Method and device for constructing virtual temperature field in vehicle, electronic device, and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for constructing a virtual temperature field inside a vehicle. The method includes: acquiring air conditioning execution status signals, environmental auxiliary signals, and cabin status signals at multiple times within a preset time window before a specified time; constructing a thermal boundary condition vector for a specified area inside the vehicle at multiple times based on the air conditioning execution status signals, environmental auxiliary signals, and cabin status signals of the specified area at multiple times; using the thermal boundary condition vector at multiple times as input to a trained virtual temperature field prediction model, and outputting the virtual temperature of the specified area at the specified time through the virtual temperature field prediction model; and constructing a virtual temperature field at the specified time based on the virtual temperatures of different areas inside the vehicle at the specified time. The above solution solves the mass production problems caused by the use of high-density sensors to measure the temperature of different areas in the prior art, such as high hardware cost, complex wiring harnesses, difficult layout, and complex calibration and maintenance.

Description

Technical Field

[0001] This application relates to the field of automotive technology, and in particular to a method and apparatus for constructing a virtual temperature field inside a vehicle, an electronic device, a computer-readable storage medium, and a computer program product. Background Technology

[0002] As automotive air conditioning control technology evolves from traditional "set temperature control" to "human comfort control," the single in-vehicle ambient temperature is no longer sufficient to accurately represent the actual thermal environment. Currently, most solutions rely on mathematical models to predict the overall humidity or temperature inside the vehicle. However, the real in-vehicle thermal environment is characterized by strong spatial non-uniformity, high time-varying nature, high costs of traditional measurements, and higher requirements for upstream environmental input for downstream comfort algorithms. Sunlight exposure, vent patterns, air outlet direction, blower capacity, and seat position can lead to significant differences in the thermal environment of areas such as the head, torso, legs, and hands; while changes in the external environment, air conditioning actuator operation, door opening, and passenger entry and exit can also cause rapid changes in the local thermal environment.

[0003] Using high-density temperature, wind speed, radiation, and humidity sensors for real-world temperature field measurement typically faces mass production challenges such as high hardware costs, complex wiring harnesses, difficult layout, and complex calibration and maintenance. Therefore, achieving online construction of a virtual temperature field in the occupant area with low cost and low sensor dependence has become an important research direction in the field of automotive thermal management and comfort control. Summary of the Invention

[0004] This application provides a method and apparatus for constructing a virtual temperature field inside a vehicle, electronic equipment, and storage medium to solve mass production problems such as high hardware costs, complex wiring harnesses, difficult layout, and complex calibration and maintenance caused by using high-density sensors to measure the temperature of different areas.

[0005] In a first aspect, embodiments of this application provide a method for constructing a virtual temperature field inside a vehicle, including: Acquire air conditioning execution status signals, environmental auxiliary signals, and cabin status signals at multiple times within a preset time window prior to a specified time. For a specific area inside the vehicle, a thermal boundary condition vector for the specified area at multiple times is constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal of the specified area at multiple times. The thermal boundary condition vectors at multiple times are used as input to a trained virtual temperature field prediction model, which outputs the virtual temperature of the specified region at the specified time. Based on the virtual temperatures of different areas inside the vehicle at the specified time, a virtual temperature field is constructed for the specified time.

[0006] The above embodiments of this application provide a technical solution that constructs a boundary condition vector corresponding to a region by acquiring air conditioning execution status signals, environmental auxiliary signals, and cabin status signals, and predicts the virtual temperature of the region through a virtual temperature field prediction model. The virtual temperatures of different regions constitute a virtual temperature field. This solution eliminates the need to deploy sensors to measure temperature in each region, solving the mass production problems caused by the use of high-density sensors to measure the temperature of different regions in existing technologies, such as high hardware costs, complex wiring harnesses, difficult layout, and complex calibration and maintenance. The virtual temperature field can accurately reflect the differences in local thermal environment of different seats and different body areas, facilitating the control of local comfort.

[0007] In one embodiment, the step of constructing a thermal boundary condition vector for a designated area within the vehicle at multiple times, based on air conditioning execution status signals, environmental assistance signals, and cabin status signals of the designated area at multiple times, includes: Based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at any target time among the multiple time points, calculate the regional equivalent air supply capacity, regional equivalent air supply temperature, regional solar radiation heat index, and regional air disturbance index at the target time. Based on the area equivalent air supply capacity, area equivalent air supply temperature, area solar radiation heat index, area air disturbance index, vehicle outside temperature, and whether there are occupants in the designated area at the target time, construct the thermal boundary condition vector for the target time.

