Indoor environment coordination control system and method for personalized thermal comfort and building energy saving

Abstract

The invention discloses an indoor environment coordination control system and method for personalized thermal comfort and building energy saving. The indoor environment coordination control system comprises a physiological parameter acquisition module, an overall environment adjustment module and a local thermal comfort adjustment module. The physiological parameter acquisition module uses an infrared camera to obtain the temperature of a human face key area in a non-intrusive manner, and predicts the human body thermal sensation through a personal thermal comfort model in combination with the environment temperature and humidity. And the overall environment adjusting module adopts a fuzzy control algorithm to determine a temperature set point and a wind speed gear of the heating ventilation air conditioner according to the thermal inductance predicted value and the environmental parameters, and indoor overall environment adjustment is implemented. The local thermal comfort adjusting module adjusts the temperature and speed of the air flow at the outlet of the personal thermoelectric comfort device through a gear control algorithm, carries out rapid compensation on the local environment, and adjusts the personal thermal inductance to a comfort range. The cooperation of the heating ventilation air conditioner and the personal thermal comfort device in space, time and target is realized, and the operation energy consumption of the building is remarkably reduced while the personalized thermal comfort is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of intelligent control and energy-saving technology for building environments, specifically to an indoor environment coordination control system and method for personalized thermal comfort and building energy conservation. It is applicable to the coordinated adjustment of building indoor heating, ventilation and air conditioning systems and personal thermal comfort devices, ensuring personalized thermal comfort while effectively reducing building operating energy consumption. Background Technology

[0002] With the acceleration of urbanization, the proportion of building operation energy consumption in total social energy consumption is constantly increasing, with HVAC systems typically accounting for more than 50% of building energy consumption. Existing HVAC systems often employ a "one-size-fits-all" control mode for centralized regulation of the indoor environment. While this approach can maintain overall thermal comfort levels, it has the following shortcomings: Firstly, there are significant spatial differences in the indoor environment regarding occupant distribution, heat load, and airflow organization, making centralized control inadequate to address these spatial differences and individual variations in thermal comfort. Secondly, to meet thermal comfort requirements, HVAC systems are often set at lower cooling setpoints or higher heating setpoints, leading to increased energy consumption.

[0003] Localized environmental regulation targets only occupied areas, maintaining unoccupied spaces in a relatively energy-efficient state, which helps improve individual thermal comfort and reduce building energy consumption. However, existing personal thermal comfort devices for regulating localized environments mostly use on / off control or fixed-level control, lacking real-time sensing capabilities of human thermal sensation. Furthermore, they typically operate independently, failing to effectively coordinate with HVAC control, thus hindering their role in overall energy conservation. Therefore, developing a coordinated control system for HVAC and personal thermal comfort devices that balances personalized thermal comfort with building energy efficiency has significant practical implications and application value. Summary of the Invention

[0004] To address the common problems of high energy consumption due to centralized control in existing building indoor environmental control systems, difficulty in considering individual differences in thermal comfort, and lack of effective coordination between personal thermal comfort devices and HVAC control, this invention proposes a coordinated indoor environmental control system for personalized thermal comfort and building energy conservation. This invention aims to guide HVAC control through non-invasive personal thermal sensation prediction, introduce personal thermoelectric comfort devices to compensate for HVAC control errors, and achieve coordinated control of HVAC and thermal comfort devices on both spatial and temporal scales, thereby reducing building operating energy consumption while ensuring individual thermal comfort.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an indoor environment coordination control system for personalized thermal comfort and building energy conservation, comprising a physiological parameter acquisition module, an overall environment adjustment module, and a local thermal comfort adjustment module. The physiological parameter acquisition module includes a physiological / environmental detection unit and a personal thermal sensation prediction unit; the overall environment adjustment module includes a fuzzy control unit and an infrared communication unit; and the local thermal comfort adjustment module includes a personal thermoelectric comfort device and a coordination control unit.

