Air conditioner seat collaborative cooling control method based on neural network

Abstract

The invention is suitable for the technical field of automobile intelligent control, and provides an air conditioner seat collaborative cooling control method based on a neural network, which comprises the following steps of: constructing a human body-environment comfortable temperature original index model by collecting a large amount of human body and environment temperature distribution data in advance; on the basis, data acquisition and model comparison are carried out on the real-time human body temperature, the environment temperature and the temperature in a passenger compartment before the passenger gets on the bus, so that the heat comfort in the bus is pre-adjusted most reasonably before the passenger gets on the bus, and a method for cooperatively controlling ventilation and heating of an air conditioner and a seat is utilized; the two are combined to improve the thermal comfort of the passenger and the passenger compartment. And meanwhile, real-time sensing data is quickly processed when a passenger feels uncomfortable in temperature during riding and adjustment is carried out, the body temperature of the passenger and the ambient temperature of a passenger compartment are recorded and are input into a model for correction, so that the body comfort requirement of the passenger is met in a more personalized manner, and the comfort of the passenger in the vehicle running process is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of automobile intelligent control, and in particular relates to a coordinated cooling control method of air-conditioning seats based on a neural network. Background Art

[0002] With the development of science and technology, cockpit intelligence, automation and personalization have become important development directions of the automotive industry, and people's requirements for comfort are also increasing day by day. Among them, thermal comfort, as an important part, plays a key role in improving people's cockpit environment and driving experience.

[0003] The traditional air conditioning temperature adjustment method or seat ventilation and heating adjustment is limited to active human operation, and the independent operating systems of the two make it very inconvenient for people to adjust thermal comfort. It is difficult for people to operate both of them in a timely manner in the face of complex driving environments to best meet their own comfort requirements. This is very unfavorable to the development of intelligent cockpits and convenience and safety during driving. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a method for coordinated cooling control of air-conditioned seats based on a neural network, aiming to solve the problems raised in the above-mentioned background technology.

[0005] The embodiment of the present invention is implemented as follows: a neural network-based air-conditioning seat coordinated cooling control method comprises the following steps:

[0006] Step 1: In the data model preparation stage, a large amount of human body comfortable temperature distribution data and corresponding ambient temperature data are collected and preprocessed. The human body temperature distribution data set and the corresponding ambient temperature data set are used to build the human body-environment comfort temperature original index model in combination with the neural network algorithm.

[0007] A large amount of human body temperature and corresponding environmental temperature data is collected through infrared thermal imagers, temperature sensors and other detection and sensing equipment, and effective data preprocessing is performed to obtain high-quality training data. Data preprocessing includes data cleaning, data conversion, data sorting, processing unbalanced data, data verification and quality inspection to ensure model performance and accuracy, so as to obtain the best human-environment temperature comfort index model for people to feel comfortable.

[0008] Step 2: In the practical operation stage, after deploying the human-environment comfort temperature index model, the infrared thermal imager and the in-vehicle temperature sensor are used to obtain the corresponding data, including the thermal image of the occupant, the temperature of the surrounding environment and the temperature data of the passenger compartment. The input data is used to form a human-environment temperature data map using a neural network, and compared with the original human-environment comfort temperature index model. The difference between the two data is used to preliminarily preset the ventilation and heating control strategy method of the air conditioner-seat;

[0009] Step 3: During the adaptive learning stage, when the occupants adjust the air-conditioning temperature, air volume, air outlet angle, or any gear of seat ventilation and heating due to temperature discomfort, the occupant thermal imaging and cabin temperature data are collected in real time and uploaded to the controller to modify the original model of the human body-environment comfort temperature index using a neural network, and the updated human body-environment comfort index model is recorded to achieve personalized adaptive learning to improve the adjustment of occupant thermal comfort.

