Method and device for controlling comfort of occupant in vehicle, electronic equipment and storage medium

Abstract

This application provides a method, device, electronic device, and storage medium for controlling the comfort of vehicle occupants. The solution includes: determining whether a clothing change event has occurred based on an image sequence of the occupants; determining the current clothing thermal resistance parameter when a clothing change event is confirmed; determining the current predicted thermal sensation based on the current clothing thermal resistance parameter; and adjusting the cabin environment actuator based on the current predicted thermal sensation and the target thermal sensation value. This application predicts heat changes through an algorithm and adjusts the cabin environment actuator in advance when a clothing change event is detected, before a physical change in body temperature occurs. This achieves proactive feedback, completely eliminating the lag in traditional automatic air conditioning adjustments, ensuring the user remains in a constant state of comfort throughout the entire clothing change process, achieving a "zero-wait" temperature control experience and seamless thermal comfort.

Description

Technical Field

[0001] This application relates to the field of automotive technology, and in particular to a method and device for controlling the comfort of occupants in a vehicle, electronic equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] Vehicle air conditioning control systems have evolved from manual to automatic to intelligent systems. Current mainstream intelligent air conditioning systems incorporate the PMV (Predicted Mean Vote) thermal comfort evaluation index. This index comprehensively considers six parameters: air temperature, mean radiant temperature, air velocity, air humidity (four environmental factors), human metabolic rate, and clothing thermal resistance (two human factors). It aims to accurately predict the actual temperature sensation of occupants through mathematical models. Clothing thermal resistance (Clo value) is a highly weighted variable; it is approximately 0.3-0.5 Clo for summer short sleeves and can reach 1.0-1.5 Clo for winter down jackets. At the same in-car temperature, occupants wearing different clothing may experience drastically different sensations. Therefore, accurately acquiring and dynamically updating occupant clothing thermal resistance parameters is crucial for achieving precise in-car thermal comfort control.

[0003] Currently, the closest existing technologies related to occupant clothing recognition in the industry mainly include the following solutions. One type is a temperature control system based on infrared thermal imaging, which determines the hot or cold state by detecting the temperature distribution on the occupant's body surface and initiates adjustment when the body surface temperature is below a threshold. This type of solution is a negative feedback control, which must wait for the body surface temperature to actually drop before responding, resulting in a "cooling window" of several minutes, and is easily affected by environmental heat sources such as direct sunlight and hot drinks. Another type is a static initial setting solution based on visual recognition, which identifies the occupant's clothing and sets the initial air conditioning strategy when the vehicle starts. This type of solution assumes that the occupant's clothing remains constant during the journey and cannot perceive dynamic changes in clothing putting on or taking off during driving.

[0004] In summary, the existing technology has the following main technical problems: First, the response is lagging. The existing negative feedback control mechanism cannot respond the moment when the passenger takes off their clothes, causing the passenger to experience a period of discomfort due to temperature changes. Second, the model is inaccurate. The thermal resistance parameter of clothing is statically set or only identified once at the beginning of the trip. When the passenger takes off their coat during the journey, the PMV model calculation results deviate significantly from the actual feeling. Third, it relies on manual intervention. Due to the above defects, users are forced to be distracted while driving to adjust the air conditioning, which poses a driving safety hazard. Summary of the Invention

[0005] This application provides a method, device, electronic device, and storage medium for controlling the comfort of vehicle occupants, in order to solve the problems of existing intelligent air conditioning adjustment having lag, poor accuracy, and the need for manual user intervention, which poses safety hazards.

[0006] This application provides a method for controlling the comfort of vehicle occupants, including: Based on the image sequence of the occupants inside the vehicle, determine whether a clothing change event has occurred; Once a clothing change event is determined, the thermal resistance parameters of the current clothing are determined. Based on the current clothing thermal resistance parameters, determine the current predicted thermal sensation; The cabin environment actuators are adjusted based on the current predicted thermal sensation and the target thermal sensation value.

[0007] The technical solution provided in this application, before the physical change in body temperature occurs, predicts the heat change through an algorithm when a clothing change event is detected, and adjusts the cabin environment actuator in advance, realizing proactive feedback. This completely eliminates the lag in traditional automatic air conditioning adjustment, keeping the user in a constant and comfortable state throughout the entire process of clothing change, achieving a "zero-wait" temperature control experience and seamless connection of thermal comfort.

[0008] In one embodiment, determining whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle includes: The image sequence of the occupants inside the vehicle is subjected to clothing-wearing and undressing action detection to determine whether clothing-wearing and undressing actions exist. The image sequence of the occupants inside the vehicle is subjected to clothing type change detection to determine whether the clothing type has changed; A clothing change event is determined when there is a simultaneous action of putting on or taking off clothes and a change in the type of clothing.

[0009] The embodiments described above in this application use the action characteristics of putting on and taking off clothes and the characteristics of clothing type changes. Only when both characteristics are satisfied is it determined that a clothing change event has occurred. This two-layer verification mechanism can improve the accuracy of clothing change event detection and avoid false triggering of "tidying clothes / mistaken action".

[0010] In one embodiment, the step of detecting dressing and undressing actions in the image sequence of the vehicle occupants to determine whether dressing and undressing actions exist includes: Skeletal keypoint detection is performed on the image sequence to determine the wrist-neck distance feature, wrist height above shoulder feature, wrist velocity value, and hand involvement feature of each image; An action state machine is constructed based on the wrist-neck distance features, wrist-to-shoulder height features, wrist velocity values, and hand-to-hand involvement features of each image. Based on the action state machine, when the sequence of undressing actions or the sequence of dressing actions are continuously satisfied, it is determined that the dressing or undressing action has occurred.

[0011] The embodiments described above in this application use an action state machine to avoid errors in single-frame detection, effectively distinguish between dressing and undressing actions and interfering actions, have a low false trigger rate, avoid meaningless frequent actions of the air conditioner, and improve user experience.

