JPH04148142A - Air conditioning control device - Google Patents

Abstract

PURPOSE:To enable a temperature to be rapidly set to a target temperature feeling by a method wherein a future temperature feeling is estimated on the basis of a skin temperature of an occupant in a period ranging from the past to the present time and then an air conditioning control is carried out. CONSTITUTION:A skin temperature of an occupant got through an infra-red ray thermometer acting as a temperature feeling information sensing means 1 is inputted to a temperature feeling estimating means 2. As this input, the present skin temperature and some skin temperatures measured for every 30 seconds before 1 minute and 30 seconds are applied and then a temperature feeling after one minute is estimated with a neural network (a temperature feeling estimating means 2). Then, the estimated temperature feeling and the stored temperature feeling stored by a control pattern memory means 3 at the same time as that of the estimated temperature are compared to each other by a controlling amount determining means 4 to get their deviation. In the event that the estimated temperature feeling is displaced toward a colder temperature, the heating intensity during a heating season is increased and then a cooling capability during a cooling season is decreased to cause such a controlling amount to be transmitted to an air conditioning means 5. However, a control for 30 seconds after starting the air conditioning operation is carried out to keep a room temperature of 25 deg.C and this temperature feeling estimating means is utilized after 30 seconds of starting of the air conditioning operation and then the controlling operation is carried out.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an air conditioning control device, and more particularly to a vehicle air conditioning control device suitable for controlling air conditioning inside a vehicle.

(Prior art) Conventionally, in this type of air conditioning control device,
As disclosed in Japanese Patent No. 9713, there is known an air conditioning control device that estimates the occupant's thermal sensation at a given time based on the occupant's skin temperature and controls air conditioning based on the thermal sensation.

(Problem to be solved by the invention) However, in conventional air conditioning control devices, the thermal sensation of the occupant is estimated at the time when the occupant's skin temperature is measured. can be estimated, but future changes in thermal sensation cannot be predicted. Therefore,
Air conditioning may not be installed properly, resulting in inconveniences such as being too cold or too warm.

(Means for Solving the Problems) In order to solve the above problems, the air conditioning control device of the present invention includes:
A thermal sensation information detecting means for detecting information regarding a person's thermal sensation in a room, and a thermal sensation for predicting a thermal sensation in the near future based on the history of thermal information within a predetermined time output from the thermal sensation information detecting means. a prediction means, a control pattern storage means for storing a control pattern of a predetermined thermal sensation in order to reach a target thermal sensation, a control pattern of a thermal sensation stored in the storage means, and the thermal sensation prediction means. control amount determining means for determining control amounts of temperature-controlled air such as air volume and temperature so that the near future thermal sensation output from the The room is characterized by comprising an air conditioning control means for controlling the temperature of the temperature-conditioned air blown into the room, and a blower for controlling the volume of the temperature-conditioned air blown into the room based on the control amount output from the control amount determining means. .

(Function) The air conditioning control device of the present invention configured as described above detects information regarding the thermal sensation of a person indoors by the thermal sensation information detection means, and detects information regarding the thermal sensation of a person in the room within a predetermined period of time output from the thermal sensation information detection means. A storage means for predicting a near future thermal sensation by a thermal sensation prediction means based on the history of thermal sensation information and storing a predetermined thermal sensation control pattern in order to reach a target thermal sensation while maintaining a comfortable state. determining control amounts of temperature-controlled air such as air volume, temperature, etc. by a control amount determining means so that the temperature sensation control pattern stored in the thermal sensation control pattern and the thermal sensation output from the thermal sensation prediction means in the near future match; The temperature of the temperature-conditioned air to be blown into the room is controlled by the air conditioning control means based on the control amount output from the control amount determining means, and the temperature-conditioned air whose temperature has been adjusted by the temperature control means is controlled by the control amount. The blower blows air into the room based on the control amount determined by the determining means.

(Effects) In the air conditioning control device of the present invention, information regarding a large temperature sensation in a room is detected by a thermal sensation information detection means, and a history of thermal sensation information within a predetermined period of time outputted from the thermal sensation information detection means is used. Based on this, it is possible to predict the thermal sensation in the near future using the thermal sensation prediction means, so it is possible to predict changes in the thermal sensation prior to the point of time when the thermal sensation information is detected, and to accurately set the desired control pattern for the thermal sensation. It will come true.