[0008] The technical solution provided in the above embodiments converts discrete air conditioning execution status signals, environmental auxiliary signals, and cabin status signals into thermal boundary condition vectors with more physical meaning. Instead of directly inputting several sensor values ​​into the model, it transforms fragmented controller signals into a unified input representation that can directly drive the construction of a virtual temperature field, thus solving the problems of discrete input, weak physical meaning, and difficult model learning.

[0009] In one embodiment, before using the thermal boundary condition vectors at multiple times as input to a trained virtual temperature field prediction model, and outputting the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model, the method further includes: A loss function constraint is introduced during the training process of the virtual temperature field prediction model; the loss function is: L = λ1·L_temp + λ2·L_smooth + λ3·L_boundary + λ4·L_time Where L_temp is the error between the predicted temperature and the calibration true value, L_smooth is the spatial smoothing constraint between adjacent regions, L_boundary is the consistency constraint between the predicted temperature and the supply air boundary conditions, L_time is the continuity constraint of the predicted temperature in the time dimension, and λ1, λ2, λ3 and λ4 are the weight coefficients of each loss term.

[0010] The above embodiments incorporate physical prior constraints during model training or inference, such as constraints on the continuity of temperature field changes, the correlation between the air supply area and adjacent body areas, and the requirement that the temperature field should not exhibit non-physical jumps before and after damper switching. By constructing the aforementioned loss function, the divergence of pure black-box models under limited sample conditions can be avoided, thereby improving interpretability and cross-condition generalization ability.

[0011] In one embodiment, the spatial smoothing constraint between adjacent regions is calculated using the following formula: L_smooth = (1 / T)•ΣΣ_(i,j∈N)(P_i(t)-P_j(t))^2 Where T is the number of time samples, i represents the specified region, j is the neighboring region of the specified region i, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, and P_j(t) is the predicted temperature of neighboring region j of region i.

[0012] The calculation of the spatial smoothing constraints between adjacent regions can prevent temperature abrupt changes in adjacent body regions without physical basis, thereby improving the accuracy of model predictions.

[0013] In one embodiment, the consistency constraint between the predicted temperature and the supply air boundary conditions is calculated using the following formula: L_boundary=(1 / N•T))•ΣΣ[P_i(t)-(T_cab(t)+k_i•Q_air,i(t)•(T_supply,i(t)-T_cab(t)))]^2 Where T is the number of time samples, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, T_cab(t) is the average temperature inside the vehicle at time t, k_i is the region sensitivity coefficient, Q_air,i(t) is the equivalent air supply capacity of region i at time t, and T_supply,i(t) is the equivalent air supply temperature of region i at time t.

[0014] The aforementioned consistency constraint between the predicted temperature and the air supply boundary conditions can keep the predicted temperature between the average temperature inside the vehicle and the air supply temperature. The weights are determined by the air supply capacity and regional characteristics, which forces the virtual temperature output by the model to not violate the basic physical laws of the air conditioning system, thus ensuring that the model has physical interpretability.

[0015] In one embodiment, after outputting the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model, the method further includes: Based on the virtual temperature of the specified region at the specified time, the known dynamic temperature difference compensation amount of the specified region at the specified time is superimposed to obtain the corrected virtual temperature.

[0016] The technical solution provided in the above embodiments takes into account blower aging, actual damper opening deviation, evaporator efficiency decay, and seasonal differences, and performs zero-bias correction, vehicle aging correction, seasonal correction, and actuator deviation correction on the output of the virtual temperature field prediction model. Thus, it can maintain long-term prediction stability without relying on a large number of new sensors.

[0017] In one embodiment, before obtaining the corrected virtual temperature by superimposing a known dynamic temperature difference compensation amount for the specified region at the specified time onto the virtual temperature of the specified region at the specified time, the method further includes: The measured temperature values ​​of the specified area at the specified time are compared with the virtual temperature output by the virtual temperature field prediction model to determine the dynamic temperature difference compensation amount.

[0018] This application allows the use of a small number of existing sensors as a low-cost calibration source. By calculating the deviation from the virtual temperature output by the model, the dynamic temperature difference compensation amount is determined, and online correction is achieved using a small amount of measured data, thereby improving robustness across vehicle models and operating conditions.

[0019] In one embodiment, constructing the virtual temperature field at the specified time based on the virtual temperatures of different areas inside the vehicle at the specified time includes: Based on the prediction confidence of the virtual temperature field prediction model for a specified area, when the prediction confidence is greater than a first threshold, the virtual temperature of the specified area output by the virtual temperature field prediction model is used to construct the virtual temperature field. When the prediction confidence of the specified region is greater than the second threshold and less than the first threshold, the virtual temperature field is constructed using the corrected virtual temperature. When the prediction confidence level of the specified region is less than the second threshold, a downgrade control is performed.