[0012] In one embodiment, the step of detecting changes in clothing type in the image sequence of the vehicle occupants to determine whether the clothing type has changed includes: The image sequence is identified, and the probability distribution of clothing categories corresponding to images at different times is output. Based on the probability distribution of clothing categories corresponding to images at different times and the equivalent thermal resistance corresponding to each category, the instantaneous clothing thermal resistance values ​​at different times are calculated by weighting. Based on the instantaneous thermal resistance values ​​of clothing at multiple moments within the first time window before the dressing and undressing action occurs, and the instantaneous thermal resistance values ​​of clothing at multiple moments within the second time window after the dressing and undressing action occurs, the thermal resistance before the event and the thermal resistance after the event are determined. When the absolute value of the difference between the thermal resistance before and after the event is greater than a preset threshold, it is determined that the clothing type has changed.

[0013] The above embodiments calculate the thermal resistance value based on the probability distribution of each category and the equivalent thermal resistance of each category using a weighted average. This eliminates the constraint of discrete category boundaries, allowing for a smooth transition between different categories and avoiding abrupt changes in thermal resistance parameters caused by category jumps. When the recognition model is uncertain about the clothing category (e.g., a flat multi-category probability distribution), the weighted average naturally converges to a conservative estimate, avoiding the extreme biases caused by forced decision-making in hard classification methods. Furthermore, the above embodiments of this application calculate the pre-event and post-event thermal resistance using instantaneous clothing thermal resistance values ​​at multiple moments within a time window. This ensures that even if recognition errors occur in individual frames, the thermal resistance estimate within the window remains stable and does not fluctuate drastically due to instantaneous noise, effectively suppressing instantaneous recognition errors and improving the stability of thermal resistance calculation.

[0014] In one embodiment, determining the pre-event thermal resistance and post-event thermal resistance based on the instantaneous clothing thermal resistance values ​​at multiple moments within a first time window before the clothing action occurs, and the instantaneous clothing thermal resistance values ​​at multiple moments within a second time window after the clothing action occurs, includes: Using the quality factors of multiple moments within the first time window as weights, calculate the weighted average of the instantaneous clothing thermal resistance values ​​of multiple moments within the first time window, and use this average as the thermal resistance before the event. Using the quality factors at multiple moments within the second time window as weights, the weighted average of the instantaneous clothing thermal resistance values ​​at multiple moments within the second time window is calculated as the post-event thermal resistance.

[0015] The above embodiments are based on the thermal resistance calculation method of quality factor weighting, which realizes selective trust of visual recognition results. When the recognition quality is poor, the weight of the recognition result is automatically reduced; while when the recognition quality is high, the recognition result is automatically trusted. This can improve the accuracy of thermal resistance calculation before and after the event, thereby improving the accuracy of judging whether the clothing type has changed and avoiding false triggering of cabin environment actuators.

[0016] In one embodiment, determining a clothing change event when both dressing / undressing actions and a change in clothing type occur simultaneously includes: The direction of the action is determined based on the action state machine. The direction of clothing change is determined based on the thermal resistance before and after the event. A clothing change event is determined to have occurred when there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than a first threshold, the confidence level of the clothing status is greater than a second threshold, and the direction of the action is consistent with the direction of the change in clothing.

[0017] The above embodiments, based on dual verification of action and clothing type, combine confidence level and further perform directional consistency verification to ensure that action and clothing can corroborate each other, avoid false triggers caused by "tidying clothes / mistaken action", and further improve the reliability of judgment.

[0018] In one embodiment, determining a clothing change event when both dressing / undressing actions and a change in clothing type occur simultaneously includes: When there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than the first threshold, and the confidence level of the clothing status is greater than the second threshold, a clothing change event is determined to have occurred.

[0019] The above embodiments not only use dual verification of dressing and undressing actions and changes in clothing type to determine whether a clothing change event has occurred, but also incorporate confidence in dressing and undressing actions and confidence in clothing status, further improving the accuracy of clothing change event recognition and avoiding false triggers.

[0020] In one embodiment, determining the current predicted thermal sensation based on the current clothing thermal resistance parameter includes: Based on the current clothing thermal resistance parameters, the thermal comfort value is determined using the existing PMV thermal comfort prediction model. The current predicted thermal sensation is obtained by overlaying a user-personalized offset onto the thermal comfort value.

[0021] The above embodiments, based on the model-predicted thermal sensation, superimpose a user-personalized offset, which makes the predicted thermal sensation more accurate and in line with the user's actual experience, thus meeting the personalized needs of different users.

[0022] In one embodiment, the method provided in this application further includes: when the clothing change event indicates a clothing removal event, maintaining or reducing the seat heating level while increasing the air temperature.

[0023] This application automatically completes this complex balance through algorithms, completely freeing the user's hands, avoiding the problem of local overheating caused by simple linkage, and providing a more scientific and delicate comfort experience that conforms to human thermophysiology than manual adjustment by the user, truly achieving "warm but not dry" and eliminating driving safety hazards.

[0024] Secondly, embodiments of this application also provide a control device for the comfort of vehicle occupants, including: The image discrimination module is used to determine whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle; The thermal resistance calculation module is used to determine the thermal resistance parameters of the current clothing when a clothing change event is detected. The thermal sensation prediction module is used to determine the current predicted thermal sensation based on the current clothing thermal resistance parameters; The adjustment module is used to adjust the cabin environment actuators based on the current predicted thermal sensation and the target thermal sensation value.

[0025] Thirdly, embodiments of this application also provide an electronic device, including: Memory is used to store executable instructions for a computer; The processor, when executing computer-executable instructions stored in the memory, implements the method for controlling the comfort of vehicle occupants provided in the above embodiments.