Further, the thermal sensation prediction means includes thermal sensation information at the reference time detected by the thermal sensation information detection means and within a predetermined time period before that time, and an output value representing the thermal sensation after the reference time (in the near future). Using , the input of the thermal sensation information and the output value representing the thermal sensation in the near future are multiplied by the thermal sensation information by a weighting constant adjusted in advance to correspond to the output value, and then added. It is a calculation device configured by combining a large number of calculation elements that perform non-linear transformation and obtain an output, and when predicting the thermal sensation in the near future, it uses thermal sensation information within a predetermined time output from the thermal sensation information detection means and adjustment. The temperature sensation in the near future can be predicted based on the magnitude of the final output value obtained by multiplying and adding the weight constants obtained through non-linear transformation.

(Example) First example (composition) The first embodiment will be explained based on FIG. 14.

The infrared thermometer 116 that detects the skin temperature of the face as the thermal information detection means 1 is an infrared detection type sensor that detects the skin temperature of the passenger's face in a non-contact manner. Air conditioning unit 92
A control device 111 that controls the is comprised of elements surrounded by broken lines in FIG. Among the elements constituting this control device 111, the control pattern storage means 3, the control amount determining means 4, and the thermal sensation prediction means 2 are composed of a microcomputer, and the thermal sensation prediction means 2 is composed of a neural network.

The thermal sensation prediction means 2 is a neural network as shown in FIG. 2, and has an input layer that predicts the thermal sensation together with the skin temperature for 1 minute and 30 seconds every 30 seconds for the past few minutes, and the current skin temperature. is input. The neural network that is the thermal sensation prediction means of this embodiment is composed of three layers: an input layer 11, an intermediate layer 12, and an output layer 13. The input layer 11 has 4 elements, the intermediate layer 12 has 4 elements, and the output layer has 4 elements. 13 is composed of one element. Each element has nonlinear input and output, as shown in FIG. The relational expression is: is a weight, θ is a threshold value for each element, and Σ is a summation for all input xi.

However, only the input layer was made to have an identity change of Y=X.

The output layer 13 shown in FIG. 2 outputs a value corresponding to the future thermal sensation.

The threshold value of each element of this thermal sensation prediction neural network and the weight between elements are determined through prior training.

In this education, experiments are conducted with several male and female subjects under various air conditioning conditions, and the skin temperature and thermal sensation at that time are obtained as educational data. Based on this educational data, threshold values and weights are calculated in steps as shown in FIG.

Step 1: In Sl of FIG. 4, skin temperature history data of educational data is inputted from the input layer. In advance, the threshold values of the elements and the weights between the elements are given by random numbers, and in this state, calculations are performed in the intermediate layer 12 and the output layer 13 to calculate a predicted future thermal sensation.

Step 2: At 82 in FIG. 4, the difference (error) E between the warm sensation Y calculated by Sl and the actually declared warm sensation is calculated using the following formula.

E=0.5*(Y-D) 2 Step 3: At 83 in FIG. 4, if the error in S2 is sufficiently small, the weight and threshold at that time are stored in the memory and the process ends. If it is large, the weight change amount is calculated in S4 and thereafter.

Step 4: At 84 in FIG. 4, weights and threshold correction amounts are calculated based on the following equations.

aE/ciY=Y-D The threshold correction amount of the output layer 13 is aE/aθ=aE/aY-δY/aθ =-BE/aY-Y (1-Y) Next, aE/aX=aE/aY -BY/aX aE/aY-Y (1-Y) Weight between the i-th element of the intermediate layer 12 and the output layer 13 (W
The amount of correction for i OUT ) is a E/aW i O[JT = aE/aX-aX/aWiOUT = aE/aX''3'i where yi is the i-th output value of the intermediate layer 12. , the contribution rate of the i-th output value yi of the intermediate layer 12 to the error is BE/ay 1=aE/aX-aX/ay 1=aE/
The i-th threshold θi of the aX-WiOUT intermediate layer 12 or the contribution rate to the error is BE/aθ1 = aE/ay 1-ay i/aθ1 - aE/ay i -y i (1-y i) If the value input to the i-th element of the intermediate layer 12 is xi, then aE/axi 20E/ay 1-ay i/ax iaE/ayi-
yi (1-yi) The contribution rate of the weight WiJ between the j-th element of the input layer 11 and the i-th element of the intermediate layer 12 to the error is aE/δWiJ = aE/axi-ax i/aWij = aE/δX1 −yj Here, yj′ is the j-th output of the input layer 11. The threshold value of each element and the correction amount of the weight between elements are calculated using the above formula.

Step 5: At 85 in FIG. 4, aE/a obtained in S4
Using W, BE/aθ, etc., the correction amount is calculated by ΔW (t) - ε·δE/aW+αΔW (t−1). Here, ΔW(t-1) is the amount of correction from the previous correction, and ε and α are constants. The actual weight is modified as follows: W=W+ΔW.

