JPH07228125A - Air conditioner for vehicle - Google Patents

Abstract

PURPOSE:To reduce the number of sensors and simplify the contents of control by providing a neural network which has neuro-learned the effects of various air conditioning load factors on comfortableness beforehand, inputting an air conditioning load factor into the network to output a PMV value, and correcting a temperature adjusting command value as a correction value to input it. CONSTITUTION:A neural network 67 is provided with an input layer (n), an output layer (m), and multiple middle layers Q which are located between the layers, and stores the effects of various air conditioning load factors on comfortableness through preceding neuro-learning. When detected values from a room temperature sensor 44, an outside air temperature sensor 43, a sun ray sensor 42, and an inside/outside air temperature mode detecting means 46 and the amount of air from a blower which is determined by a blower air amount determining means 66 are inputted into the input layer (n) of the neural network, the neural network 67 computes an occupant comfortableness index for the interior of a vehicle, that is, a PMV value, based on the air conditioning load factors. The PMV value is outputted from the output layer (m), inputted into a temperature adjustment command value compensating means 62, and displayed on a displaying device 68.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle air conditioner.

[0002]

2. Description of the Related Art In a conventional vehicle air conditioner,
A temperature control command value is calculated from the vehicle interior temperature and the set temperature, and the temperature, air volume, etc. of the conditioned air blown into the vehicle interior are automatically controlled based on the temperature control command value so that the vehicle interior temperature reaches the set temperature. Was there.

[0003]

In the above-mentioned conventional apparatus, if fine control is performed in order to improve the comfort of passengers in the passenger compartment, the number of sensors increases and the control content becomes complicated and the cost increases. There was a problem.

[0004]

The present invention has been invented in order to solve the above-mentioned problems, and its gist is to obtain a temperature control command value from at least the vehicle interior temperature, the outside air temperature and the set temperature. In a vehicle air conditioner that includes a calculating unit that calculates and automatically controls the blown air temperature, the blown air amount, and the like so that the vehicle interior temperature becomes a set temperature based on the temperature adjustment command value calculated by the calculating unit. Equipped with a neural network in which the effect of air conditioning load factors on comfort was previously learned, the air conditioning load factors were input to this neural network, PMV values were output, and the PMV values were calculated as correction values by the above calculation means. According to another aspect of the present invention, there is provided a vehicle air conditioner characterized in that the temperature control command value is input to a correction unit.

As the air conditioning load factor, at least the vehicle interior temperature, the outside air temperature, the amount of solar radiation, the blower air amount, and the inside / outside air mode can be input.

It is possible to provide a display device for informing an occupant of the PMV value output from the neural network.

[0007]

In the present invention, the air-conditioning load factor is input to the neural network in which the effects of various air-conditioning load factors on the comfort have been neuro-learned in advance, the PMV value is output, and this is input to the correction means as the correction value. Therefore, the temperature control command value is corrected.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be specifically described with reference to the drawings. As shown in FIG. 2, an inside air intake port 13 and an outside air intake port 14 are formed on the upstream side of the casing 12 of the air conditioner 10, and a face outlet 20, a defrost outlet 23, and a foot outlet 24 are provided on the downstream side. Are formed. Inside air inlet
The inside air and outside air intake 14 are switched by an inside / outside air switching damper 15. The face outlet 20, the defrost outlet 23, and the foot outlet 24 are blowout mode switching dampers 25 and 2, respectively.
It is opened and closed by 6 and 27.

A blower 16, an evaporator 17, an air mix damper 18, and a heater 19 are arranged in the casing 12. The evaporator 17 is a compressor 2 via a refrigerant pipe.
A refrigerant circulation circuit is configured by being connected to 8, a condenser 29, a receiver 30, an expansion valve 31, and the like, and the compressor 28 is driven by a traveling engine 35 via a pulley 33 and a magnet clutch 34. The cooling water of the traveling engine 35 circulates in the heater 19, and the air volume passing through the heater 19 is changed by the opening degree of the air mix damper 18.

When the blower 16 is driven by the motor 21,
The inside air or the outside air is sucked into the casing 12 from the inside air inlet 13 or the outside air inlet 14 selected by switching the inside / outside air switching damper 15, and is urged by the blower 16 and then passes through the evaporator 17. Cooled by. Next, this air is diverted by the air mix damper 18, a part of which is heated by passing through the heater 19 and then mixed with the remaining air bypassing the heater 19 to become conditioned air at a predetermined temperature. Then, this conditioned air is blown into the vehicle interior from any one or two of the face outlet 20, the defrost outlet 23, and the foot outlet 24 selected by opening and closing the blowout mode switching dampers 25, 26, 27. .

The detected values of the outside air temperature sensor 43, the room temperature sensor 44, the solar radiation sensor 42, etc. are input to the control device 11 which is a microcomputer.

Further, the operating switches installed on the control operation panel in front of the driver's seat, the operating mode changeover switch, the blowout mode changeover switch 40, the inside / outside air changeover damper 15 changeover switch 46, and the blower 16 are rotated at different speeds. Air volume selector switch 41, room temperature setting device 45 for setting the temperature inside the passenger compartment arbitrarily
Are also input to the control unit 11.

The output of the control device 11 is supplied through a driver (not shown) to the magnet clutch 34, the drive motor 21 of the blower 16, and the motor actuators of the dampers 15, 25, 26 and 27.
It is sent to 56, 57, 58, etc., and by operating them, the vehicle interior temperature is maintained at the set temperature set by the room temperature setting device.