[0026] Fourthly, embodiments of this application also provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the method for controlling the comfort of vehicle occupants provided in the above embodiments.

[0027] Fifthly, embodiments of this application also provide a computer program product, which includes computer-executable instructions stored in a computer-readable storage medium; When the processor of the electronic device reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, it implements the method for controlling the comfort of vehicle occupants provided in the above embodiments.

[0028] For the technical effects of the aforementioned control devices, electronic devices, storage media, and computer program products for occupant comfort, please refer to the technical effects of the control method for occupant comfort. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the hardware architecture of the vehicle occupant comfort control system provided in the embodiments of this application; Figure 2 This is a functional principle block diagram of the vehicle occupant comfort control system provided in the embodiments of this application; Figure 3 This is a flowchart illustrating a method for controlling passenger comfort in a vehicle, as provided in an embodiment of this application. Figure 4 yes Figure 3 Detailed flowchart of step S10 in the corresponding embodiment; Figure 5 yes Figure 4 Detailed flowchart of step S110 in the corresponding embodiment; Figure 6 yes Figure 4 Detailed flowchart of step S120 in the corresponding embodiment; Figure 7 This is a block diagram of a vehicle occupant comfort control device provided in an embodiment of this application; Figure 8 This is a block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0030] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this specification as detailed in the appended claims.

[0031] The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of this specification. The singular forms “a,” “the,” and “the” as used in this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0032] It should be understood that although the terms first, second, third, etc., may be used in this specification to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this specification, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0033] Figure 1 This is a schematic diagram of the hardware architecture of the vehicle occupant comfort control system provided in an embodiment of this application. Figure 1 As shown, it mainly includes the following hardware components: a sensing and acquisition module, a central processing unit, and an execution system connected in sequence.

[0034] The sensing and acquisition module includes an in-vehicle camera, an in-vehicle temperature sensor, a sunlight intensity sensor, vehicle speed sensors, and information on window opening / closing status. The in-vehicle camera is used to capture image sequences (i.e., continuous video frames) of the occupants inside the vehicle.

[0035] The central processing unit includes an image processing engine, a thermal comfort algorithm module, and a strategy control unit. The image processing engine determines whether a clothing change event has occurred based on a sequence of images of the occupants inside the vehicle. The thermal comfort algorithm module, when a clothing change event is determined, determines the thermal resistance parameter of the clothing after the change and, based on the thermal resistance parameter, determines the current predicted thermal sensation. The strategy control unit adjusts the cabin environment actuators based on the current predicted thermal sensation and the target thermal sensation value. The central processing unit can be used to execute the vehicle occupant comfort control method provided in the following embodiments of this application.

[0036] The actuator system refers to the aforementioned cabin environment actuators, including HVAV (Heating, Ventilation and Air Conditioning) systems and contact thermal systems (such as seat heating and ventilation systems).

[0037] Figure 2 This is a functional principle block diagram of the vehicle occupant comfort control system provided in an embodiment of this application. For example... Figure 2As shown, the system comprises a visual analysis layer, a model calculation layer, and a decision control layer. The visual analysis layer takes continuous video frames as input. By performing motion capture and clothing classification on these frames, it identifies clothing change events and determines the thermal resistance parameters of the clothing after the change. The model calculation layer takes environmental parameters, human body parameters, and metabolic rate as input. It calculates the real-time predicted thermal sensation value using the Fanger PMV equation. The decision control layer calculates the deviation PMV between the real-time predicted thermal sensation value and the user-set thermal sensation value (i.e., the thermal sensation target value). Based on the deviation PMV, it adjusts the outlet air temperature and airflow, regulates the seat heating or ventilation level, and can also adjust the airflow direction or dampers as needed.

[0038] Figure 3 This is a flowchart illustrating a method for controlling occupant comfort in a vehicle, as provided in an embodiment of this application. Figure 3 As shown, the method includes steps S10-S40.

[0039] Step S10: Determine whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle.

[0040] In one embodiment, the image sequence of the occupants inside the vehicle can be used to detect dressing and undressing actions to determine whether such actions exist. When dressing and undressing actions are detected, it is determined that a clothing change event has occurred.

[0041] In another embodiment, the image sequence of the vehicle occupants is subjected to clothing type change detection to determine whether the clothing type has changed. When the clothing type changes, it is determined that a clothing change event has occurred.

[0042] In other embodiments, such as Figure 4 As shown, step S10 specifically includes: step S110, performing dressing and undressing action detection on the image sequence of the occupants in the vehicle to determine whether dressing and undressing actions exist; step S120, performing clothing type change detection on the image sequence of the occupants in the vehicle to determine whether the clothing type has changed; step S130, when both dressing and undressing actions and clothing type changes exist simultaneously, determining that a clothing change event has occurred.

[0043] The embodiments described above in this application use the action characteristics of putting on and taking off clothes and the characteristics of clothing type changes. Only when both characteristics are satisfied is it determined that a clothing change event has occurred. This two-layer verification mechanism can improve the accuracy of clothing change event detection and avoid false triggering of "tidying clothes / mistaken action".

[0044] In one embodiment, such as Figure 5 As shown, step S110 specifically includes steps S111-S113.

[0045] Step S111: Perform skeletal keypoint detection on the image sequence to determine the wrist-neck distance feature, wrist-to-shoulder height feature, wrist velocity value, and hand-hand involvement feature for each image.

[0046] The detection of skeletal key points can be achieved using existing technologies. These key points can include the positional coordinates of the shoulder, elbow, wrist, neck, and hip. In one embodiment, after detecting the key points, scale normalization can be performed first. Specifically, the key point coordinates are normalized using shoulder width s(t) = ||shoulder_L - shoulder_R|| to reduce the influence of distance / viewpoint. shoulder_L represents the positional coordinates of the left shoulder, and shoulder_R represents the positional coordinates of the right shoulder.