[0006] The system includes the following operational steps: First, in the physiological parameter acquisition module, the infrared camera in the physiological / environmental detection unit non-invasively acquires the temperature of key areas of the human face, and the temperature and humidity sensor in the physiological / environmental detection unit samples the temperature and humidity of the surrounding environment. The personal thermal comfort model, trained offline in the personal thermal sensation prediction unit, predicts the human body's thermal sensation. Second, in the overall environmental adjustment module, based on the predicted human body thermal sensation value and the sampled environmental parameters, the fuzzy control unit uses a fuzzy control algorithm to obtain a suitable air conditioning temperature setpoint and fan speed level, and sends this information to the commercial HVAC system via the infrared communication unit to implement overall indoor environmental adjustment. Simultaneously, in the local thermal comfort adjustment module, the coordination control unit uses a speed control algorithm to adjust the temperature and speed of the airflow at the outlet of the personal thermoelectric comfort device, and through local environmental compensation, adjusts the individual's thermal sensation to a comfortable range, ensuring individual thermal comfort while reducing building operating energy consumption.

[0007] Furthermore, non-invasive temperature acquisition of key areas of the human face includes the following steps:

[0008] A1. Set the installation position of the infrared camera so that it faces the person's face; acquire infrared thermal imaging images of the human face, and select multiple temperature acquisition areas in the facial region, including the cheek area and the nose area.

[0009] A2. Infrared Image Preprocessing: The acquired facial infrared thermal imaging images are denoised, calibrated, and normalized to improve image quality and the stability of temperature data.

[0010] A3. Training of Facial Key Region Detection Model: The preprocessed infrared facial image is manually annotated to mark the location information of the cheek and nose regions; a facial key feature region extraction model is trained based on the annotated data so that it can automatically identify and locate the cheek and nose regions in the facial infrared image.

[0011] A4. Temperature Extraction of Key Areas: Using a trained facial key feature region extraction model, the real-time acquired infrared facial images are detected, the cheek and nose areas are automatically located, and the corresponding temperature feature parameters are extracted from the areas.

[0012] Furthermore, the offline training of the personal thermal comfort model includes the following steps:

[0013] B1. Invite 5-10 participants to establish a thermal comfort dataset. Obtain the temperature of each participant's facial cheeks and nose area according to steps A1-A4. Adjust the indoor air conditioning temperature to allow participants to experience five different thermal sensations: cold, slightly cold, moderate, warm, and hot. After each temperature adjustment, participants remained seated for 15-20 minutes to allow their body's thermal response to stabilize again.

[0014] B2. Synchronously collect the corresponding indoor ambient temperature and relative humidity parameters.

[0015] B3. A questionnaire was used to correlate the temperature of the cheek and nose areas, environmental parameters, and individual subjective thermal sensation levels. The questionnaire was completed simultaneously with the facial temperature data collection described in step B1. The questionnaire included the participant's name, gender, height, weight, and age, as well as their actual thermal sensation under the current environmental parameters and their corresponding thermal sensation level vote. The thermal sensation and their corresponding thermal sensation level votes were: cold (-2), slightly cold (-1), moderate (0), slightly warm (+1), and warm (+2). Based on the questionnaire results, a correlation was established between different facial area temperatures and environmental parameters and individual thermal sensation levels.

[0016] B4. Outliers in the temperature data of the cheek and nose areas are removed using the quartile method. The processed temperature data of these areas is then fused with the corresponding environmental parameters to form a personal thermal feature vector, which is used for offline training of the personal thermal comfort model.

[0017] B5. Establish a data-driven personal thermal comfort model. The model can be trained offline using a convolutional neural network. Its input is the personal thermal sensation feature vector described in step B4, and its output is the thermal sensation level voting value described in step B3. The data is divided into training, validation, and test sets. The model is trained using the K-fold cross-validation method, and the key parameters of the model are adjusted based on the test results, including the convolution kernel size, activation function type, and regularization method and parameters.

[0018] Furthermore, the real-time temperature of the cheek and nose areas, as well as the ambient temperature and humidity parameters, are input into the trained personal thermal comfort model, and the output is the human body thermal sensation prediction result.

[0019] Furthermore, in the overall environmental control module, the fuzzy control algorithm used by the fuzzy control unit includes the following steps:

[0020] C1. Design two decoupled multi-input single-output fuzzy controllers, namely a temperature controller and a fan speed controller, to adjust the HVAC temperature setpoint and fan speed level, respectively.

[0021] C2. Based on the deviation between the predicted human body thermal sensation value and the preset target value, calculate the output of the temperature controller and the wind speed controller using fuzzy inference and rule base.

[0022] C3. The calculated temperature setpoint and fan speed setting are sent to the commercial HVAC system via an infrared communication unit to implement closed-loop control of the overall indoor environment, enabling individual thermal tracking of preset target values.