[0020] The solution provided in the above embodiments not only outputs the virtual temperature but also the predicted confidence level for each region. When the confidence level is lower than the second threshold, it can switch to a simplified temperature field, revert to the traditional in-vehicle temperature closed-loop control, or reduce the weight of local comfort control. This improves mass production usability and functional safety friendliness, and prevents erroneous amplification when the model is abnormal.

[0021] Secondly, embodiments of this application also provide a device for constructing a virtual temperature field inside a vehicle, including... Acquire air conditioning execution status signals, environmental auxiliary signals, and cabin status signals at multiple times within a preset time window prior to a specified time. For a specific area inside the vehicle, a thermal boundary condition vector for the specified area at multiple times is constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal of the specified area at multiple times. The thermal boundary condition vectors at multiple times are used as input to a trained virtual temperature field prediction model, which outputs the virtual temperature of the specified region at the specified time. Based on the virtual temperatures of different areas inside the vehicle at the specified time, a virtual temperature field is constructed for the specified time.

[0022] 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 constructing a virtual temperature field inside a vehicle provided in the above embodiments.

[0023] 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 constructing a virtual temperature field inside a vehicle provided in the above embodiments.

[0024] 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 constructing a virtual temperature field inside the vehicle provided in the above embodiments.

[0025] For the technical effects of the in-vehicle virtual temperature field construction apparatus, electronic device, computer-readable storage medium, and computer program product provided in the above embodiments of this application, please refer to the above-described in-vehicle virtual temperature field construction method. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating a method for constructing a virtual temperature field inside a vehicle, as provided in an embodiment of this application. Figure 2 yes Figure 1 Detailed flowchart of step S120 in the corresponding embodiment; Figure 3This is a block diagram of the in-vehicle virtual temperature field construction device provided in the embodiments of this application; Figure 4 This is a block diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0027] 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.

[0028] 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.

[0029] 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."

[0030] In the first related solution, there is an intelligent airflow control scheme based on temperature distribution field and BP neural network. This approach involves acquiring the cabin temperature distribution field and the setpoint of the air outlet working vector, then using a BP neural network to obtain the theoretical value of the air outlet working vector and achieve intelligent adjustment. This type of solution already involves key concepts such as "cabin temperature distribution field," "neural network," and "intelligent airflow control."

[0031] However, the core objective of this type of solution is to deduce vent control based on the existing temperature distribution field, without solving the problem of "how to construct the temperature distribution field of the vehicle cabin".

[0032] In related solution two, there is a virtual humidity / virtual temperature sensor solution based on mathematical models. This type of solution mainly uses mathematical models to predict the overall humidity or temperature inside the vehicle, aiming to reduce the number of components and costs. Its essence is "virtual sensor / virtual environmental quantity estimation." However, it does not emphasize the spatialized temperature field representation for the occupant's body area, nor does it highlight the temporal neural network reconstruction based on the internal operating state signals of the air conditioning system.

[0033] Among the relevant solutions, there is an air conditioning control scheme based on a 3D cockpit model and deep learning / reinforcement learning. Another scheme establishes a thermal comfort prediction model based on a 3D model of the passenger compartment, a human thermal comfort model, and deep learning, and further uses reinforcement learning for air conditioning cooling control. These schemes often use signals such as solar radiation, air temperature after the evaporator, air velocity after the evaporator, and interior and exterior temperatures, and set multiple thermal comfort monitoring points through a 3D model. However, this type of scheme deviates further from the online simulation and control training framework, relying on 3D cabin models and co-simulation, and does not emphasize "real-time, lightweight, and low-sensor-cost online construction of a virtual temperature field on the vehicle."

[0034] In related scheme four, there is a scheme for estimating virtual solar load / virtual ambient temperature. Additionally, there is a technical approach that estimates virtual solar load or virtual ambient temperature using external vehicle temperature, cabin temperature, mass flow rate, and actual solar radiation, and then uses this estimation for climate control. This demonstrates the engineering rationale for using existing signals to infer thermal environment parameters that are difficult to measure directly. However, these schemes mainly output a single scalar, not the spatial temperature field oriented towards the occupant's body area, and do not solve the problem of online reconstruction involving "spatial distribution + temporal variation + multiple actuator states driving together."

[0035] Figure 1 This is a flowchart illustrating a method for constructing a virtual temperature field inside a vehicle, as provided in an embodiment of this application. Figure 1 As shown, the method includes steps S110-S140.

[0036] Step S110: Obtain the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at multiple times within a preset time window before the specified time.