[0047] Wherein, the distance from wrist to neck, d_neck(t), is equal to ||p_wrist - p_neck||. Here, p_wrist represents the coordinates of the wrist, and p_neck represents the coordinates of the neck.

[0048] The wrist is higher than the shoulder h(t) = y_shoulder - y_wrist (defined as positive to indicate the wrist is higher; note the coordinate system direction) where y_shoulder represents the y-axis coordinate of the shoulder and y_wrist represents the y-axis coordinate of the wrist.

[0049] The wrist velocity value v_wrist(t) = ||p_wrist(t) - p_wrist(t-Δt)|| / Δt, where p_wrist(t) represents the wrist position coordinate at time t, p_wrist(t-Δt) represents the wrist position coordinate at time t-Δt, and Δt represents the time difference.

[0050] The hand involvement characteristic refers to whether both hands simultaneously satisfy one of the following actions: "neck-to-neck / shoulder-to-arm / arm-raising".

[0051] Step S112: Construct an action state machine based on the wrist-neck distance feature, wrist-to-shoulder height feature, wrist velocity value, and hand involvement feature of each image.

[0052] Since dressing and undressing is a "active action" with a clear purpose, relatively strong force, and relatively fast speed, while "unintentional actions" such as scratching, adjusting collar, and touching face are usually slow and have short trajectories, the wrist speed value can be used to determine whether it is a "powerful and fast" action, and it is more likely to be a real dressing and undressing action.

[0053] The state machine for undressing includes at least the following steps: S0 Normal (initial state) → S1 Hand-to-Neck (wrists approaching the neck) → S2 Pull (over the shoulder / raised arm to pull) → S3 End (drop and pause trigger). The state machine for dressing includes at least the following steps: S0 Normal (initial state) → S1 (hands carrying fabric upwards over the head) → S2 wrists moving quickly downwards from above towards the neck → S3 End (wrists moving downwards towards the torso).

[0054] Under normal circumstances, both undressing and dressing require the participation of both hands. Based on the wrist-neck distance d_neck(t), the wrist-shoulder height difference h(t), the wrist speed value v_wrist(t), and whether both hands are involved, the user's current action (such as the wrist approaching the neck, passing over the shoulder, or falling back) can be determined, thus forming an action state machine.

[0055] Step S113: Based on the action state machine, when the undressing action sequence or the dressing action sequence is continuously satisfied, it is determined that the dressing and undressing action has occurred.

[0056] When the action state machine contains a sequence of actions to undress, it is determined that the undressing action has occurred. When the action state machine contains a sequence of actions to put on clothes, it is determined that the dressing action has occurred.

[0057] The above embodiments are based on wrist-neck distance (whether it is close to the collar), wrist height above shoulder (whether it is pulling overhead), wrist speed (whether it is a forceful and rapid movement), and hand involvement (whether it is a coordinated operation). A distracting action usually only satisfies one or two of these features (for example, scratching satisfies "hand close to the neck" but not "high speed" and "hand involvement"), while a genuine dressing and undressing action needs to satisfy multiple features simultaneously, thus achieving precise filtering. This application, through an action state machine, avoids the error of single-frame detection, effectively distinguishes dressing and undressing actions from distracting actions, has a low false trigger rate, avoids meaningless frequent actions of the air conditioner, and improves the user experience.

[0058] In one embodiment, the action of dressing or undressing is determined to have occurred only when the action state machine continuously satisfies the sequence of undressing or dressing actions and the confidence level of dressing or undressing actions is greater than a first threshold.

[0059] Specifically, the confidence score C_a for the action of dressing and undressing can be calculated using the following formula: C_a = σ( w1·I(min d_neck<τ_neck) + w2·I(max h>τ_h) + w3·max v_wrist + w4·I(hands involved) - b ) Where σ is a sigmoid function, w1, w2, w3, and w4 represent weights, I is an indicator function, min_d_neck is the minimum distance from wrist to neck, τ_neck is the wrist-to-neck distance threshold, max_h is the maximum height difference between wrist and shoulder, max_v_wrist is the maximum wrist speed, and b is the action decision bias term used to adjust the sensitivity of action detection, obtained through experimental calibration. Specifically, τ_neck can be 0.35–0.50; τ_h can be 0.10–0.20; and τ_v is the 90th percentile of the sample.

[0060] In one embodiment, such as Figure 6 As shown, the above step S120 specifically includes: steps S121-S124.

[0061] Step S121: Identify the image sequence and output the probability distribution of clothing categories corresponding to images at different times.

[0062] Specifically, an image classification model can be trained in advance to output the probability of each image belonging to each clothing category. The clothing category probability distribution includes the probability value of each category. The clothing category probability distribution at time T can be represented by q(t)={q1,…,qK}, where qK represents the probability value of belonging to the Kth category.

[0063] Step S122: Based on the probability distribution of clothing categories corresponding to images at different times and the equivalent thermal resistance corresponding to each category, the instantaneous clothing thermal resistance value at different times is calculated by weighting.

[0064] The equivalent thermal resistance of the Kth category can be represented by Ik, and the instantaneous clothing thermal resistance at time t can be represented by I_cl(t) = Σ_k qk(t)·Ik. The probability value of each category at time t is used as a weight, multiplied by the equivalent thermal resistance corresponding to each category, and the weighted sum is used to calculate the instantaneous clothing thermal resistance at time t.

[0065] Step S123: Determine the thermal resistance before the event and the thermal resistance after the event based on the instantaneous thermal resistance values ​​of clothing at multiple moments within the first time window before the dressing and undressing action occurs, and the instantaneous thermal resistance values ​​of clothing at multiple moments within the second time window after the dressing and undressing action occurs.