[0010] A further technical solution is that in step 1, when collecting data, three wind measuring points are arranged on the head, trunk and lower limbs respectively, and the human body is divided into three parts: head and neck, trunk and lower limbs. The average values ​​of the transient temperatures of these three parts and the average value of the passenger compartment temperature are obtained as data sets, and the collected human body temperature partition data and the corresponding passenger compartment ambient temperature data are substituted into the PMV calculation formula, and the calculation formula is as follows:

[0011] PMV = [0.303e -0.036M +0.0275]TL

[0012] TL=M0-W0-3.05[5.73-0.07(MW)-P a ]-0.0014M(34-t a )

[0013] -0.0173M(5.87-P a )-0.42(MW-58)

[0014] -3.96×10 -8 f cl [(t cl +273) 4 -(t r +273) 4 ]-f cl h c (t cl -t a )

[0015] Where M is the energy metabolism rate of the human body, TL is the difference between the heat produced by the human body and the heat dissipated by the human body to the outside world, M0 is the human body metabolic rate under the state of thermoneutral thermal equilibrium, W0 is the mechanical work done by the human body to the outside world under the state of thermoneutral thermal equilibrium, P a is the partial pressure of water vapor around the human body, t a is the air temperature around the human body, W is the mechanical work done by the human body, f cl is the area coefficient of clothing, t cl is the outer surface temperature of the clothes, t r is the mean radiation temperature, h c is the convective heat transfer coefficient, t a is the air temperature around the human body;

[0016] The obtained average temperatures of the three parts, the average temperature of the passenger compartment and the calculated corresponding PMV values ​​are respectively input into the neural network for training, and finally the human body-environment temperature comfort index model is obtained.

[0017] A further technical solution is that in step 2, the control objects of the air-conditioning-seat ventilation and heating control strategy method include the automobile air-conditioning and the seat ventilation and heating functions, and the two are collaboratively controlled according to the set control strategy; the control strategy method is specifically as follows:

[0018] The three parts of human-environment comfort temperature index data obtained are input into the cockpit controller for comparison. The part with the worst thermal comfort will be adjusted with priority to air conditioning and dual adjustment of seat ventilation and heating, while the rest will be adjusted with seat ventilation and heating for thermal comfort.

[0019] A further technical solution is that in step 3, when the original model of the human-environment comfort temperature index is corrected by using a neural network, the model can be improved according to the occupant's operating behavior under different driving environments and driver behaviors, so as to automatically adjust and continuously update the structure and parameters of the neural network model to adapt to the complex and changeable driving environment, temperature and occupant's physical condition, thereby continuously improving the performance and accuracy of the model.

[0020] The embodiment of the present invention provides a method for coordinated cooling control of air-conditioning seats based on neural networks. The method constructs a human-environment comfort temperature original index model by pre-collecting a large amount of human body and environmental temperature distribution data. On this basis, data collection and model comparison are performed for the real-time human body temperature, environmental temperature and passenger cabin temperature before the occupant gets on the vehicle, so as to achieve the most reasonable pre-adjustment of the thermal comfort in the vehicle before the occupant gets on the vehicle, and the method for coordinated control of air conditioning and seat ventilation and heating is used so that the two are jointly committed to improving the thermal comfort of the occupant and the passenger cabin. At the same time, when the occupant feels uncomfortable with the temperature and makes adjustments during riding, the real-time sensor data is quickly processed, the occupant's human body temperature and the passenger cabin ambient temperature are recorded, and the input model is corrected to achieve a more personalized fit for the occupant's physical comfort needs, so that the cabin thermal comfort adjustment can respond to environmental changes and passenger needs more quickly, and the personalized air-conditioning seat coordinated heating and cooling system design method based on neural networks can optimize energy utilization and operation efficiency, reduce operating costs, and realize automation, intelligence, and personalized settings, thereby improving the comfort of the occupant during vehicle operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of a data preparation stage in a neural network-based air-conditioning seat coordinated cooling control method provided by an embodiment of the present invention;

[0022] Figure 2 A schematic diagram of a practical operation phase of a neural network-based air-conditioning seat coordinated cooling control method provided by an embodiment of the present invention;

[0023] Figure 3 A schematic diagram of an adaptive learning phase in a neural network-based air-conditioning seat coordinated cooling control method provided by an embodiment of the present invention;

[0024] Figure 4 It is a schematic diagram of the three-dimensional model of the passenger compartment, seats and monitoring dummy;

[0025] Figure 5 The temperature distribution of the passenger cabin and human body after the ventilation and cooling simulation is completed;

[0026] Figure 6 It is the PMV change data of the head and neck;

[0027] Figure 7 This is the PMV change curve of the trunk;

[0028] Figure 8 This is the PMV change curve of the lower limbs;

[0029] Fig. 9 is the human-environment comfort temperature index model obtained after fitting;

[0030] Fig.10 It is the comfort index corresponding to the three groups of zones. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0032] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0033] like Figure 1-3 As shown, a neural network-based air-conditioning seat coordinated cooling control method provided by an embodiment of the present invention includes the following steps:

[0034] Step 1: In the data model preparation stage, a large amount of human body comfortable temperature distribution data and corresponding ambient temperature data are collected and preprocessed. The human body temperature distribution data set and the corresponding ambient temperature data set are used to build the human body-environment comfort temperature original index model in combination with the neural network algorithm.