[0037] The specified time can be any time, represented by 't'. The preset time window is a period of time before and including the specified time. For example, the preset time window can be 1 minute, 3 minutes, or 5 minutes. Multiple times within 3 minutes before time 't' can be represented by time 't', time 't-1', time 't-2', time 't-3', ..., time 'tn'. The time interval between these multiple times can be 100ms, 200ms, or 500ms.

[0038] In one embodiment, the air conditioning operation status signals may include: evaporator exhaust air temperature, main outlet air temperature, blower voltage / duty cycle / speed, mixing damper opening, mode damper opening, internal / external circulation damper opening, and the enabling status of air outlets in each zone. Environmental auxiliary signals may include: solar radiation intensity, outside temperature, vehicle speed, and vehicle orientation / sun incidence angle. Cabin status signals may include: seat occupancy status, seat rail position, backrest angle, steering wheel position, and door opening status. These signals can be read from the vehicle bus, the air conditioning controller, and the cabin domain controller.

[0039] Step S120: For a designated area inside the vehicle, construct a thermal boundary condition vector for the designated area at multiple times based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal of the designated area at multiple times.

[0040] The designated area refers to the body area of ​​the occupants inside the vehicle, such as the driver's head area, driver's chest area, driver's legs and feet area, driver's hands area, the corresponding area for the front passenger, and the corresponding areas for the left / middle / right sides of the second row. Virtual temperature can be calculated for each area inside the vehicle individually, thereby constructing a virtual temperature field.

[0041] For any given region, the actual in-vehicle thermal environment in that region will also change over time. Therefore, the thermal boundary condition vector at time t can be constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at time t; the thermal boundary condition vector at time t-1 can be constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at time t-1; the thermal boundary condition vector at time t-2 can be constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at time t-2, and so on.

[0042] In one embodiment, such as Figure 2 As shown, the above step S120 specifically includes step S121 and step S122.

[0043] Step S121: Based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at any target time among the plurality of times, calculate the regional equivalent air supply capacity, regional equivalent air supply temperature, regional solar radiation heat index, and regional air disturbance index at the target time.

[0044] Step S122: Based on the regional equivalent air supply capacity, regional equivalent air supply temperature, regional solar radiation heat index, regional air disturbance index, vehicle outside temperature, and whether there are occupants in the designated area at the target time, construct the thermal boundary condition vector for the target time.

[0045] Here, "target time" refers to any one of multiple times. "Regional equivalent air supply capacity" is the quantified value of the effective airflow capacity actually allocated to the i-th region for heat exchange. "Regional equivalent air supply temperature" is the equivalent temperature of the mixed air supplied to the i-th region. "Regional solar radiation heat index" refers to the relative intensity of local heat load caused by solar radiation projected onto the i-th region, taking into account vehicle orientation and window geometry (typically normalized to a range of 0-1 or 0-10). "Regional air disturbance index" is a quantified indicator of airflow fluctuations, draft sensation, or turbulence intensity within the i-th region.

[0046] Taking time t as an example, the regional equivalent air supply capacity, regional equivalent air supply temperature, regional solar radiation heat index, and regional air disturbance index at time t can be calculated using the following formulas.

[0047] For example, (1) the equivalent air supply capacity of the area is Q_air,i(t) = f1(V_blower(t), D_mode(t),D_mix(t), D_zone,i(t)).

[0048] Where Q_air,i(t) represents the equivalent air supply capacity of the i-th region; V_blower(t) represents the blower voltage or equivalent speed; D_mode(t) represents the mode gate state; D_mix(t) represents the mixing damper state; and D_zone,i(t) represents the damper or air outlet path allocation state corresponding to the i-th region.

[0049] (2) The equivalent supply air temperature of the region is T_supply,i(t) = f2(T_evap(t), D_mix(t), Recirc(t)).

[0050] Where T_supply,i(t) represents the equivalent supply air temperature of the region; T_evap(t) represents the air temperature after the evaporator; D_mix(t) represents the mixing damper status; and Recirc(t) represents the opening degree of the internal and external circulation dampers.

[0051] (3) Regional solar heat index S_i(t) = f3(Solar(t), SeatPos_i, VehicleHeading,WindowGeom_i).

[0052] Where S_i(t) represents the solar heat index of the region; Solar(t) represents the solar radiation intensity; SeatPos_i represents the seat position corresponding to region i; VehicleHeading represents the vehicle orientation; and WindowGeom_i represents the window parameters corresponding to region i.

[0053] (4) Regional air disturbance index A_i(t) = f4(Q_air,i(t), VehicleSpeed(t),OutletMode(t)).

[0054] Where A_i(t) represents the regional air disturbance index; Q_air,i(t) represents the equivalent air supply capacity of the i-th region; VehicleSpeed(t) represents the vehicle speed; and OutletMode(t) represents the air outlet mode.

