includes: The zone division module is used to acquire zone physical data of multiple whole-house zones from the whole-house physical data, and generate baseline temperature control commands based on the zone physical data. The environmental analysis module is used to collect and analyze regional environmental data for the whole house according to the baseline temperature control command. Based on the collected and analyzed data, it determines the factors that cause the temperature to deviate from the baseline, including the factors that cause the temperature to rise and fall. It also performs data fusion to obtain multi-factor fused data. The collaborative control module is used to perform positive and negative impact analysis based on AI models using multi-factor fusion data, perform temperature control offset analysis based on the positive and negative impact analysis data, generate regional temperature control adjustment data based on the temperature control offset analysis data, and perform temperature control adjustment of the whole house area data to obtain temperature optimization control data.

Images