Step 6: At 86 in FIG. 4, after the correction in S5,
Perform the same calculation as l and return to S2.

Steps 1 to 6 are repeated, and when the error becomes sufficiently small, the weight, threshold, or thermal sensation prediction neural network is set.

Next, the control pattern storage means 3 predetermines the temporal change pattern of the target temperature sensation, for example, as shown in FIG.
When creating a state of insensitivity that is neither hot nor cold, Figure 5 (b
As shown in FIG. 1, there are cases in which the state is gradually brought closer to the insensible state from the beginning, and there are also cases in which the state is maintained in the insensitive state from the beginning, as shown in FIG. 5(C).

Similarly, several types of heating agents can be considered.

In this first embodiment, the insensitive state is maintained from the beginning as shown in FIG.
), or any other temperature control pattern including the patterns listed above may be used. It is also possible to store several types of control patterns in the control pattern storage means and to allow the occupant to select the most suitable one according to his or her preference.

The control amount determination means 4 calculates the deviation between the temperature sensation predicted by the temperature sensation prediction means after one minute and the temperature sensation at the same time determined from the temperature sensation control pattern stored in advance in the control pattern storage means 3. The amount of change in wind temperature, wind speed, etc. is calculated based on the deviation.

The air conditioning unit 92 has a generally known configuration, and includes an inside/outside air switching device 101, a blower 102, an evaporator 03, a heater core 104, and an air mix damper 105.
, and a vent outlet 106.

The air conditioning control means 5 controls the temperature of the temperature-controlled air blown into the room based on the control amount output from the control amount determining means 4 using the evaporator 103, the heater core 104, and the air mix damper 105.

Further, the blower 102 is a blade-shaped blower having a large number of blades on the outer peripheral side, and is provided near the air intake of the air conditioning unit 92, and controls the temperature-conditioned air adjusted by the air conditioning control means 5 to determine the control amount. Air is blown into the vehicle compartment 91 according to the air volume determined by means 4.

(Operation) The operation of the first embodiment will be explained below. An infrared thermometer as the thermal information detection means 1 measures thermal information of the occupant, for example, the facial skin temperature of the driver. The skin temperature of the occupant obtained by this infrared thermometer is input to the thermal sensation prediction means 2.

As the second input, the current skin temperature and the skin temperature measured every 30 seconds up to 1 minute and 30 seconds before are used to predict the thermal sensation one minute later using a neural network (thermal sensation prediction means 2). Next predicted thermal sensation and control pattern storage means 3 at the same time
The control amount determination means 4 compares the temperature sensation stored in advance with the temperature sensation, calculates the deviation, and if the predicted temperature sensation deviates to the cold side, if it is a heating agent, increase the heating intensity, and if it is a cooling period, increase the heating intensity. A control amount such as lowering the cooling capacity is transmitted to the air conditioning control means 5. However, since there is little data on changes in skin temperature in the early stages of air conditioning, the accuracy of predicting thermal sensations deteriorates. Therefore, the control for 30 seconds after the start of air conditioning, when the necessary data for the thermal sensation prediction means is available, is performed to maintain the car room temperature at 25 degrees, and from 30 seconds after the start of air conditioning, the temperature sensation prediction means is used. Take control.

FIG. 15 shows the thermal sensation predicted by the thermal sensation prediction means of the first embodiment, which was conducted in a laboratory under the conditions of an outside temperature of 35°C, an initial car temperature of 60°C, and a solar radiation of 900 W/m2-h. , the thermal sensation estimated using a conventional regression equation that uses the measured skin temperature and the rate of change of the skin temperature as variables to estimate the subject's thermal sensation at the time of skin temperature measurement, and the actual thermal sensation of the subject. This is the result of comparing the declared value.

The conventional method uses the current skin temperature and its rate of change,
Since it is only an estimate of the thermal sensation, the value matches well with the reported value reported by the subject at the start of air conditioning when the rate of change in skin temperature is large, but it soon deviates from the reported value. Then, in the latter half of the air conditioning period, the value approaches the declared value again. This is because human sensations are not determined at any given moment, but are also influenced by the history of changes over time. For this reason, in the conventional method that does not take past skin temperature history into consideration, the accuracy of estimating thermal sensation deteriorates in the early stages of air conditioning when the skin temperature changes are rapid and the influence of the temporal history of the skin temperature change is large. . In addition, because the conventional method estimates the thermal sensation based on the skin temperature measured at the time of estimating the thermal sensation, it is strongly affected by the measurement accuracy of the skin temperature, and the thermal sensation may be affected by variations in the measured skin temperature values. Variations in estimated values may occur.