[0035] A large amount of human body temperature and corresponding environmental temperature data is collected through infrared thermal imagers, temperature sensors and other detection and sensing equipment, and effective data preprocessing is performed to obtain high-quality training data. Data preprocessing includes data cleaning, data conversion, data sorting, processing unbalanced data, data verification and quality inspection to ensure model performance and accuracy, so as to obtain the best human-environment temperature comfort index model for people to feel comfortable.

[0036] Step 2: In the practical operation stage, after deploying the human-environment comfort temperature index model, the infrared thermal imager and the in-vehicle temperature sensor are used to obtain the corresponding data, including the thermal image of the occupant, the temperature of the surrounding environment and the temperature data of the passenger compartment. The input data is used to form a human-environment temperature data map using a neural network, and compared with the original human-environment comfort temperature index model. The difference between the two data is used to preliminarily preset the ventilation and heating control strategy method of the air conditioner-seat;

[0037] Step 3: During the adaptive learning stage, when the occupants adjust the air-conditioning temperature, air volume, air outlet angle, or any gear of seat ventilation and heating due to temperature discomfort, the occupant thermal imaging and cabin temperature data are collected in real time and uploaded to the controller to modify the original model of the human body-environment comfort temperature index using a neural network, and the updated human body-environment comfort index model is recorded to achieve personalized adaptive learning to improve the adjustment of occupant thermal comfort.

[0038] In the embodiment of the present invention, a large amount of human body comfortable temperature distribution data and corresponding environmental temperature data are collected and preprocessed in the model data preparation stage. The human body temperature distribution data set and the corresponding environmental temperature data set after preprocessing are used to build a human body-environment comfortable temperature original index model in combination with a neural network algorithm. The neural network can specifically adopt a product neural network or a recurrent neural network.

[0039] After deploying the human-environment comfort temperature index model, in the practical operation stage, the corresponding data are obtained through infrared thermal imagers and in-vehicle temperature sensors, including occupant thermal imaging images, human surrounding temperature images and in-vehicle passenger compartment temperature data. The input data constitutes a human-environment temperature data map, which is compared with the original human-environment comfort temperature index model. The difference between the two data is used to preliminarily preset the ventilation and heating control strategy method of the air conditioner-seat.

[0040] During the adaptive learning stage, when the occupants adjust the air-conditioning temperature, air volume, air outlet angle, or any gear of seat ventilation and heating due to temperature discomfort, the occupants' thermal imaging and cabin temperature data are collected in real time and uploaded to the controller, where the original model of the human body-environment comfort temperature index is corrected using a neural network. The updated human body-environment temperature comfort index model is recorded to achieve personalized adaptive learning to improve the adjustment of the occupants' thermal comfort.

[0041] As a preferred embodiment of the present invention, in step 1, by establishing a three-dimensional model of the passenger compartment, seats and monitoring dummy, starCCM+ is used to perform a ventilation and cooling simulation test of the passenger compartment, and the feasibility is verified by taking summer working conditions as an example. The initial simulation temperature is set to 40°C. Figure 4 As shown in the figure, three wind measurement points are arranged on the head, torso and lower limbs, the simulation step is set to 1s, the simulation time is 100s, the human body is divided into three parts: head and neck, torso and lower limbs, and the average values ​​of transient simulation temperatures of these three parts and the average value of passenger compartment temperature are obtained as data sets. The temperature distribution of the passenger compartment and the human body after the ventilation and cooling simulation is shown in the figure below. Figure 5 shown.