[0055] f1, f2, f3, and f4 represent the equivalent air supply capacity function, the equivalent air supply temperature function, the solar radiation heating index function, and the air disturbance index function, respectively. The specific form of each function can be achieved by looking up a table, polynomial fitting, or physical simplification formula, and is not limited to a single mathematical expression.

[0056] For example, the thermal boundary condition vector at time t is X_i(t) = [Q_air,i(t), T_supply,i(t), S_i(t), A_i(t), T_out(t), Occ_i(t), …].

[0057] Where T_out(t) represents the outside temperature, and Occ_i(t) indicates whether there are occupants in the i-th region. Depending on the needs, the thermal boundary condition vector at time t can also include the door opening status, backrest angle, etc.

[0058] The technical solution provided in the above embodiments converts discrete air conditioning execution status signals, environmental auxiliary signals, and cabin status signals into thermal boundary condition vectors with more physical meaning. Instead of directly inputting several sensor values ​​into the model, it transforms fragmented controller signals into a unified input representation that can directly drive the construction of a virtual temperature field, thus solving the problems of discrete input, weak physical meaning, and difficult model learning.

[0059] Step S130: Use the thermal boundary condition vectors at multiple times as input to the trained virtual temperature field prediction model, and output the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model.

[0060] Specifically, the thermal boundary condition vectors X_i(t), X_i(t-1), …, X_i(tn) of region i at times t1, t-1, t-2…tn can be used together as input to the virtual temperature field prediction model to obtain the virtual temperature of region i at time t output by the virtual temperature field prediction model.

[0061] The virtual temperature field prediction model can be trained in advance using sample data. This sample data includes air conditioning execution status signals, environmental auxiliary signals, cabin status signals, and temperature measurements of a specified area at a target time, occurring multiple times prior to that time. The virtual temperature field prediction model can be trained using NARX (Nonlinear Autoregressive with eXogenous inputs), LSTM (Long Short-Term Memory), GRU (Gated Recurrent Unit), temporal convolutional networks, or a hybrid model combining a physical constraint layer and a lightweight multilayer perceptron.

[0062] In one embodiment, a loss function constraint is introduced during the training process of the virtual temperature field prediction model; the loss function is: L = λ1·L_temp + λ2·L_smooth + λ3·L_boundary + λ4·L_time Where L_temp is the error between the predicted temperature and the calibration true value, L_smooth is the spatial smoothing constraint between adjacent regions (used to prevent temperature abrupt changes without physical basis in adjacent body regions), L_boundary is the consistency constraint between the predicted temperature and the air supply boundary conditions (used to ensure that the model output is consistent with the direction of the air supply boundary conditions), L_time is the continuity constraint of the predicted temperature in the time dimension (used to suppress non-physical jumps in time), and λ1, λ2, λ3 and λ4 are the weight coefficients of each loss term.

[0063] The above embodiments incorporate physical prior constraints during model training or inference, such as constraints on the continuity of temperature field changes, the correlation between the air supply area and adjacent body areas, the absence of non-physical jumps in the temperature field before and after damper switching, and directional differences in heat increments between the sunlit and non-sunlit sides. By constructing the aforementioned loss function, the divergence of pure black-box models under limited sample conditions can be avoided, improving interpretability and cross-condition generalization ability.

[0064] In one embodiment, (1) the error between the predicted temperature and the calibration true value: L_temp = (1 / (N·T))·ΣΣ(P_i(t)-T_gt,i(t))^2 Where T is the number of time samples, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, and T_gt,i(t) represents the true value of region i at time t; the predicted temperature refers to the virtual temperature output by the model during the training phase.

[0065] (2) Spatial smoothing constraints between adjacent regions: L_smooth = (1 / T)·ΣΣ_(i,j∈N)( P_i(t)- P_j(t))^2 Where T is the number of time samples, i represents the specified region, j is the neighboring region of the specified region i, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, and P_j(t) is the predicted temperature of neighboring region j of region i.

[0066] (3) Consistency constraint between predicted temperature and supply air boundary conditions: L_boundary=(1 / N·T))·ΣΣ[P_i(t)-(T_cab(t)+k_i·Q_air,i(t)·(T_supply,i(t)-T_cab(t)))]^2.

[0067] Where T is the number of time samples, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, T_cab(t) is the average temperature inside the vehicle at time t, k_i is the region sensitivity coefficient, Q_air,i(t) is the equivalent air supply capacity of region i at time t, and T_supply,i(t) is the equivalent air supply temperature of region i at time t.

[0068] (4) The continuity constraint of predicted temperature in the time dimension: L_time = (1 / (N·T))·ΣΣ(P_i(t)-P_i(t-1))^2.