On the other hand, the actual thermal sensation prediction means uses the history of skin temperature as input data. Therefore, as shown in Figure 15, when there is a sudden change in skin temperature as seen at the start of air conditioning (I) and at the beginning of air conditioning (■), in other words, when the effect of the temporal history of skin temperature changes greatly affects the occupant's sense of warmth. It is also possible to predict the temperature sensation with high accuracy. In addition, by using the history of skin temperature, it is possible to take into account whether the skin temperature is increasing or decreasing, and also consider how the rate of change is changing, so it is possible to accurately predict future thermal sensations. Can be predicted well.

Based on the temperature sensation prediction results as described above, the control amount determination means 4 determines the control amount, and according to this control amount, the air conditioning control means 5 determines the opening degree of the air mix damper 105, the evaporator 103, and the heater core 104 shown in FIG. Change the set temperature. The warm air adjusted by the air conditioning control means 5 is blown into the vehicle compartment 91 with the wind speed adjusted by the blower 102 shown in FIG. 14 according to the output of the control amount determining means.

(Effects) The air conditioning control device of the first embodiment is able to predict the future thermal sensation based on the skin temperature of the occupants from the past to the present and adjust the air temperature and speed to an appropriate level in advance. air conditioning control. Therefore, compared to the conventional technology in which the current thermal sensation is estimated and control is performed based only on the estimated current thermal sensation, the first embodiment of the present invention can quickly approach the target thermal sensation. .

Second Embodiment (Structure) In addition to the first embodiment, the second embodiment of the present invention, as shown in the part enclosed by the broken line in FIG. The history of the thermal sensation information outputted from the thermal sensation operation button 114 (FIG. 14) as the input means 27 and the thermal sensation information detection means 21, and the desired thermal sensation input from the thermal sensation operation button 114 are stored. A learning data storage means 28 and a thermal sensation prediction correction means 29 for changing the weights and thresholds of the neural network constituting the thermal sensation prediction means 22 based on the learning data stored in the learning data storage means 28 are provided. It is characterized by: All of these are built into a control device 111 (FIG. 14) consisting of a microcomputer.

(Function) Regarding the function of the second embodiment, mainly the differences from the first embodiment will be explained below.

The characteristic feature of the second embodiment added in addition to the first embodiment is a thermal sensation operation button 114 (shown in FIG. 6) that allows correction of differences in thermal sensation among individuals.
Desired thermal sensation input means 27).

The sensation of warmth differs slightly depending on the person, but in this second example,
Since the occupant's desired thermal sensation can be input using the thermal sensation operation button, the weights and threshold values of the neural network constituting the thermal sensation prediction means 22 can be changed in accordance with such individual differences. This makes it possible to quickly create air-conditioned conditions that are more comfortable for passengers.

In the second embodiment, the thermal sensation predicted by the thermal sensation prediction means 22 based on the temperature history every 30 seconds for 1 minute and 30 seconds from the past to the present, and the temperature actually felt by the occupant are calculated. If the temperature and temperature do not match, the passenger operates the thermal sensation operation button 114. When an input is received from the thermal sensation operation button 114, the output from the thermal sensation operation button 114 (desired thermal sensation input means 27) is inputted directly to the control amount determining means 24 in order to quickly control the air conditioning. Thereby, the control amount determining means 24 determines an appropriate control amount and outputs the control amount to the air conditioning control means 25.

Based on this control amount, the air conditioning control means 25 performs air conditioning control in an amount corresponding to the temperature sensing operation button 114.

On the other hand, the skin temperature data used for the thermal sensation prediction and the signal input from the thermal sensation operation button 114 (desired thermal sensation input means 27) are stored in the learning data storage means 28.

The thermal sensation prediction correction means 29 calculates the thermal sensation prediction correction amount using the data stored in the learning data storage means when air conditioning in the vehicle interior is not required.

The method for calculating the thermal sensation prediction correction amount here is the same as the method used to determine the weights and thresholds in the first embodiment. Based on the results from the thermal sensation prediction correction means 29, the weights and threshold values of the neural network constituting the thermal sensation prediction means 22 are corrected.

FIG. 16 and FIG. 17 show the results of an experiment in which the effect of the thermal sensation operation button on improving the accuracy of thermal sensation prediction was confirmed. The initial environmental conditions for this experiment were: outside temperature 28°C, initial car room temperature 28°C.
The test was carried out at 1 °C without solar radiation.

FIG. 16 shows the state before correction. Except for the early stages of air conditioning, the predicted thermal sensation and the subjects' reported values deviate from the middle to late stages of air conditioning. At this point, the temperature sensing operation button was operated at the point indicated by the arrow in FIG.