[0042] After the simulation, a total of 100 sets of human body temperature partition data and corresponding cabin ambient temperature data were obtained, which were used as data to substitute into the PMV calculation formula. The PMV evaluation index is the expected average thermal sensation, which represents the thermal sensation of most people in the current environment and is used to evaluate the comfort of the human body under different environmental conditions and different activity levels. PMV takes into account factors such as activity level, clothing thermal resistance, air temperature, relative humidity, average radiation temperature and gas flow rate, and its calculation formula is as follows:

[0043] PMV = [0.303e -0.036M +0.0275]TL

[0044] TL=M0-W0-3.05[5.73-0.07(MW)-P a ]-0.0014M(34-t a )

[0045] -0.0173M(5.87-P a )-0.42(MW-58)

[0046] -3.96×10 -8 f cl [(t cl +273) 4 -(t r +273) 4 ]-f cl h c (t cl -t a )

[0047] Where M is the energy metabolism rate of the human body, TL is the difference between the heat produced by the human body and the heat dissipated by the human body to the outside world, M0 is the human body metabolic rate under the state of thermoneutral thermal equilibrium, W0 is the mechanical work done by the human body to the outside world under the state of thermoneutral thermal equilibrium, P a is the partial pressure of water vapor around the human body, t a is the air temperature around the human body, W is the mechanical work done by the human body, f cl is the area coefficient of clothing, t cl is the outer surface temperature of the clothes, t r is the mean radiation temperature, h c is the convective heat transfer coefficient, t a is the temperature of the air surrounding the human body.

[0048] Finally, the average temperatures of the three parts (head and neck, trunk, and lower limbs), the average temperature of the passenger compartment, and the corresponding PMV values ​​obtained were input into the neural network for training, and independent neural network fitting models were established. The PMV change data are as follows: Figure 6-Figure 8 As shown. This is only used as a method verification, so this example uses a simple feedforward neural network as an example for fitting, and the final model is imported into Simulink, as shown Fig. 9 shown.

[0049] The input part 1 is the average temperature of the head and neck, 2 is the average temperature of the torso, and 3 is the average temperature of the lower limbs. The output part 1 is the PMV of the head and neck, 2 is the PMV of the torso, and 3 is the PMV of the lower limbs. The last output box (blue text) is the average value of the PMV of the three parts, which is the PMV index of the overall comfort of the human body.

[0050] As a preferred embodiment of the present invention, in step 2, taking the data obtained in a simulation process as an example, the average head and neck temperature is 35.1°C, the trunk temperature is 37.5°C, and the lower limb temperature is 39.3°C. The three groups of data are input as the test group into the established human-environment comfort temperature index model, and the corresponding comfort indexes output by the three models can be obtained, such as Fig.10 When the index is 0, it means that the human body area is comfortable. Therefore, the corresponding index of the head and neck area is 3.682, the corresponding index of the trunk area is 3.662, and the corresponding index of the lower limbs is 4.931.

[0051] As a preferred embodiment of the present invention, in step 2, the control objects of the air-conditioning-seat ventilation and heating control strategy method include the automobile air-conditioning and the seat ventilation and heating functions, which are collaboratively controlled according to the set control strategy; the control strategy method is specifically as follows:

[0052] The three parts of human-environment comfort temperature index data obtained are input into the cockpit controller for comparison. The part with the worst thermal comfort will be adjusted with priority to air conditioning and dual adjustment of seat ventilation and heating, while the rest will be adjusted with seat ventilation and heating for thermal comfort.

[0053] In the embodiment of the present invention, it can be seen from the previous simulation data that the thermal comfort of the lower limbs is the worst, followed by the head and neck, and finally the torso. Therefore, based on this situation, the air conditioner will automatically adjust the horizontal air outlet angle downward to blow directly on the lower limbs of the human body, and increase the air volume, and turn on the seat ventilation of the leg rest part and increase the gear to increase the air volume, so as to accelerate the adjustment of the thermal comfort of the lower limbs. The head is adjusted by opening the seat ventilation function of the seat headrest vents, and the torso is adjusted by opening the seat ventilation function of the seat back vents.