[0069] Where T is the number of time samples, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, and P_i(t-1) is the predicted temperature of region i at time t-1.

[0070] The calculation of the spatial smoothing constraints between adjacent regions can prevent temperature abrupt changes in adjacent body regions without physical basis, thereby improving the accuracy of model predictions.

[0071] The aforementioned consistency constraint between the predicted temperature and the air supply boundary conditions can keep the predicted temperature between the average temperature inside the vehicle and the air supply temperature. The weights are determined by the air supply capacity and regional characteristics, which forces the virtual temperature output by the model to not violate the basic physical laws of the air conditioning system, thus ensuring that the model has physical interpretability.

[0072] In one embodiment, in addition to outputting the virtual temperature of a specified region at a specified time, the temperature change rate ΔP_i(t) and the prediction confidence level C_i(t) of the specified region can also be calculated. The prediction confidence level can be calculated based on the variance of the model output.

[0073] Step S140: Construct a virtual temperature field at the specified time based on the virtual temperatures of different areas inside the vehicle at the specified time.

[0074] In one embodiment, after predicting the virtual temperature of each region using a virtual temperature field prediction model, the virtual temperature field is obtained. The virtual temperature field includes the virtual temperature of each region, and the temperature differences between different regions can be determined through the virtual temperature field.

[0075] In another embodiment, step S140 may specifically include: based on the prediction confidence of the virtual temperature field prediction model for a specified area, when the prediction confidence is greater than a first threshold, constructing the virtual temperature field using the virtual temperature of the specified area output by the virtual temperature field prediction model; when the prediction confidence of the specified area is greater than a second threshold but less than the first threshold, constructing the virtual temperature field using the corrected virtual temperature; and when the prediction confidence of the specified area is less than the second threshold, performing degradation control.

[0076] The first threshold is greater than the second threshold. That is, when the prediction confidence for a specified area is high (greater than the first threshold), the virtual temperature of that area output by the model can be directly used to construct a virtual temperature field. When the prediction confidence for a specified area tends to be between the first and second thresholds, the virtual temperature output by the model needs to be corrected, and the corrected virtual temperature is used to construct the virtual temperature field. When the prediction confidence is low (below the second threshold), degradation control is implemented. Specifically, degradation control can be to simplify the virtual temperature field (excluding the virtual temperature of the specified area), stop constructing the virtual temperature field and revert to the traditional air conditioning temperature control method, or reduce the weight of the specified area when using its virtual temperature in subsequent calculations (e.g., to calculate comfort).

[0077] The solution provided in the above embodiments not only outputs the virtual temperature but also the predicted confidence level for each region. When the confidence level is lower than the second threshold, it can switch to a simplified temperature field, revert to the traditional in-vehicle temperature closed-loop control, or reduce the weight of local comfort control. This improves mass production usability and functional safety friendliness, and prevents erroneous amplification when the model is abnormal.

[0078] In one embodiment, the corrected virtual temperature is obtained by superimposing a known dynamic temperature difference compensation amount of the specified region at the specified time on the virtual temperature of the specified region at the specified time.

[0079] It should be noted that since the virtual temperature field prediction model is trained based on historical data, deviations that could not be fully predicted during training may occur in real vehicles. For example, an aging blower can lead to a decrease in actual airflow, deviations in the actual opening of the dampers can result in inaccurate air distribution, evaporator efficiency decline can lead to a decrease in cooling capacity, and seasonal differences (winter / summer, sunny / cloudy days) can cause the model to drift. These deviations will result in a slowly changing error between the virtual temperature output by the model and the actual temperature felt by the occupants over time. Therefore, the dynamic temperature difference compensation amount can vary in different areas, and the dynamic temperature difference compensation amount can also vary in the same area at different times. The dynamic temperature difference compensation amount is used to compensate for the virtual temperature output by the model. The specific update time of the dynamic temperature difference compensation amount can be determined based on the in-vehicle temperature sensor, the air outlet temperature, the sunlight sensor, and the operating condition label. For example, when the detection result of the in-vehicle temperature sensor deviates significantly from the model prediction result, it is necessary to determine the dynamic temperature difference compensation amount.

[0080] Specifically, the corrected virtual temperature can be calculated using the following formula: P_i^corr(t) = P_i(t) + b_i(t) Where P_i^corr(t) represents the virtual temperature of region i at time t after correction, P_i(t) represents the virtual temperature of region i at time t, and b_i(t) represents the dynamic temperature difference compensation amount of region i at time t.

[0081] The technical solution provided in the above embodiments takes into account blower aging, actual damper opening deviation, evaporator efficiency decay, and seasonal differences, and performs zero-bias correction, vehicle aging correction, seasonal correction, and actuator deviation correction on the output of the virtual temperature field prediction model. Thus, it can maintain long-term prediction stability without relying on a large number of new sensors.