After the above experiment was completed, the thermal sensation prediction means 22 was modified based on the input data from the thermal sensation operation button, and the modification effect was investigated. The results are shown in FIG. 17. It can be seen that the parts that deviated from the declared values from the middle to late stages of air conditioning have been corrected, and the accuracy of predicting thermal sensation has improved. Furthermore, it can be seen that at the start of air conditioning, for which the temperature sensation had already been predicted with sufficient accuracy before the correction, that accuracy was maintained even after the correction.

As described above, the second embodiment is equipped with a warm operation button,
Since the thermal sensation prediction means can be modified in accordance with the occupant's thermal sensation, it is possible to predict a thermal sensation that closely matches the occupant's thermal sensation.

In this case, the neural network constituting the thermal sensation prediction means 22 was modified based on the learning data, or the control pattern storage means 33 was modified using this learning data, as shown within the dashed line box in FIG. The control pattern may be modified. In other words, the neural network constituting the thermal sensation prediction means 32 is not modified, and only the pattern of the target thermal sensation is changed in accordance with the signal from the thermal sensation operation button 114.

FIGS. 18 and 19 show experimental results confirming the effect of improving the degree of achievement of the target temperature sensation when only the control pattern is modified as described above. In Figure 18, the air conditioner is
Control is performed to maintain target temperature sensation -2 (cool). However, in this experiment, the reported value of the test subject was -1 (slightly cool) for the feeling of warmth. Therefore, the subject inputted a signal to modify the target thermal sensation in a cooler direction using the thermal sensation operation button 114 at the position of the arrow in FIG.

This input signal and the target temperature sensation at the time when the signal is input are input into the learning data storage means 38, and based on the data, the control pattern correction means 39 calculates the deviation between the input signal and the target temperature sensation. , modify the control pattern stored in the control pattern storage means 33. FIG. 19 shows the results of an experiment in which the control pattern was modified in this way and the target temperature sensation was modified accordingly. By setting the target temperature sensation to -3 (slightly cold), it was possible to maintain the subject's reported value at -2 (cool), which is the original target temperature sensation.

As described above, by inputting the thermal sensation requested by the occupant using the thermal sensation operation button and modifying the target thermal sensation, it is now possible to provide appropriate air conditioning that takes into account individual differences in thermal sensation.

Regarding the above-mentioned modification of the control pattern, another method is to store several kinds of control pattern storage means in advance, as explained in the first embodiment using FIGS. The most suitable control pattern can be selected from among them according to the passenger's preference. Further, it is possible to use the stored thermal sensation control pattern by moving the pattern toward hotter or colder according to the passenger's preference based on human power from the thermal sensation operation button 114, and modifying the pattern. .

(Effects) In addition to the effects of the first embodiment, the present second embodiment has the following advantages: Since it has the ability to make adjustments according to personal preference, air conditioning control can be performed that takes into account individual differences in the thermal sensations of occupants.

Third Embodiment (Structure) A third embodiment of the present invention is characterized in that a vehicle room temperature detection means 50 is added to the second embodiment as shown in FIG.

As the vehicle room temperature detection means 50, a vehicle room temperature sensor 115 (FIG. 14), which is provided in the center console near the driver's feet and is composed of, for example, a thermocouple, can be considered. In the third embodiment, since the vehicle room temperature is added to the input to the thermal sensation prediction means 42, the input layer of the neural network constituting the thermal sensation prediction means 42 is as shown in FIG. One element has been added compared to the example. Furthermore, the threshold values of each element of the neural network and the weights between the elements are determined in advance training using the same method as shown in the first embodiment. At this time, the vehicle room temperature is also input to the educational data.

(Function) Regarding the function of the third embodiment, mainly the differences from the second embodiment will be explained below.

The present third embodiment is characterized in that the prediction accuracy of future thermal sensation prediction based only on the thermal sensation information of the thermal sensation information detection means of the first and second embodiments is further improved by adding information on the vehicle room temperature. It has characteristics.

For example, in the first and second embodiments, if a person has been in a relatively hot space for a long time before getting into the car in the summer or gets into a car where the car temperature has risen under sunlight, the skin temperature of the passenger may be low. There is a possibility that the predicted temperature will be cooler than the actual temperature. However, if the vehicle room temperature sensor 115 is used as in the third embodiment, the environment around the occupant can be known, and thermal sensation prediction can be performed using information about the environmental temperature around the occupant other than the occupant's skin temperature. , it is possible to predict future thermal sensations more accurately.

FIG. 20 is a comparison of the predicted thermal sensation according to the third embodiment, the thermal sensation estimated by the conventional method, and the test subject's reported value when the temperature of the vehicle interior is also included in the thermal sensation prediction. The initial environmental conditions during this experiment were an outside temperature of 35° C., a car room temperature of 60° C., and an amount of solar radiation of 900 w/h.