[0023] Furthermore, the temperature controller in step C1 has the following inputs: thermal prediction value, deviation between the thermal preset target value and the predicted value, and ambient temperature sampling value; and output is the air conditioning temperature setpoint. The fan speed controller has the following inputs: thermal prediction value, deviation between the thermal preset target value and the predicted value, and ambient temperature change rate; and output is the air conditioning fan speed setting.

[0024] Furthermore, the inputs to the temperature controller and wind speed controller are fuzzified, using a triangular membership function to convert the input variables into fuzzy sets to represent the uncertainty of the input. For each controller, several fuzzy control rules are defined. These rules are determined based on thermal comfort principles and a trial-and-error method, and their logical form is as follows: , is used to describe the relationship between input and output variables.

[0025] Furthermore, the preset target value for thermal sensing in step C2 is set to +1 (cooling scenario) or -1 (heating scenario).

[0026] Furthermore, the fuzzy inference described in step C2 adopts Mamdani-type inference to perform inference calculations on the fuzzified input variables to obtain fuzzy output sets of temperature control quantities and wind speed control quantities; subsequently, the centroid method is used to defuzzify the fuzzy outputs.

[0027] Furthermore, the infrared communication unit described in step C3 sends the temperature setpoint and fan speed setting obtained by the fuzzy control unit to the HVAC system to achieve air conditioning according to the conventional infrared transceiver frequency and protocol.

[0028] Furthermore, considering the slow response characteristics of indoor air, the control cycle of the fuzzy control algorithm is 10-15 minutes.

[0029] Furthermore, in the localized thermal comfort adjustment module, the personal thermoelectric comfort device includes an electrothermal conversion component, an airflow drive component, and a flexible airflow distribution network for guiding the regulated airflow to sensitive local areas of the body. The thermoelectric comfort device cools / heats the outlet airflow of the airflow drive component through electrothermal conversion, and then uses the flexible airflow distribution network to deliver the cold / hot airflow to the localized thermally sensitive areas of the body, altering the local temperature distribution and airflow field. By controlling the input voltage of the electrothermal conversion component and the airflow drive component, the temperature and velocity of the airflow at the network outlet can be changed, thereby adjusting the body's thermal sensation to a comfortable range.

[0030] Furthermore, in the local thermal comfort adjustment module, the gear control algorithm used by the coordination control unit includes the following steps:

[0031] D1. Divide the input voltage of the electrothermal conversion component into N levels and the input voltage of the blower into M levels, for a total of N×M level combinations, with combination numbers 1, 2, 3, ..., N×M.

[0032] D2. Establish a personal thermal compensation model based on local environmental adjustment through offline training. The model's inputs are the ambient temperature, relative humidity, and gear combination number around the human body, and the model's output is the human body thermal compensation value after local adjustment (referred to as the local human body thermal compensation value).

[0033] D3. Based on the current air conditioning temperature setpoint and fan speed setting, the ambient temperature and relative humidity around the human body are acquired in real time. Using the model obtained in step D2, the local human body thermal compensation value corresponding to all settings under the current environmental parameters is predicted. The setting combination that is closest to the thermal expectation value is selected to adjust the input voltage of the electrothermal conversion component and the airflow drive component.

[0034] Furthermore, the offline training of the personal thermal compensation model includes the following steps:

[0035] E1. Invite 5-10 participants to participate in an offline data acquisition experiment. During the experiment, participants will wear thermoelectric comfort devices.

[0036] E2. Collect the ambient temperature and relative humidity of the subject's surroundings. Based on the current environmental parameters, adjust the input voltage of the electrothermal conversion component and the airflow drive component in sequence according to the gear combination number. After each adjustment, the subject should remain seated for 3-5 minutes to allow their body thermal response to reach a steady state.

[0037] E3. Using a questionnaire similar to that in step B3, record the voting values ​​for the actual human body thermal sensation level under each gear combination and corresponding environmental parameters.

[0038] E4. Adjust the indoor ambient temperature and relative humidity, repeat steps E2-E3, obtain the correspondence between different environmental parameters and settings and the compensated individual thermal perception level, and construct an offline training dataset.

[0039] E5. Establish a personal thermal compensation model based on local environment regulation. Divide the dataset into training and testing sets, and train the model offline using a convolutional neural network. Adjust the key parameters of the model based on the test results, including kernel size, activation function type, and regularization methods and parameters.