[0066] The timing of the dressing and undressing actions can be represented by t_e. The first time window refers to the preset time period before the dressing and undressing actions occur. For example, the first time window can be represented as W_pre=[t_e-T_pre, t_e-δ], where T_pre can be 2 seconds to 4 seconds and δ can be 0.8 seconds to 1.5 seconds. The second time window refers to the preset time period after the dressing and undressing actions occur. For example, the second time window can be represented as W_post=[t_e+δ, t_e+T_post], where T_post can be 2 seconds to 6 seconds and δ can be 0.8 seconds to 1.5 seconds.

[0067] The pre-event thermal resistance refers to the thermal resistance of the user's clothing before the dressing / undressing event occurs. The post-event thermal resistance refers to the thermal resistance of the user's clothing after the dressing / undressing event occurs.

[0068] In one embodiment, the thermal resistance before the event can be the average of the instantaneous clothing thermal resistance values ​​at multiple moments within a first time window. The thermal resistance after the event can be the average of the instantaneous clothing thermal resistance values ​​at multiple moments within a second time window.

[0069] In another embodiment, the quality factors of multiple moments within the first time window can be used as weights to calculate the weighted average of the instantaneous clothing thermal resistance values ​​of multiple moments within the first time window, which is used as the pre-event thermal resistance; the quality factors of multiple moments within the second time window can be used as weights to calculate the weighted average of the instantaneous clothing thermal resistance values ​​of multiple moments within the second time window, which is used as the post-event thermal resistance.

[0070] Specifically, the thermal resistance before the event can be expressed as I_cl^pre, which can be calculated using the following formula: I_cl^pre = (Σ_{t∈W_pre} Q(t)•I_cl(t)) / (Σ_{t∈W_pre} Q(t)).

[0071] The post-event thermal resistance can be expressed as I_cl^post, which can be calculated using the following formula: I_cl^post = (Σ_{t∈W_post}Q(t)·I_cl(t)) / (Σ_{t∈W_post}Q(t)).

[0072] Where Q(t) represents the quality factor at time t, and the quality factor Q(t)∈[0,1]. The quality factor can be represented by occlusion rate, confidence of skeletal keypoint detection, confidence of clothing category classification, etc.

[0073] The above embodiments are based on the thermal resistance calculation method of quality factor weighting, which realizes selective trust of visual recognition results. When the recognition quality is poor, the weight of the recognition result is automatically reduced; while when the recognition quality is high, the recognition result is automatically trusted. This can improve the accuracy of thermal resistance calculation before and after the event, thereby improving the accuracy of judging whether the clothing type has changed and avoiding false triggering of cabin environment actuators.

[0074] Step S124: When the absolute value of the difference between the thermal resistance before the event and the thermal resistance after the event is greater than a preset threshold, it is determined that the clothing type has changed.

[0075] Specifically, the difference between the thermal resistance before and after the event can be represented by ΔI_cl, where ΔI_cl = I_cl^post - I_cl^pre. ΔI_cl < 0 indicates undressing (thermal resistance decreases); ΔI_cl > 0 indicates dressing (thermal resistance increases). |ΔI_cl| > τ_cl (i.e., a preset threshold) indicates a change in clothing type. If ΔI_cl < -τ_cl, it indicates undressing; ΔI_cl > τ_cl indicates dressing.

[0076] In another embodiment, a change in clothing type can be determined when the absolute value of the difference between the thermal resistance before the event and the thermal resistance after the event is greater than a preset threshold and the confidence level of clothing status is greater than a second threshold.

[0077] The clothing status confidence score is used to characterize the accuracy of determining whether the clothing type has changed. The clothing status confidence score C_s can be calculated using the following formula: ,in The average quality of the window before and after the dressing and undressing actions occurs, with α and β being weighting coefficients.

[0078] It should be noted that existing technologies identify clothing categories by looking up the corresponding thermal resistance value in a table. This application, however, calculates the thermal resistance value based on the probability distribution of each category and the equivalent thermal resistance of each category using a weighted average. This eliminates the constraint of discrete category boundaries, allowing for a smooth transition between different categories and avoiding abrupt changes in thermal resistance parameters caused by category jumps. When the identification model is uncertain about the clothing category (e.g., a flat probability distribution across multiple categories), the weighted average naturally converges to a conservative estimate, avoiding the extreme biases caused by forced decision-making in hard classification methods. Furthermore, the above embodiments of this application calculate the pre-event and post-event thermal resistance using instantaneous clothing thermal resistance values ​​at multiple moments within a time window. Therefore, even if identification errors occur in individual frames, the thermal resistance estimate within the window remains stable and does not fluctuate drastically due to instantaneous noise, effectively suppressing instantaneous identification errors and improving the stability of thermal resistance calculation.

[0079] In one embodiment, step S130 may specifically include: when there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than a first threshold and the confidence level of the clothing status is greater than a second threshold, a clothing change event is determined to have occurred.

[0080] The calculation methods for the confidence scores of dressing and undressing actions and clothing status are described above.

[0081] The above embodiments not only use dual verification of dressing and undressing actions and changes in clothing type to determine whether a clothing change event has occurred, but also incorporate confidence in dressing and undressing actions and confidence in clothing status, further improving the accuracy of clothing change event recognition and avoiding false triggers.

[0082] In another embodiment, step S130 may specifically include: determining the direction of action based on the action state machine; determining the direction of clothing change based on the thermal resistance before and after the event; and determining that a clothing change event has occurred when there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than a first threshold, the confidence level of the clothing state is greater than a second threshold, and the direction of action is consistent with the direction of clothing change.

[0083] The direction of the action can include dressing and undressing, and the direction of clothing change includes dressing and undressing. The direction of the action being consistent with the direction of clothing change means that if the action is dressing, the direction of clothing change is also dressing; or, if the action is undressing, the direction of clothing change is also undressing.