When the vehicle room temperature is not heated as in the first embodiment, the skin temperature is determined when the difference between the declared value after the middle of the air conditioning period and the predicted thermal sensation value of the present invention, shown in The change was so slight that it was unclear whether heat was going out from the body or going into the surrounding air, making it impossible to estimate the exchange of heat between the occupant's skin and the surrounding air.
As a result, it took a long time for the subjects' reported thermal sensations to match their predicted thermal sensations after the air conditioning was sufficiently adequate and the skin temperature changed little.

However, if the car room temperature is also included as in the third embodiment,
It is possible to take into account the exchange of heat between the skin and the surrounding air, which was insufficient based on skin temperature alone, and it is possible to more accurately predict the sensation of warmth.

In FIG. 15, where the vehicle room temperature is not taken into consideration, there is a slight deviation between the predicted temperature sensation and the actual temperature sensation from the mid-air conditioning stage (■) to the late air-conditioning stage (2). However, in FIG. 20, which also takes into account the vehicle room temperature, this deviation is corrected and the temperature sensation is estimated with high accuracy.

As described above, if the car room temperature is not taken into account, the accuracy of the thermal sensation prediction may decrease in a state where skin temperature changes are small. If this is also taken into consideration, thermal sensation prediction can be performed with high accuracy even in a state where the skin temperature change is small.

The predicted thermal sensation and the control pattern for the thermal sensation stored in the control pattern storage means 43 are compared by the control amount determining means 44, and if the predicted thermal sensation deviates to the colder side, the heating is switched off during the heating season. A control amount is transmitted to the air conditioning control means 45, such as increasing the cooling capacity and lowering the cooling capacity in the cooling season. According to this control amount, the air conditioning control means 45 changes the opening degree of the air mix damper 105, etc., and controls the air conditioning in the vehicle interior 91 by the blower 102.

Since the third embodiment includes the same thermal sensation prediction correction section 51 as the second embodiment, it is possible to predict the thermal sensation control more in accordance with the passenger's preference, and to perform more accurate air conditioning control.

(Effects) As described above, in addition to the effects of the first and second embodiments, the air conditioning control device of the third embodiment inputs the temperature inside the vehicle, so that it is possible to control the temperature of the environment around the occupants. It is possible to predict the temperature sensation by taking into account the vehicle room temperature, making it possible to predict the temperature sensation with high accuracy and enable accurate air conditioning control.

Fourth Embodiment (Structure) As shown in FIG. The feature is that a regression formula or a neural network is provided as a vehicle room temperature prediction means 71 that predicts the vehicle room temperature in the near future based on the history of vehicle room temperature information.

The vehicle room temperature prediction means 71 inputs the current and past vehicle room temperatures detected by the vehicle room temperature detection means 70, which is the same vehicle room temperature sensor as in the third embodiment, and calculates a regression formula or
A neural network predicts the car's room temperature one minute later.

When predicting the vehicle room temperature using a regression equation as one method for predicting the vehicle room temperature, a linear equation expressed as T=at+b is used as the equation. Here, a and b are constants, t is time, and T is the vehicle room temperature. a, b is 1 minute 3
Using the car temperature every 30 seconds from 0 seconds ago to the present, minimum 2
Determined each time temperature prediction is performed by multiplication. a determined by the above method.

Using b, input the time after 1 minute as t, and obtain the predicted car room temperature 1 minute after T.

Here, we have used the − next equation, or if necessary, we have used the quadratic equation,
It may also be a cubic equation. It is also possible to find the constant using a method other than the method of least squares.

Furthermore, there is no problem in creating an equation using the vehicle room temperature at a predetermined time interval instead of the time t as an input variable, and calculating the predicted vehicle room temperature at a time after the predetermined time.

Next, as another method of predicting the car room temperature, Fig. 11 is a configuration diagram of a neural network when a neural network is used as the car room temperature prediction means 71.
This system inputs the car room temperature history every 0 seconds and predicts and outputs the car room temperature one minute later.

The input vehicle room temperature history and the future vehicle room temperature (after 1 minute) that will be output are calculated using a vehicle room temperature prediction neural network using the same method as shown in the first embodiment, based on data obtained through experiments in advance. Weights and thresholds within the net are set.

Details of the neural network constituting the thermal sensation prediction means 62 are shown in FIG. The structure is the same as that of the third embodiment except that one element is added to the input layer because the predicted vehicle room temperature (after 1 minute) is increased as a human input to the input layer 81.