[0040] Furthermore, the thermal sensation expectation value mentioned in step D3 is set to 0.

[0041] Furthermore, considering the fast response characteristics of the personal thermoelectric comfort device, the control cycle of the gear control algorithm is 2-3 minutes.

[0042] Beneficial effects:

[0043] 1. Significantly reducing building energy consumption while ensuring individual thermal comfort. This invention allows HVAC systems to operate within a more relaxed and energy-efficient temperature setting range (e.g., setting the overall room temperature at 28°C instead of 26°C or below in summer), thereby reducing system energy consumption. For individuals experiencing discomfort under these settings, the system rapidly adjusts their local environment through personal thermoelectric comfort devices. This "overall energy-saving setting, local compensation adjustment" approach, combined with slow-fast matching of control cycles, fully demonstrates the coordinated control of building indoor HVAC and personal thermal comfort devices across spatial, temporal, and objective dimensions, achieving a balance between energy conservation and comfort.

[0044] 2. Improved fairness and personalized service capabilities in indoor thermal environment control. Addressing the issue of uneven heating and cooling caused by factors such as distance from air conditioning vents, differences in sunlight exposure, and subjective feelings, the system can provide differentiated compensation by allocating appropriate thermoelectric comfort devices to users in different locations, ensuring that users in different locations within the same space can achieve a relatively balanced state of thermal comfort.

[0045] 3. Based on infrared thermal imaging technology and deep learning theory, it achieves non-contact, real-time, and accurate prediction of individual thermal sensation. No sensors need to be worn by the user, demonstrating truly "intrusive" thermal comfort perception, and it is easy to deploy long-term. Attached Figure Description

[0046] Figure 1 This invention proposes an indoor environment coordination and control system architecture that is geared towards personalized thermal comfort and building energy conservation.

[0047] Figure 2This is an example of the integrated packaging of the coordination and control unit in the physiological parameter acquisition module, the overall environment adjustment module, and the local thermal comfort adjustment module.

[0048] Figure 3 This is an example development board for deploying personal thermal comfort models, personal thermal compensation models, fuzzy control algorithms, and gear control algorithms. This development board is packaged in... Figure 2 Inside the 3D printing control box.

[0049] Figure 4 This is an example of the wearing effect of a personal thermoelectric comfort device.

[0050] Figure 2 In the middle, 1: MAG-RT384 infrared camera; 2: YT2000 gimbal; 3: 5.5-inch MIPI DSI touch screen; 4: antenna; 5: 3D printing control box; 6: infrared camera power interface; 7: infrared communication unit; 8: infrared camera network port; 9: DHT22 temperature and humidity sensor.

[0051] Figure 3 In the diagram, 10: Ethernet port; 11: Power interface; 12: RV1126 main control chip.

[0052] Figure 4 In the middle, 13: PT1000 temperature sensor; 14: airflow drive component; 15: electrothermal conversion component; 16: drive circuit; 17: power supply battery; 18: wearable vest with flexible airflow distribution network. Detailed Implementation

[0053] The following specific embodiments further illustrate the indoor environment coordination control system and method for personalized thermal comfort and building energy conservation described in this invention, but the scope of protection of this invention is not limited thereto.

[0054] This invention discloses an indoor environment coordination control system for personalized thermal comfort and building energy conservation, comprising a physiological parameter acquisition module, an overall environment adjustment module, and a local thermal comfort adjustment module. The physiological parameter acquisition module includes a physiological / environmental detection unit and a personal thermal sensation prediction unit; the overall environment adjustment module includes a fuzzy control unit and an infrared communication unit; and the local thermal comfort adjustment module includes a personal thermoelectric comfort device and a coordination control unit. The overall system architecture is as follows: Figure 1 As shown. The system uses the Rockchip RV1126 processor as the main control chip (see...). Figure 3 It completes tasks such as facial infrared thermal image preprocessing and analysis, personal thermal perception prediction, overall environmental fuzzy control, and thermoelectric comfort device gear control.

[0055] The system includes the following operating steps:

[0056] In the physiological parameter acquisition module, the infrared camera in the physiological / environmental detection unit is used to non-invasively acquire the temperature of key areas of the human face, the temperature and humidity sensor in the physiological / environmental detection unit is used to sample the temperature and humidity of the surrounding environment, and the personal thermal comfort model trained offline in the personal thermal sensation prediction unit is used to predict the human thermal sensation.