[0084] The above embodiments, based on dual verification of action and clothing type, combine confidence level and further perform directional consistency verification to ensure that action and clothing can corroborate each other, avoid false triggers caused by "tidying clothes / mistaken action", and further improve the reliability of judgment.

[0085] Step S20: When a clothing change event is determined, determine the current clothing thermal resistance parameters.

[0086] It should be noted that if no clothing change event occurs, the status quo can be maintained without further calculations or adjustments.

[0087] Here, the current clothing thermal resistance parameter refers to the clothing thermal resistance *clo* at the current moment. The current clothing thermal resistance parameter can be represented by *I_cl^now*. To avoid PMV fluctuations caused by instantaneous changes in clothing recognition, a smooth transition is recommended. Therefore, *I_cl^now(t)* = λ•I_cl^now(t-Δt) + (1-λ)•I_cl^post, where *I_cl^now(t)* represents the clothing thermal resistance parameter at time *t*, and *I_cl^now(t-Δt)* represents the clothing thermal resistance parameter at time *t-Δt*. For example, λ = 0.6–0.85.

[0088] Step S30: Determine the current predicted thermal sensation based on the current clothing thermal resistance parameters.

[0089] In one embodiment, the thermal comfort value (i.e., PMV value) can be determined based on the current clothing thermal resistance parameter using an existing PMV thermal comfort prediction model. The PMV value is directly used as the current predicted thermal sensation. The PMV thermal comfort prediction model can be an existing model, and its inputs can include the current clothing thermal resistance parameter I_cl^now, cabin temperature T_a, humidity RH, air velocity v, mean radiant temperature T_r, and human metabolic rate M. PMV = (0.303•exp(-0.036•M) + 0.028) • L. Where L is the human body thermal balance residual (a combination of convection / radiation / evaporation / respiration, etc.).

[0090] In another embodiment, a thermal comfort value (i.e., PMV value) can be determined based on the current clothing thermal resistance parameters using an existing PMV thermal comfort prediction model; then, a user-personalized offset is superimposed on the thermal comfort value to obtain the current predicted thermal sensation.

[0091] Specifically, the user-personalized offset can be represented by b_user, where the user's current predicted thermal sensation TSV_personal = PMV - b_user. This user-personalized offset can be determined based on historical user feedback data. For example, the user might report feeling cold or hot in a specific environmental state or cockpit actuator setting, or the user might manually adjust actuator settings to determine the user-personalized offset.

[0092] The above embodiments, based on the model-predicted thermal sensation, superimpose a user-personalized offset, which makes the predicted thermal sensation more accurate and in line with the user's actual experience, thus meeting the personalized needs of different users.

[0093] Step S40: Adjust the cabin environment actuators based on the current predicted thermal sensation and the target thermal sensation value.

[0094] The thermal sensation target value can be represented by PMV_target, which defaults to 0 or the user-neutral point b_user. Specifically, the cockpit environment actuators can be adjusted based on the deviation e between the current predicted thermal sensation and the thermal sensation target value. e = PMV^now - PMV_target. PMV^now represents the current predicted thermal sensation.

[0095] The cabin environment actuators can include HVAC systems and seat heating and ventilation systems. For example, |e|≤0.2 represents the comfort zone, which can be maintained without adjustment; 0.2<|e|≤0.6 represents mild discomfort, mainly compensated with airflow / seat adjustment (low disturbance); 0.6<|e|≤1.2 represents moderate discomfort, requiring coordinated adjustment of airflow and target / supply air temperature; and |e|>1.2 represents severe discomfort, requiring rapid correction (still needing to prevent vibration and implement "miscontrol" protection).

[0096] Specifically, if the interface is the target temperature T_set, the change in target temperature ΔT_set = -K_T • sat(e). K_T represents the temperature control gain, and sat(e) represents the saturation function with respect to the deviation e. Limiting the magnitude of each change (e.g., ≤1.0°C), when e>0 (too hot), ΔT_set is negative, and the temperature is lowered; when e<0 (too cold), ΔT_set is positive, and the temperature is raised.

[0097] For controlling the airflow F, prioritize increasing airflow for hotter areas; for colder areas, avoid increasing airflow to prevent excessive cooling, and instead use a combination of heating and reducing airflow. The change in airflow ΔF = K_F•|e|•I(|e|>e_F), with an upper limit for the setting and a cooling time. The cooling time refers to the interval between two airflow adjustments (e.g., 10-20 seconds) to prevent frequent fluctuations. K_F represents the airflow control gain, and I(|e|>e_F) represents the indicator function. This value is 1 when |e| is greater than the threshold e_F; otherwise, it is 0. In other words, airflow is only adjusted when the deviation exceeds a certain threshold.

[0098] For seat control, if e < -0.6 (significantly cold): prioritize seat heating +1 level; HVAC slightly increase temperature; fan speed can be -1 level to avoid blowing cold air. If e > +0.6 (significantly hot): prioritize seat ventilation +1 level; HVAC slightly decrease temperature; fan speed can be +1 level to enhance heat dissipation.

[0099] Furthermore, to prevent jitter and implement "miscontrol" protection, the minimum adjustment interval for the same actuator is 10 to 20 seconds; the entry threshold e_on is higher than the exit threshold e_off (e.g., e_on=0.6, e_off=0.3) to suppress back-and-forth flipping. After each control is issued, u_stable (stable control point) is recorded; the error improvement is compared within the evaluation window Δt: Δe = |e(t)| - |e(t+Δt)|. If Δe < ε (ε represents the deviation improvement threshold), it indicates insufficient improvement or deterioration, and a rollback / degradation is initiated. If the system has just executed "warming up" but the user immediately executes "cooling down" (or reverse adjustment), it is judged as miscontrol: triggering rollback (i.e., canceling the warming control) and increasing the subsequent trigger threshold (increasing the confidence level T_a for dressing / undressing actions / confidence level T_s for clothing status or the entry threshold e_on in the short term).