[0082] In one embodiment, several measured temperature values ​​of the designated area at the designated time are compared with the virtual temperature output by the virtual temperature field prediction model to determine the dynamic temperature difference compensation amount.

[0083] Specifically, the difference between the actual temperature measured by a small number of sensors and the virtual temperature is used as the dynamic temperature difference compensation amount.

[0084] This application allows the use of a small number of existing sensors as low-cost calibration sources, such as in-vehicle temperature sensors, evaporator exhaust temperature, air vent temperature, sunlight sensors, or optional local skin temperature / infrared temperature points. By calculating the deviation from the virtual temperature output by the model, the dynamic temperature difference compensation amount is determined, and online correction is achieved using a small amount of measured data, improving robustness across vehicle models and operating conditions.

[0085] Current technologies for accurately acquiring spatial temperature fields in mass-produced vehicles often require a large number of sensors, leading to a significant increase in cost, layout space, wiring complexity, and calibration difficulty. Furthermore, most existing solutions focus on predicting scalar quantities such as overall virtual humidity, overall virtual solar load, and overall virtual ambient temperature, which are insufficient to accurately reflect the differences in local thermal environments between different seats and different body areas.

[0086] The technical solution provided in this application does not require a large number of sensors. Instead of outputting a single "in-vehicle temperature" or the average temperature of the entire cabin, it constructs a virtual temperature field around areas relevant to occupant thermal comfort, such as the driver's head, chest, abdomen, hands, and legs / feet, as well as corresponding areas in the front passenger seat and the left / middle / right of the second row. It maintains predictive stability under multiple passengers, multiple seats, different sunlight conditions, and different airflow modes. When necessary, it can also output regional wind intensity indices, regional temperature change rates, regional cumulative heat exposure, and regional prediction confidence levels. This naturally aligns upstream environmental estimation with downstream comfort prediction models, avoiding the problem of single-point temperature failing to support local comfort control. The virtual temperature field constructed in this application can serve as a general intermediate layer for intelligent air conditioning, outputting air temperature, thermal disturbance index, equivalent heating / cooling intensity, temperature change trend, and prediction confidence levels for each occupant and their body parts. The results output by this application can simultaneously serve comfort control, energy-optimal control, active interaction, and preference learning, enhancing system reusability.

[0087] 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.

[0088] Figure 3 This is a block diagram of a device for constructing a virtual temperature field inside a vehicle, as provided in an embodiment of this application. Figure 3 As shown, the device includes: a signal acquisition module 310, a vector construction module 320, a temperature prediction module 330, and a temperature field construction module 340.

[0089] The signal acquisition module 310 is used to acquire air conditioning execution status signals, environmental auxiliary signals and cabin status signals at multiple times within a preset time window before a specified time. The vector construction module 320 is used to construct a thermal boundary condition vector for a specified area inside the vehicle at multiple times, based on the air conditioning execution status signal, environmental auxiliary signal and cabin status signal of the specified area at multiple times. The temperature prediction module 330 is used to take the thermal boundary condition vectors at multiple times as input to the trained virtual temperature field prediction model, and output the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model. Temperature field construction module 340 is used to construct a virtual temperature field at the specified time based on the virtual temperature of different areas inside the vehicle at the specified time.

[0090] 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.

[0091] 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.

[0092] Figure 4 This specification illustrates a hardware structure diagram of an electronic device containing a vehicle interior virtual temperature field construction device according to an exemplary embodiment, such as... Figure 4 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.

[0093] 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 method for constructing the virtual temperature field inside the vehicle provided in the embodiments of this specification.

[0094] 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.

[0095] 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.

[0096] 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.).

[0097] 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).

[0098] 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.

[0099] This specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method for constructing a virtual temperature field inside a vehicle as provided in any embodiment of this specification.

[0100] 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 method for constructing a virtual temperature field inside a vehicle as described in the above embodiments of this application.

[0101] 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.

[0102] 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.

[0103] 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 for constructing a virtual temperature field inside a vehicle, characterized in that, include Acquire air conditioning execution status signals, environmental auxiliary signals, and cabin status signals at multiple times within a preset time window prior to a specified time. For a specific area inside the vehicle, a thermal boundary condition vector for the specified area at multiple times is constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal of the specified area at multiple times. The thermal boundary condition vectors at multiple times are used as input to a trained virtual temperature field prediction model, which outputs the virtual temperature of the specified region at the specified time. Based on the virtual temperatures of different areas inside the vehicle at the specified time, a virtual temperature field is constructed for the specified time.