[0057] In the overall environmental control module, based on the predicted human body thermal sensation value and the sampled environmental parameters, the fuzzy control unit uses a fuzzy control algorithm to obtain the appropriate air conditioning temperature set point and fan speed level, and sends them to the commercial HVAC system through the infrared communication unit to implement the overall indoor environmental control.

[0058] In the local thermal comfort adjustment module, the coordination control unit uses a gear control algorithm to adjust the temperature and speed of the airflow at the outlet of the personal thermoelectric comfort device. Through local environmental compensation, the individual's thermal sensation is adjusted to a comfortable range, ensuring individual thermal comfort while reducing building operating energy consumption.

[0059] The physiological parameter acquisition module includes a physiological / environmental detection unit and a personal thermal prediction unit. The physiological / environmental detection unit uses a MAG-RT384 infrared thermal imaging camera, mounted in a fixed indoor location with a YT2000 pan-tilt unit, ensuring the camera is always facing the person's face. It continuously or periodically acquires facial infrared thermal images, ensuring the facial area remains stably within the camera's field of view. Based on human thermophysiological characteristics, the cheek and nose areas are selected as key facial regions to characterize the body's current thermal state. The acquired infrared thermal images undergo preprocessing, including infrared noise suppression, temperature calibration, and pixel-level normalization, to reduce the impact of environmental interference on regional temperature results and improve the stability and accuracy of temperature measurements. After preprocessing, the YOLO11 target detection algorithm deployed on the RV1126 chip is used to analyze the infrared thermal images, automatically identifying and locating the spatial positions of the cheek and nose areas, avoiding subjective errors caused by manual selection. For the key thermally sensitive areas detected, feature parameters such as average temperature, maximum temperature, and temperature distribution of the area are further extracted as facial temperature characteristics reflecting the human body's thermophysiological state. Simultaneously, indoor ambient temperature and humidity parameters are sampled using a DHT22 temperature and humidity sensor, and time-aligned with the facial temperature characteristics at the corresponding moments.

[0060] The personal thermal sensation prediction unit fuses facial temperature features with environmental temperature and humidity parameters to form a personal thermal sensation feature vector, which is then input into an offline-trained convolutional neural network personal thermal comfort model to output a personal thermal sensation prediction value. The convolutional neural network model is also deployed on the RV1126 chip, and its network structure includes, in sequence: an input layer, a first one-dimensional convolutional layer (containing 64 3x1 convolutional kernels), a first ReLU activation function layer, a second one-dimensional convolutional layer (containing 64 3x1 convolutional kernels), a second ReLU activation function layer, an SE channel attention enhancement module, an adaptive average pooling layer, a first fully connected layer (with 32 neurons), a third ReLU activation function layer, and an output layer (with 5 neurons, corresponding to 5 categories of thermal sensation voting results). The training parameters are configured as follows: the Adam optimizer is used, the initial learning rate is set to 0.001, the number of training epochs is set to 120, and the batch size is 32; the time series window length of the input data is 5; the loss function is the weighted cross-entropy loss function based on the inverse of class frequency, combined with 5-fold hierarchical cross-validation to ensure the generalization performance of the model under different sample distributions.

[0061] The overall environmental control module includes a fuzzy control unit and an infrared communication unit. The fuzzy control unit employs a fuzzy control algorithm based on individual thermal perception prediction to adapt to the fuzzy and nonlinear characteristics of human thermal perception. To achieve refined adjustment of the overall indoor thermal environment, this embodiment designs two decoupled multi-input single-output fuzzy controllers, used for air conditioning temperature setpoint control and fan speed control respectively, avoiding system oscillations caused by coupled temperature and fan speed adjustments. The temperature controller's inputs include the thermal perception prediction value, the deviation between the preset thermal perception target value and the predicted value, and the ambient temperature; its output is the air conditioning temperature setpoint. The fan speed controller's inputs include the thermal perception prediction value, the deviation between the preset thermal perception target value and the predicted value, and the rate of change of ambient temperature; its output is the air conditioning fan speed setting. From an energy-saving perspective, the preset thermal perception target value is set to +1 (for cooling scenarios) or -1 (for heating scenarios). Considering the slow response characteristics of indoor air, the control cycle of the fuzzy control algorithm is set to 10 minutes.