A control block diagram is shown in FIG.
As shown in FIG. 1, the detected values of the room temperature sensor 44, the outside air temperature sensor 43, the solar radiation sensor 42 and the set temperature set in the room temperature setter 45 are input to the temperature adjustment command value calculation means 61, where the temperature adjustment command is issued. The value is calculated. The temperature control command value is corrected by the temperature control command value correction means 62, and then the blowout mode switching means 63, the on / off determination means 64 of the compressor 28, the air mix damper.
It is input to the opening degree determining means 65 of 18 and the air volume determining means 66 of the blower 16.

The switching damper opening degree determining means 63 determines the blowout mode in accordance with the map stored therein, determines the opening degree of each of the switching dampers 25, 26 and 27, and switches each by the actuator 58 in response to a command from this. Damper 25, 26, 27
Is driven to a predetermined opening.

The compressor on / off determining means 64 turns on / off the compressor 28 according to a map stored therein.
The compressor 28 is turned on or stopped by deciding OFF and turning on and off the magnet clutch 34 according to a command from this time.

The air mix damper opening degree determining means 65 determines the opening degree of the air mix damper 18 according to the map stored therein, and the actuator according to a command from this time.
The air mix damper 18 is driven by 57 to have a predetermined opening.

The air volume determining means 66 determines the air volume of the blower 16 in accordance with the map stored in the air volume determining means 66, and the motor 21 rotates at a rotational speed according to a command from now on.

A neural network 67 has an input layer n.
, An output layer m, and a plurality of intermediate layers Q interposed therebetween, the influence of various air conditioning load factors on comfort is neuro-learned and stored in advance. Therefore, the detection values of the room temperature sensor 44, the outside air temperature sensor 43, the solar radiation sensor 42 and the inside / outside air mode detecting means 46, and the blower air volume determining means.
When the blower air volume determined in 66 is input to the input layer n of the neural network, the neural network 67 calculates the passenger comfort index, that is, the PMV value, based on these air conditioning load factors.

The formula for calculating the PMV value is shown in the following formula. PMV = A ₀ + A ₁ · X ₁ + A ₂ · X ₂ + A ₃ · X ₃ + A ₄ · X ₄ + A ₅ · X ₅ where X ₁ is the outside air temperature X ₂ is the amount of solar radiation X ₃ is the cabin temperature X ₄ The blower air volume X ₅ is the inside / outside air mode Note that A ₀ , A ₁ , A ₂ , A ₃ , A ₄ , and A ₅ are coefficients and are determined according to the learning algorithm in neurolearning and the generalized method of learning results.

PM calculated by the neural network 67
The V value is output from the output layer m, and is input to the temperature adjustment command value correction means 62 and the display device 68 at the same time.
V value is displayed.

When the PMV value is input, the temperature control command value correction means 62 corrects the temperature control command value calculated by the temperature control command value calculation means 61 according to a predetermined map.

Thus, since the temperature control command value is corrected by the PMV value, the vehicle interior environment can be finely adjusted and the comfort of the occupant can be further improved.

The neural network 67 has air-conditioning load factors that affect comfort such as the traveling speed of the automobile, the humidity of the air in the passenger compartment, the air flow, the wind direction, the amount of clothing of the occupants, the amount of activity, and the preference of cooling and warming. If there is, this can be input together. Further, it is possible to control the wind direction of the conditioned air blown into the vehicle compartment, the capacity of the compressor (inverter frequency), etc. according to the temperature control command value.

[0025]

According to the present invention, the PMV value is calculated by inputting the air conditioning load factor into the neural network, and the PMV value is calculated.
Since the temperature control command value is corrected by the V value, the comfort of passengers in the vehicle compartment can be further improved.

Further, since the neural network preliminarily neuro-learns the influence of various air conditioning load factors on the comfort, it is possible to simplify the pre-tuning when the vehicle air conditioner is mounted on the vehicle.

PM output from the neural network
If a display device for notifying the passenger of the V value is provided, it is possible to notify the passenger in the passenger compartment of the comfort.

[Brief description of drawings]

FIG. 1 is a control block diagram of a vehicle air conditioner according to an embodiment of the present invention.

FIG. 2 is a schematic configuration diagram of the vehicle air conditioner.

[Explanation of symbols]

 44 Room temperature sensor 43 Outside air temperature sensor 42 Solar radiation sensor 45 Room temperature setter 46 Inside / outside air mode detection means 61 Temperature control command value calculation means 62 Temperature control command value correction means 67 Neural network 68 Display device

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor, Kominori Yokoyama, 1-chome, Asahi-cho, Nishibiwajima-cho, Nishi-Kasugai-gun, Aichi Prefecture Inside the Mitsubishi Heavy Industries, Ltd. air-conditioning factory (72) Inventor Tadao Suzuki, Nishibashijima-cho, Nishi-Kasugai-gun, Aichi Prefecture Asahi-cho 3-chome, Mitsubishi Heavy Industries, Ltd. Air conditioner factory (72) Inventor Futa Futa, Iwazuka-cho, Nakamura-ku, Nagoya

Claims (3)

[Claims] 1. A calculation means for calculating a temperature control command value from at least the vehicle interior temperature, the outside air temperature, and the set temperature, and the vehicle interior temperature becomes the set temperature based on the temperature control command value calculated by the calculation means. In a vehicle air conditioner that automatically controls the blown air temperature, blown air volume, etc., a neural network is preliminarily neurolearned about the effects of various air conditioning load factors on comfort, and the air conditioning load factors are input to this neural network. A PMV value is output as a correction value, and the PMV value is input as a correction value to a correction unit that corrects the temperature control command value calculated by the calculation unit. 2. The vehicle air conditioner according to claim 1, wherein at least the vehicle interior temperature, the outside air temperature, the amount of solar radiation, the blower air amount, and the inside / outside air mode are input as the air conditioning load factor. 3. The vehicle air conditioner according to claim 1, further comprising a display device for notifying an occupant of a PMV value output from the neural network.