2. The method according to claim 1, characterized in that, For a designated area inside the vehicle, based on the air conditioning execution status signals, environmental auxiliary signals, and cabin status signals of the designated area at multiple times, a thermal boundary condition vector for the designated area at multiple times is constructed, including: Based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal at any target time among the multiple time points, calculate the regional equivalent air supply capacity, regional equivalent air supply temperature, regional solar radiation heat index, and regional air disturbance index at the target time. Based on the area equivalent air supply capacity, area equivalent air supply temperature, area solar radiation heat index, area air disturbance index, vehicle outside temperature, and whether there are occupants in the designated area at the target time, construct the thermal boundary condition vector for the target time.

3. The method according to claim 1, characterized in that, Before using the thermal boundary condition vectors at multiple times as input to a trained virtual temperature field prediction model, and outputting the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model, the method further includes: A loss function constraint is introduced during the training process of the virtual temperature field prediction model; the loss function is: L = λ1·L_temp + λ2·L_smooth + λ3·L_boundary + λ4·L_time Where L_temp is the error between the predicted temperature and the calibration true value, L_smooth is the spatial smoothing constraint between adjacent regions, L_boundary is the consistency constraint between the predicted temperature and the supply air boundary conditions, L_time is the continuity constraint of the predicted temperature in the time dimension, and λ1, λ2, λ3 and λ4 are the weight coefficients of each loss term.

4. The method according to claim 3, characterized in that, The spatial smoothing constraint between adjacent regions is calculated using the following formula: L_smooth = (1 / T)•ΣΣ_(i,j∈N)(P_i(t)-P_j(t))^2 Where T is the number of time samples, i represents the specified region, j is the neighboring region of the specified region i, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, and P_j(t) is the predicted temperature of neighboring region j of region i.

5. The method according to claim 3, characterized in that, The consistency constraint between the predicted temperature and the supply air boundary conditions is calculated using the following formula: L_boundary=(1 / N•T))•ΣΣ[P_i(t)-(T_cab(t)+k_i•Q_air,i(t)•(T_supply,i(t)-T_cab(t)))]^2 Where T is the number of time samples, N is the number of regions, P_i(t) is the predicted temperature of region i at time t, T_cab(t) is the average temperature inside the vehicle at time t, k_i is the region sensitivity coefficient, Q_air,i(t) is the equivalent air supply capacity of region i at time t, and T_supply,i(t) is the equivalent air supply temperature of region i at time t.

6. The method according to claim 1, characterized in that, After outputting the virtual temperature of the specified region at the specified time through the virtual temperature field prediction model, the method further includes: Based on the virtual temperature of the specified region at the specified time, the known dynamic temperature difference compensation amount of the specified region at the specified time is superimposed to obtain the corrected virtual temperature.

7. The method according to claim 6, characterized in that, Before obtaining the corrected virtual temperature by superimposing a known dynamic temperature difference compensation amount for the specified region at the specified time onto the virtual temperature of the specified region at the specified time, the method further includes: The measured temperature values ​​of the specified area at the specified time are compared with the virtual temperature output by the virtual temperature field prediction model to determine the dynamic temperature difference compensation amount.

8. The method according to claim 6, characterized in that, The step of constructing a virtual temperature field at a specified time based on the virtual temperatures of different areas inside the vehicle at the specified time includes: Based on the prediction confidence of the virtual temperature field prediction model for a specified area, when the prediction confidence is greater than a first threshold, the virtual temperature of the specified area output by the virtual temperature field prediction model is used to construct the virtual temperature field. When the prediction confidence of the specified region is greater than the second threshold and less than the first threshold, the virtual temperature field is constructed using the corrected virtual temperature. When the prediction confidence level of the specified region is less than the second threshold, a downgrade control is performed.

9. A device for constructing a virtual temperature field inside a vehicle, characterized in that, include Acquire air conditioning execution status signals, environmental auxiliary signals, and cabin status signals at multiple times within a preset time window prior to a specified time. For a specific area inside the vehicle, a thermal boundary condition vector for the specified area at multiple times is constructed based on the air conditioning execution status signal, environmental auxiliary signal, and cabin status signal of the specified area at multiple times. The thermal boundary condition vectors at multiple times are used as input to a trained virtual temperature field prediction model, which outputs the virtual temperature of the specified region at the specified time. Based on the virtual temperatures of different areas inside the vehicle at the specified time, a virtual temperature field is constructed for the specified time.

10. 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 constructing a virtual temperature field inside a vehicle as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The device stores computer-executable instructions, which, when executed by a processor, implement the method for constructing a virtual temperature field inside a vehicle as described in any one of claims 1 to 8.

12. 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 constructing a virtual temperature field inside the vehicle as described in any one of claims 1 to 8.

Images