[0062] Furthermore, taking the wind speed controller as an example, the fuzzy rules are explained. The wind speed controller has a three-input single-output structure. Its input variables include the personal thermal perception prediction value (TSP), the deviation (e) between the preset thermal perception target value and the prediction value, and the rate of change of ambient temperature (dT / dt). The output is the air conditioning fan speed setting. Based on the above variables, the fuzzy rules are derived from recognized thermal comfort theories such as the human body thermal balance model, and are obtained through induction and optimization using a large amount of prior experimental data. A fuzzy control rule table is constructed (as shown in Table 1). The rule table for the temperature controller is constructed using similar logic, and will not be elaborated upon due to space limitations. Its inputs are the thermal perception prediction value, the deviation, and the ambient temperature, and its output is the temperature setpoint adjustment amount. The domain of the thermal perception deviation e is set to [-3, +1]. Here, a negative e indicates that the current environment is relatively cold, and a positive e indicates that the current environment is relatively hot. The thermal perception deviation is divided into five fuzzy subsets: negative large (NB), negative small (NS), zero (ZO), positive small (PS), and positive large (PB). Each fuzzy subset is described using a triangular membership function, and its distribution is as follows: NB corresponds to an interval of approximately [-3, -2], with a membership degree of 1 at -3, and linearly decreasing to 0 at -2; NS corresponds to an interval of approximately [-3, -1], with a membership degree of 1 at -2, and linearly decreasing to both sides; ZO corresponds to an interval of approximately [-2, 0], with a membership degree of 1 at -1, indicating a state close to the target; PS corresponds to an interval of approximately [-1, 1], with a membership degree of 1 at 0, indicating a slightly hot state; PB corresponds to an interval of approximately [0, 1], with a membership degree of 1 at 1, and linearly decreasing to 0.

[0063] Table 1 shows the fuzzy control rules for the wind speed controller:

[0064] ;

[0065] The infrared communication unit consists of four infrared light-emitting diodes (model LF5038), which transmit the temperature setpoint and fan speed level obtained by the fuzzy control unit to the HVAC unit according to the conventional infrared transceiver frequency and protocol.

[0066] The localized thermal comfort adjustment module includes a personal thermoelectric comfort device and a coordination control unit. The personal thermoelectric comfort device includes an electrothermal conversion component, an airflow drive component, a power supply / drive circuit, and a wearable vest with a flexible airflow distribution network, etc. Figure 4As shown. The device uses a XIAO ESP32C3 chip to receive the gear combination sent by the main control chip, and then generates a corresponding pulse width modulation signal through the drive circuit. The electrothermal conversion component receives the pulse width modulation signal output by the XIAO ESP32C3 through an XY-GMOS driver board, and then controls its input voltage to adjust the outlet airflow temperature. The airflow drive component receives the pulse width modulation signal output by the XIAO ESP32C3, and then controls its input voltage to adjust the outlet airflow speed. The wearable vest with a flexible airflow distribution network is made of lightweight and breathable material, and the network is embedded inside the clothing to deliver cold / hot airflow to the heat-sensitive area of ​​the neck.

[0067] The coordination and control unit employs a gear-level control algorithm, dividing the input voltage of the electrothermal conversion component into 5 levels and the input voltage of the airflow drive component into 3 levels (a total of 15 gear-level combinations). Based on the current air conditioning temperature setpoint and fan speed setting, the system acquires real-time temperature and humidity parameters of the human body's surrounding environment. Using a personal thermal comfort compensation model, it derives the local human thermal comfort compensation values ​​corresponding to all gear-level combinations under the current environmental parameters. The gear-level combination closest to the desired thermal comfort value is selected and sent to the XIAOESP32C3 chip, thereby adjusting the input voltage of the electrothermal conversion component and the airflow drive component. The desired thermal comfort value is set to 0, corresponding to a "moderate" thermal comfort level. Considering the fast response characteristics of the personal thermoelectric comfort device, the control cycle of the gear-level control algorithm is set to 2 minutes.

[0068] To enable visualized monitoring of the system status, this embodiment designs a human-machine interface based on the QT framework. This interface uses a 5.5-inch 720×1280 resolution MIPI DSI touchscreen to display real-time facial thermal images, cheek and nose area temperatures, air conditioning temperature setpoints, and the operating status of personal thermoelectric comfort devices (such as device power and outlet airflow temperature).