[0100] In one embodiment, when the clothing change event indicates a clothing removal event, the seat heating level is maintained or reduced while the air temperature is increased.

[0101] It's important to note that thinner clothing not only affects air convection heat transfer but also significantly alters the efficiency of heat transfer between the body and the seat. When a user removes thick clothing (becomes thinner), the system may increase the air conditioning outlet temperature (non-contact compensation) while simultaneously reducing the seat heating power level (contact compensation). This is because thinner clothing allows heat from the seat to be more easily transferred to the body, and maintaining the original high setting might result in a "burning back" or excessive sweating. This precise, counterintuitive adjustment (heating the air and cooling the seat) is difficult for ordinary users to achieve manually. This application automatically balances this complex situation through an algorithm, completely freeing the user's hands, avoiding the localized overheating problems caused by simple linkage, and providing a more scientific and thermophysiologically compliant, delicate and comfortable experience than manual adjustment, truly achieving "warm but not dry" and eliminating driving safety hazards.

[0102] Existing infrared (IR) thermal imaging temperature control solutions are essentially feedback control. They require the occupants to remove clothing, allowing body heat to dissipate and causing an actual drop in surface temperature, which is then detected by the sensors (typically requiring several minutes of physiological reaction time) before the system intervenes to regulate temperature. This inevitably results in occupants experiencing an uncomfortable period of "first cold, then warm."

[0103] The technical solution provided in this application, before the physical change in body temperature occurs, predicts the heat change through an algorithm when a clothing change event is detected, and adjusts the cabin environment actuator in advance, realizing proactive feedback. This completely eliminates the lag in traditional automatic air conditioning adjustment, keeping the user in a constant and comfortable state throughout the entire process of clothing change, achieving a "zero-wait" temperature control experience and seamless connection of thermal comfort.

[0104] Existing infrared thermal imaging (IR) temperature control solutions are susceptible to interference from non-human heat sources. For example, in winter, when sunlight shines directly on an occupant's face, or when an occupant is holding a hot cup of coffee, the infrared sensor may detect localized high temperatures, mistakenly interpreting the occupant as "overheated" and reducing the heating output, causing other parts of the occupant's body to feel cold.

[0105] This application determines the current thermal resistance parameter through image analysis. This determination of physical properties is unaffected by sunlight or the temperature of the object being held. Even in scenarios with complex lighting or external heat sources, the system can accurately determine the actual warmth level of the occupants' clothing, avoiding "reverse misadjustment" of the air conditioning caused by sensor signal interference, and significantly improving the system's environmental adaptability and robustness.

[0106] Existing static vision initialization schemes typically only check clothing once when the vehicle starts, assuming that clothing remains unchanged during the journey. If occupants remove their coats during long-distance driving, the system still controls the temperature according to the "thick clothing" logic (e.g., maintaining a low level of heating), causing occupants to rapidly lose body temperature after removing their clothing.

[0107] This application introduces a real-time monitoring mechanism for clothing change events, which can dynamically reconstruct the thermal resistance parameters in the PMV model. This fills the gap in the existing technology for "dynamic changes during driving" scenarios, ensuring that the thermal comfort model (PMV) remains synchronized with the actual state of the occupants throughout the entire life cycle of the vehicle, thus avoiding "comfort gaps" during long-distance driving.

[0108] Traditional automatic air conditioning systems that rely on manual adjustment often require drivers to be distracted when their clothing changes and causes discomfort due to temperature changes, as they need to find and operate the air conditioning panel, fan speed knob, or screen menu, posing a safety hazard as their line of sight is deviated from the road.

[0109] The fully automated perception and compensation logic in this application's embodiments enables the system to perform all complex parameter adjustments on behalf of the user. This achieves "hands on the wheel, eyes on the road," improving comfort while objectively eliminating the risk of distracted driving caused by adjusting the air conditioning, thus enhancing the overall vehicle's active safety.

[0110] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

[0111] Figure 7 This is a block diagram of a vehicle occupant comfort control device provided in an embodiment of this application, such as... Figure 7As shown, the device includes: an image discrimination module 710, a thermal resistance calculation module 720, a sensory prediction module 730, and an execution adjustment module 740.

[0112] The image discrimination module 710 is used to determine whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle; The thermal resistance calculation module 720 is used to determine the thermal resistance parameters of the current clothing when a clothing change event is detected. The thermal sensation prediction module 730 is used to determine the current predicted thermal sensation based on the current clothing thermal resistance parameters; The adjustment module 740 is used to adjust the cabin environment actuators based on the current predicted thermal sensation and the target thermal sensation value.

[0113] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0114] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. Each method step is implemented by a corresponding module, which will not be described in detail here. The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of the solution in this specification according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0115] Figure 8 This specification illustrates a hardware structure diagram of an electronic device housing a control device for occupant comfort in a vehicle, according to an exemplary embodiment. Figure 8 As shown, the device may include: a processor 301, a memory 302, an input / output interface 303, a communication interface 304, and a bus 305. The processor 301, memory 302, input / output interface 303, and communication interface 304 are interconnected within the device via the bus 305.

[0116] The processor 301 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the steps of the vehicle occupant comfort control method provided in the embodiments of this specification.

[0117] The memory 302 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 302 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 302 and is called and executed by the processor 301.

[0118] Input / output interface 303 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0119] Communication interface 304 is used to connect a communication module (not shown in the figure) to enable communication and interaction between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0120] Bus 305 includes a pathway for transmitting information between various components of the device (e.g., processor 301, memory 302, input / output interface 303, and communication interface 304).

[0121] It should be noted that although the above-described device only shows the processor 301, memory 302, input / output interface 303, communication interface 304, and bus 305, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0122] This specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the vehicle occupant comfort control method as described in any of the embodiments of this specification.