The output layer 83 outputs a value corresponding to a future thermal sensation taking into account the current car room temperature and the car room temperature one minute later.

The threshold values of each element and the weights between the elements of this thermal sensation prediction means were determined through prior training using the same method as shown in the first embodiment. However, at this time, the educational data includes data that also includes the predicted vehicle room temperature.

(Function) Regarding the function of the fourth embodiment, the differences from the third embodiment will be mainly explained below.

The fourth embodiment adds to the third embodiment which predicts the future thermal sensation based on the thermal sensation information and the vehicle room temperature, and further adds input of information on the future vehicle room temperature predicted based on the history of the vehicle room temperature. It was added.

An infrared thermometer 11 constituting the temperature information detection means 61
6 and a vehicle room temperature sensor 115 constituting the vehicle room temperature detection means 70.
The skin temperature of the occupant detected by the sensor, the current car room temperature, and the predicted car room temperature one minute later predicted by the car room temperature prediction unit 71 are input to the thermal sensation prediction unit 62 .

In actual air conditioning, the skin temperature of the occupant is detected by the infrared thermometer 116, which is the thermal information detection means 61, and the temperature inside the vehicle is detected by the vehicle room temperature sensor (15), which is the vehicle room temperature detection means 70. Based on the vehicle room temperature detected by means 70, vehicle room temperature predicting means 71 predicts the vehicle room temperature one minute later.

The thermal sensation prediction means 62 uses an input layer 81 as shown in FIG.
Based on the current skin temperature, the previously measured skin temperature, the current car temperature and the predicted car temperature in the near future (1 minute later), the temperature sensation in the near future (1 minute later) is calculated. Predict.

FIG. 21 is a comparison between the predicted thermal sensation according to the fourth embodiment when the thermal sensation is predicted by including the history of the vehicle room temperature, the thermal sensation estimated by the conventional method, and the test subject's reported value. The initial environmental conditions in this experiment were: outside temperature 35°C, initial car room temperature 60°C.
The amount of solar radiation per day is 900W/h・i.

In the third embodiment, only the current temperature of the vehicle interior is input, and its temperature history is not taken into consideration. In this case, it is not known whether the vehicle temperature is rising or falling, and it is not possible to fully predict the future temperature sensation.

In contrast, in the fourth embodiment, since the temperature history inside the vehicle interior is also used as input data, it is possible to predict changes in the vehicle room temperature and to accurately predict the thermal sensation by considering the predicted vehicle room temperature. Even in the early stages of air conditioning due to sudden changes in skin temperature and vehicle room temperature, thermal sensations can be predicted with high accuracy as shown in FIG.

As described above, since the fourth embodiment takes into account the history of the vehicle room temperature, it is possible to improve the thermal sensation prediction accuracy even when there are sudden changes in the skin temperature and the vehicle room temperature.

The predicted thermal sensation and the control pattern for the thermal sensation stored in the control pattern storage means 63 are compared by the control amount determining means 64, and if the predicted thermal sensation deviates to the cold side, the heating capacity is changed during the heating season. The control amount is transmitted to the air conditioning control means 65, such as increasing the cooling capacity and decreasing the cooling capacity during the cooling period. According to this control amount, the air conditioning control means 65 controls the air mix damper 10.
5 and the set temperatures of the evaporator 103 and heater core 104 to adjust the air temperature, and adjust the wind speed using the blower 102 to air condition the inside of the vehicle compartment 91.

Therefore, in this fourth embodiment, when predicting future skin temperature, the future skin temperature is calculated based on the difference between the current and future car room temperatures, taking into account whether the car room temperature is currently increasing or decreasing. Feelings can be predicted. Furthermore, in the fourth embodiment, prediction information of the vehicle room temperature predicted based on the history of the vehicle room temperature can also be used, making it possible to predict the thermal sensation with higher accuracy and realize more accurate air conditioning control.

Note that the vehicle room temperature prediction means 71 can also be included in the thermal sensation prediction means 62. In this case, as shown in FIG. 13, the inputs are the history of the passenger's skin temperature every 30 seconds from 1 minute and 30 minutes ago to the current time, and the history of the vehicle room temperature, which is the environmental temperature around the passenger.

Furthermore, as shown in FIG.
It goes without saying that the thermal sensation prediction correction section 72 of the embodiment can be added.

(Effects) In addition to the effects of the first to third embodiments, the fourth embodiment uses the predicted vehicle room temperature to determine whether the current vehicle room temperature is trending upward or downward. The above-mentioned third method predicts the thermal sensation based on the current temperature inside the vehicle and the skin temperature.
It becomes possible to perform more accurate prediction of thermal sensation than in the case of the embodiment, and quick and accurate air conditioning control is realized.