[0054] As a preferred embodiment of the present invention, in step 3, when the original model of the human-environment comfort temperature index is corrected by using a neural network, the model can be improved according to different driving environments and driver behaviors, and according to the operating behaviors of the occupants, so as to automatically adjust and continuously update the structure and parameters of the neural network model, so as to adapt to the complex and changeable driving environment, temperature and occupant physical condition, thereby continuously improving the performance and accuracy of the model.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A neural network-based air-conditioning seat coordinated cooling control method, characterized in that: The following steps are involved: Step 1: In the data model preparation stage, a large amount of human body comfortable temperature distribution data and corresponding ambient temperature data are collected and preprocessed. The human body temperature distribution data set and the corresponding ambient temperature data set are used to build the human body-environment comfort temperature original index model in combination with the neural network algorithm. Step 2: In the practical operation stage, after deploying the human-environment comfort temperature index model, the infrared thermal imager and the in-vehicle temperature sensor are used to obtain the corresponding data, including the thermal image of the occupant, the temperature of the surrounding environment and the temperature data of the passenger compartment. The input data is used to form a human-environment temperature data map using a neural network, and compared with the original human-environment comfort temperature index model. The difference between the two data is used to preliminarily preset the ventilation and heating control strategy method of the air conditioner-seat; Step 3: During the adaptive learning stage, when the occupants adjust the air-conditioning temperature, air volume, air outlet angle, or any gear of seat ventilation and heating due to temperature discomfort, the occupant thermal imaging and cabin temperature data are collected in real time and uploaded to the controller to modify the original model of the human body-environment comfort temperature index using a neural network, and the updated human body-environment comfort index model is recorded to achieve personalized adaptive learning to improve the adjustment of occupant thermal comfort.

2. The neural network-based air-conditioning seat coordinated cooling control method according to claim 1, characterized in that: In step 1, when collecting data, three wind measuring points are arranged on the head, trunk and lower limbs respectively, and the human body is divided into three parts: head and neck, trunk and lower limbs. The average values ​​of the transient temperatures of these three parts and the average value of the passenger compartment temperature are obtained as data sets, and the collected human body temperature partition data and the corresponding passenger compartment ambient temperature data are substituted into the PMV calculation formula, which is as follows: <h2 style=";text-align:left;direction:ltr">PMV=[0.303e<h2 style=";text-align:left;direction:ltr"> -0.036M <h2 style=";text-align:left;direction:ltr"> +0.0275]TL TL=M0-W0-3.05[5.73-0.07(M-W)-P a ]-0.0014M(34-t a ) -0.0173M(5.87-P a )-0.42(M-W-58) -3.96×10 -8 f cl [(t cl +273) 4 -(t r +273) 4 ]-f cl h c (t cl -t a ) Where M is the energy metabolism rate of the human body, TL is the difference between the heat produced by the human body and the heat dissipated by the human body to the outside world, M0 is the human body metabolic rate under the state of thermoneutral thermal equilibrium, W0 is the mechanical work done by the human body to the outside world under the state of thermoneutral thermal equilibrium, P a is the partial pressure of water vapor around the human body, t a is the air temperature around the human body, W is the mechanical work done by the human body, f cl is the area coefficient of clothing, t cl is the outer surface temperature of the clothes, t r is the mean radiation temperature, h c is the convective heat transfer coefficient, t a is the air temperature around the human body; The obtained average temperatures of the three parts, the average temperature of the passenger compartment and the calculated corresponding PMV values ​​are respectively input into the neural network for training, and finally the human body-environment temperature comfort index model is obtained.

3. The neural network-based air-conditioning seat coordinated cooling control method according to claim 2 is characterized in that: In step 1, human body temperature and corresponding ambient temperature data are collected by using an infrared thermal imager and a temperature sensor; Data preprocessing includes data cleaning, data conversion, data organization, processing of unbalanced data, data validation and quality checking.

4. The neural network-based air-conditioning seat coordinated cooling control method according to claim 2, characterized in that: In step 2, the control objects of the air-conditioning-seat ventilation and heating control strategy method include the automobile air-conditioning and seat ventilation and heating functions, and the two are coordinated and controlled according to the set control strategy; the control strategy method is specifically as follows: The three parts of human-environment comfort temperature index data obtained are input into the cockpit controller for comparison. The part with the worst thermal comfort will be adjusted with priority to air conditioning and dual adjustment of seat ventilation and heating, while the rest will be adjusted with seat ventilation and heating for thermal comfort.

5. The neural network-based air-conditioning seat coordinated cooling control method according to claim 4 is characterized in that: In step 3, when the original model of the human-environment comfort temperature index is corrected by using a neural network, the model is improved according to the occupant's operating behavior under different driving environments and driver behaviors, so as to automatically adjust and continuously update the structure and parameters of the neural network model to adapt to different driving environments, temperatures and occupant physical conditions.

Images