[0123] This specification also provides a computer program product, which includes computer-executable instructions stored in a computer-readable storage medium; wherein, when the processor of an electronic device reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, it implements the steps of the vehicle occupant comfort control method described in the above embodiments of this application.

[0124] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0125] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0126] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

Claims

1. A method of controlling comfort of an occupant in a vehicle, characterized by, include: Based on the image sequence of the occupants inside the vehicle, determine whether a clothing change event has occurred; Once a clothing change event is determined, the thermal resistance parameters of the current clothing are determined. Based on the current clothing thermal resistance parameters, determine the current predicted thermal sensation; The cabin environment actuators are adjusted based on the current predicted thermal sensation and the target thermal sensation value.

2. The method according to claim 1, characterized in that, The step of determining whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle includes: The image sequence of the occupants inside the vehicle is subjected to clothing-wearing and undressing action detection to determine whether clothing-wearing and undressing actions exist. The image sequence of the occupants inside the vehicle is subjected to clothing type change detection to determine whether the clothing type has changed; A clothing change event is determined when there is a simultaneous action of putting on or taking off clothes and a change in the type of clothing.

3. The method according to claim 2, characterized in that, The step of detecting dressing and undressing actions in the image sequence of the occupants inside the vehicle to determine whether dressing and undressing actions exist includes: Skeletal keypoint detection is performed on the image sequence to determine the wrist-neck distance feature, wrist height above shoulder feature, wrist velocity value, and hand involvement feature of each image; An action state machine is constructed based on the wrist-neck distance features, wrist-to-shoulder height features, wrist velocity values, and hand-to-hand involvement features of each image. Based on the action state machine, when the sequence of undressing actions or the sequence of dressing actions are continuously satisfied, it is determined that the dressing or undressing action has occurred.

4. The method according to claim 3, characterized in that, The step of detecting changes in clothing type in the image sequence of the occupants inside the vehicle to determine whether the clothing type has changed includes: The image sequence is identified, and the probability distribution of clothing categories corresponding to images at different times is output. Based on the probability distribution of clothing categories corresponding to images at different times and the equivalent thermal resistance corresponding to each category, the instantaneous clothing thermal resistance values ​​at different times are calculated by weighting. Based on the instantaneous thermal resistance values ​​of clothing at multiple moments within the first time window before the dressing and undressing action occurs, and the instantaneous thermal resistance values ​​of clothing at multiple moments within the second time window after the dressing and undressing action occurs, the thermal resistance before the event and the thermal resistance after the event are determined. When the absolute value of the difference between the thermal resistance before and after the event is greater than a preset threshold, it is determined that the clothing type has changed.

5. The method according to claim 4, characterized in that, The determination of pre-event thermal resistance and post-event thermal resistance based on instantaneous clothing thermal resistance values ​​at multiple moments within a first time window before the clothing donning / undressing action, and instantaneous clothing thermal resistance values ​​at multiple moments within a second time window after the clothing donning / undressing action, includes: Using the quality factors of multiple moments within the first time window as weights, calculate the weighted average of the instantaneous clothing thermal resistance values ​​of multiple moments within the first time window, and use this average as the thermal resistance before the event. Using the quality factors at multiple moments within the second time window as weights, the weighted average of the instantaneous clothing thermal resistance values ​​at multiple moments within the second time window is calculated as the post-event thermal resistance.

6. The method according to claim 4, characterized in that, The determination of a clothing change event when both dressing / undressing actions and a change in clothing type occur simultaneously includes: The direction of the action is determined based on the action state machine. The direction of clothing change is determined based on the thermal resistance before and after the event. A clothing change event is determined to have occurred when there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than a first threshold, the confidence level of the clothing status is greater than a second threshold, and the direction of the action is consistent with the direction of the change in clothing.

7. The method according to claim 2, characterized in that, The determination of a clothing change event when both dressing / undressing actions and a change in clothing type occur simultaneously includes: When there is a simultaneous action of putting on or taking off clothes and a change in clothing type, and the confidence level of the action of putting on or taking off clothes is greater than the first threshold, and the confidence level of the clothing status is greater than the second threshold, a clothing change event is determined to have occurred.

8. The method according to claim 1, characterized in that, The step of determining the current predicted thermal sensation based on the current clothing thermal resistance parameters includes: Based on the current clothing thermal resistance parameters, the thermal comfort value is determined using the existing PMV thermal comfort prediction model. The current predicted thermal sensation is obtained by overlaying a user-personalized offset onto the thermal comfort value.

9. The method according to claim 1, characterized in that, Also includes: When the clothing change event indicates a clothing removal event, the seat heating level is maintained or reduced while the air temperature is increased.

10. A control device for passenger comfort in a vehicle, characterized in that, include: The image discrimination module is used to determine whether a clothing change event has occurred based on the image sequence of the occupants inside the vehicle; The thermal resistance calculation module is used to determine the thermal resistance parameters of the current clothing when a clothing change event is detected. The thermal sensation prediction module is used to determine the current predicted thermal sensation based on the current clothing thermal resistance parameters; The adjustment module is used to adjust the cabin environment actuators based on the current predicted thermal sensation and the target thermal sensation value.

11. An electronic device, characterized in that, include: Memory is used to store executable instructions for a computer; A processor, when executing computer-executable instructions stored in the memory, implements the method for controlling occupant comfort as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, The device stores computer-executable instructions, which, when executed by a processor, implement the method for controlling occupant comfort as described in any one of claims 1 to 9.

13. A computer program product, characterized in that, The computer program product includes computer-executable instructions stored in a computer-readable storage medium; When the processor of the electronic device reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, it implements the method for controlling the comfort of vehicle occupants as described in any one of claims 1 to 9.

Images