[0069] To verify the effectiveness of the coordinated control method described in this invention, a comparative experiment was conducted on coordinated control, HVAC fixed setpoint control (26℃), and HVAC single fuzzy control under a typical indoor cooling scenario. The experimental environment was a closed office space with an initial temperature set at 34℃. After one hour of system operation, the outdoor temperatures (average temperature within one hour) for the three control methods were 35.3℃, 34.7℃, and 35.0℃, respectively. Thermal comfort feedback and energy consumption measurements showed that all three control methods could guarantee steady-state thermal comfort for the human body. In terms of energy consumption, the system energy consumption corresponding to fixed setpoint control was approximately 0.80 kWh; the system energy consumption corresponding to single fuzzy control was approximately 0.73 kWh; and the system energy consumption corresponding to coordinated control was reduced to 0.65 kWh. Under the premise of ensuring individual thermal comfort, compared with commonly used fixed setpoint and single fuzzy control, the coordinated control method proposed in this invention can achieve energy savings of approximately 18.8% and 11.0%, respectively.

[0070] In summary, the present invention provides an indoor environment coordination control system and method for personalized thermal comfort and building energy conservation, comprising: (1) the system consists of a physiological parameter acquisition module, an overall environment adjustment module, and a local thermal comfort adjustment module. The physiological parameter acquisition module includes a physiological / environmental detection unit and a personal thermal sensation prediction unit; the overall environment adjustment module includes a fuzzy control unit and an infrared communication unit; and the local thermal comfort adjustment module includes a personal thermoelectric comfort device and a coordination control unit. (2) In the physiological parameter acquisition module, the infrared camera in the physiological / environmental detection unit non-invasively acquires the temperature of key areas of the human face, the temperature and humidity sensor in the physiological / environmental detection unit samples the ambient temperature and humidity, and the personal thermal comfort model obtained offline in the personal thermal sensation prediction unit predicts the human thermal sensation. (3) In the overall environment adjustment module, based on the predicted human thermal sensation value and the sampled environmental parameters, the fuzzy control unit uses a fuzzy control algorithm to obtain a suitable air conditioning temperature setpoint and fan speed level, and sends this information to the commercial HVAC system via the infrared communication unit to implement overall indoor environment adjustment. (4) In the local thermal comfort adjustment module, the coordination control unit uses a gear control algorithm to adjust the temperature and speed of the airflow at the outlet of the personal thermoelectric comfort device. Through local environmental compensation, the individual's thermal sensation is adjusted to a comfortable range, ensuring individual thermal comfort while reducing building operating energy consumption. Compared with the prior art, this invention achieves coordinated control of the building's indoor HVAC and personal thermal comfort devices in terms of spatial scale, temporal scale, and objectives through the method of "overall energy-saving setting and local compensation adjustment" combined with slow-fast matching of the control cycle, ensuring personalized thermal comfort while promoting building energy conservation.

Claims

1. An indoor environment coordination control system for personalized thermal comfort and building energy conservation, characterized in that, include: Physiological parameter acquisition module, overall environmental regulation module, and local thermal comfort regulation module; The physiological parameter acquisition module includes a physiological / environmental detection unit and a personal thermal sensation prediction unit; The overall environmental control module includes a fuzzy control unit and an infrared communication unit; The local thermal comfort adjustment module includes a personal thermoelectric comfort device and a coordination control unit; The system is configured to perform the following steps: In the physiological parameter acquisition module, the infrared camera in the physiological / environmental detection unit is used to non-invasively acquire the temperature of key areas of the human face, the temperature and humidity sensor in the physiological / environmental detection unit is used to sample the temperature and humidity of the surrounding environment, and the personal thermal comfort model trained offline in the personal thermal sensation prediction unit is used to predict the human thermal sensation. In the overall environmental control module, based on the predicted human body thermal sensation value and the sampled environmental parameters, the fuzzy control unit uses a fuzzy control algorithm to obtain the air conditioning temperature set point and fan speed level, and sends them to the HVAC system through the infrared communication unit to implement the overall indoor environmental control. In the local thermal comfort adjustment module, the coordination control unit uses a gear control algorithm to adjust the temperature and speed of the airflow at the outlet of the personal thermoelectric comfort device, and adjusts the personal thermal sensation to a comfortable range through local environmental compensation.

2. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 1, characterized in that, The non-invasive acquisition of temperature in key areas of the human face includes: Infrared thermal images of the human face are captured using an infrared camera. The acquired facial infrared thermal imaging images are preprocessed, including noise reduction, temperature calibration and normalization. Using a trained facial key feature region extraction model, the cheek and nose regions in facial infrared images are automatically identified and located. Extract the corresponding temperature feature parameters from the located cheek and nose areas.

3. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 1, characterized in that, The personal thermal comfort model is obtained through offline training via the following steps: The subjects' facial and nasal temperatures, indoor ambient temperature, and relative humidity were collected under different thermal sensation conditions, and their subjective thermal sensation level voting values ​​were recorded simultaneously to construct a dataset. Outlier handling was performed on the temperature data of the cheek and nose areas in the dataset. The processed regional temperature data is fused with environmental parameters to form a personal thermal feature vector. A data-driven personal thermal comfort model is trained using the individual thermal sensation feature vector and the corresponding subjective thermal sensation level voting value.

4. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 1, characterized in that, The fuzzy control algorithm used by the fuzzy control unit includes: Design a decoupled temperature controller and a fan speed controller, which are used to adjust the temperature setpoint and fan speed level of the HVAC system, respectively. Based on the deviation between the predicted human body thermal sensation value and the preset target value, the output of the temperature controller and the wind speed controller are calculated using fuzzy inference and rule base; The calculated temperature setpoint and fan speed setting are sent to the HVAC system via an infrared communication unit.

5. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 4, characterized in that, The temperature controller's inputs include thermal prediction values, the deviation between the preset thermal target value and the predicted values, and ambient temperature sampling values; its output is the air conditioning temperature setpoint. The fan speed controller's inputs include thermal prediction values, the deviation between the preset thermal target value and the predicted values, and the ambient temperature change rate; its output is the air conditioning fan speed setting.

6. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 1, characterized in that, The personal thermoelectric comfort device includes an electrothermal conversion component, an airflow drive component, and a flexible airflow distribution network for directing the regulated airflow to sensitive local areas of the human body; by controlling the input voltage of the electrothermal conversion component and the airflow drive component, the temperature and velocity of the outlet airflow are changed.

7. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 6, characterized in that, The gear control algorithm used by the coordination control unit includes: dividing the input voltage of the electrothermal conversion component into multiple gears, dividing the input voltage of the airflow drive component into multiple gears, and forming multiple gear combinations. A personal thermal compensation model based on local environmental adjustment is established through offline training. The model's inputs are the ambient temperature, relative humidity, and gear combination number, and its output is the local thermal compensation value. Based on the current air conditioning temperature setpoint and fan speed setting, the ambient temperature and relative humidity are acquired in real time. The personal thermal compensation model is used to predict the local thermal compensation value corresponding to all gear combinations. The gear combination closest to the expected thermal value is selected to adjust the input voltage of the electrothermal conversion component and the airflow drive component.

8. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 7, characterized in that, The personal thermal compensation model is obtained through offline training via the following steps: When subjects wore personal thermoelectric comfort devices, the voting values ​​of the actual human body thermal sensation level corresponding to each combination of gears under different environmental parameters were collected to construct an offline training dataset; The dataset was used to train a personal thermal compensation model based on local environmental regulation.

9. The indoor environment coordination control system for personalized thermal comfort and building energy conservation according to claim 1, characterized in that, The control cycle of the fuzzy control algorithm in the overall environmental adjustment module is longer than the control cycle of the gear control algorithm in the local thermal comfort adjustment module.

10. A method for coordinated indoor environmental control using the system described in any one of claims 1 to 9, characterized in that, include: The physiological parameter acquisition module non-invasively obtains the temperature of key areas of the human face and the temperature and humidity of the surrounding environment, and uses a personal thermal comfort model to predict the human body's thermal sensation. The overall environmental control module uses a fuzzy control algorithm to obtain the air conditioning temperature setpoint and fan speed level based on the predicted human body thermal sensation value and the sampled environmental parameters, and then sends them to the HVAC system to implement overall indoor environmental control. The local thermal comfort adjustment module uses a gear control algorithm to adjust the temperature and speed of the airflow at the outlet of the personal thermoelectric comfort device, and adjusts the personal thermal sensation to a comfortable range through local environmental compensation.

Images