Other Modifications In the first to fourth embodiments described above, an infrared thermometer was used as a means for detecting thermal sensation information to detect the skin temperature of the occupant. etc. may also be used. Furthermore, instead of directly measuring the thermal information of the human body, a signal from a human body equivalent sensor that outputs a signal equivalent to that of the human body can also be used as the thermal information.

In addition, in the above embodiment, the temperature inside the vehicle is detected by the vehicle interior temperature detection means, and the humidity inside the vehicle is adjusted as needed as shown in FIG. 14 in order to predict the temperature sensation more accurately. A humidity sensor (not shown) installed near the vehicle room temperature sensor 115 inputs information, solar radiation is inputted by a solar radiation sensor 113 installed on the dash, and environmental information inside the vehicle such as ceiling temperature is inputted by a humidity sensor (not shown) installed on the ceiling. The information is inputted by a temperature sensor, and furthermore, environmental information outside the vehicle interior, such as temperature and humidity outside the vehicle interior, can be inputted by an outside temperature sensor and outside air humidity sensor 112 provided in a duct at the front end of the vehicle.

Further, although information at a single measurement point was used in the first to fourth embodiments for thermal sensation information and vehicle interior temperature information, instead of this, information at multiple measurement points was used. It is also possible to predict the sensation of warmth. In this case, those skilled in the art can easily guess that this can be handled by changing the number of input elements and the number of intermediate elements of the thermal sensation prediction means according to the number of inputs.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of a first embodiment of the present invention,
FIG. 2 is a configuration diagram of the thermal sensation prediction means according to the first embodiment of the present invention;
FIG. 3 is a characteristic diagram of the element of the thermal sensation predicting means used in the first embodiment of the present invention, FIG. 4 is a flow chart of the adjustment method of the thermal sensation predicting means used in the first embodiment of the present invention, and FIG. The figure is a diagram of the thermal sensation control pattern of the first embodiment of the present invention, FIG. 6 is a block diagram showing the configuration of the air conditioning control device of the second embodiment of the present invention, and FIG. 7 is the diagram of the third embodiment of the present invention. FIG. 8 is a block diagram showing the configuration of an air conditioning control device according to a fourth embodiment of the present invention; FIG. 9 is a block diagram showing the configuration of a thermal sensation prediction means according to a third embodiment of the present invention; FIG. 10 is a block diagram showing the configuration of an air conditioning control device according to a fourth embodiment of the present invention, FIG. 1I is a configuration diagram of a vehicle room temperature prediction means according to a fourth embodiment of the present invention, and FIG. 13 is a block diagram of a modification of the thermal sensation prediction means of the fourth embodiment of the present invention; FIG. 14 is a schematic sectional view of the embodiment of the present invention; FIG. FIG. 15 is a diagram of the thermal sensation prediction results according to the first embodiment of the present invention, FIG. 16 is a diagram of the thermal sensation prediction results before modification according to the second embodiment of the present invention, and FIG.
The figure is a line diagram of the thermal sensation prediction result after correction in the second embodiment of the present invention, Figure 18 is a line diagram of the thermal sensation before correction of the control pattern in the second embodiment of the present invention, and Figure 19 is a diagram of the temperature sensation prediction result according to the present invention. 20 is a diagram of the thermal sensation prediction result of the third embodiment of the present invention, and FIG. 21 is a diagram of the temperature sensation after the control pattern correction of the second embodiment of the present invention. It is a line diagram of a feeling prediction result. Thermal sensation information detection means, Thermal sensation prediction means, Control pattern storage means, Control amount determination means, Air conditioning control means, Air blower, Desired thermal sensation input means, Learning data storage means, Warm sensation prediction correction means, Vehicle room temperature Detection means, ・Vehicle room temperature prediction means

Claims (1)

[Scope of Claims] A thermal sensation information detecting means for detecting information regarding a person's thermal sensation in a room, and a temperature information detecting means for detecting temperature in the near future based on a history of thermal sensation information within a predetermined time outputted from the thermal sensation information detecting means. a control pattern storage means for storing a predetermined control pattern for the thermal sensation in order to reach the target thermal sensation; a control pattern for the thermal sensation stored in the storage means; Controlled amount determining means for determining control amounts of temperature-controlled air such as air volume and temperature so that the thermal sensation outputted from the thermal sensation prediction means in the near future matches the thermal sensation in the near future; and a controlled amount outputted from the controlled amount determining means. An air conditioning control device comprising: an air conditioning control means for controlling the temperature of temperature-conditioned air blown into the room based on the control amount; and a blower for blowing the temperature-conditioned air into the room according to the control amount of the control amount determining means. .