JPH07120022A - Automatic ventilation controller - Google Patents

Abstract

PURPOSE:To perform controls so that the ventilation rate in a room can be suitable for bodily sensation, by a method wherein the environmental conditions inside the room are detected by two or more sensors, outputs from the sensors are input into a neural network that has been given a learning previously, and a ventilation fan is controlled so that it operates at a ventilation rate selected according to the outputs from the sensors. CONSTITUTION:Sensors 1a-1n are devices to input the environmental conditions inside a room and are connected to units of input layers of a neural network 2. The neural network 2 receives the output values from the sensors 1a-1n as inputs, and the ventilation rates are allocated to the units of output layers. A judging device 3 selects a ventilation rate based on the output values from the units, and a driving controller 4 controls a ventilation fan 5 so that it operates at a ventilation rate determined by the judging device 3. A switch 7 has ventilation keys 8 and an indication panel 9 and is used to designate learning/operation modes and the ventilation rate at the time of learning and indicate the learing/operation modes when a resident performs a learing. Furthermore, the opening degree of an air outlet is controlled in coordination with the action of the fan 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic ventilation control device for detecting an environmental condition in a room and automatically adjusting a ventilation amount in the room.

[0002]

2. Description of the Related Art Conventionally, an automatic ventilation control device of this type is known from Japanese Utility Model Laid-Open No. 60-7538. In this automatic ventilation control device, a gas sensor that detects the concentration of miscellaneous gas in the room is installed in the control unit that controls the operation of the ventilation fan.
Individually connected, the air cleanliness of the room is judged based on the level of the output level of the gas sensor, and the operation of the ventilation fan is automatically started and stopped.

[0003]

However, since it is necessary to control ventilation in consideration of the overall environmental conditions in the room including humidity, dust amount, etc. in addition to the concentration of miscellaneous gases obtained from the gas sensor, the above-mentioned conventional method is used. The automatic ventilation control device of (1) cannot accurately determine whether ventilation is necessary or not, and thus cannot perform optimal ventilation control. Further, there is a problem in that the degree of environmental conditions requiring ventilation, for example, the discomfort caused by humidity, the concentration of miscellaneous gas, etc., differs for each resident, and no measures are taken to deal with them. The present invention has been made in view of the above problems, and an object of the present invention is to provide an automatic ventilation control device that adjusts an indoor ventilation volume that matches a sensation based on a plurality of elements that indicate an indoor environmental condition. To do.

[0004]

Therefore, the present invention provides a ventilation fan for ventilating a room, a plurality of sensors for detecting an environmental condition in the room, and an output value of the sensor as an input to an input layer. A neural network that has been made to learn by allocating a ventilation amount to the, a determination unit that selects the ventilation amount from the output value of the output layer of the neural network, and drive control that controls the operation of the ventilation fan based on the selection result of the determination unit. It is composed of parts. Further, in the above configuration, an air inlet opening / closing unit for controlling opening / closing of an air inlet for supplying outside air into the room is provided,
The drive control unit is configured to control the opening degree of the air supply port in conjunction with the control of the operation of the ventilation fan.

[0005]

With the above configuration, the indoor environmental condition is detected by a plurality of sensors, the output of each sensor is input to the neural network that has been learned in advance, and the ventilation amount is determined by the output value output from each unit of the output layer. Select to control the operation of the ventilation fan. Further, when the ventilation amount is selected and the operation of the ventilation fan is controlled, the opening degree of the air supply port is controlled in conjunction with it.

[0006]

EXAMPLES The present invention will be described in detail below based on examples. FIG. 1 is a block diagram showing an embodiment of an automatic ventilation control device according to the present invention. In the figure, the sensors 1a to 1n are means for inputting environmental conditions in the room, and are connected to each unit of the input layer of the neural network 2. The neural network 2 receives the output values of the sensors 1a to 1n as inputs, and the ventilation volume is assigned to each unit in the output layer. The determination unit 3 selects the ventilation amount based on the output value from each unit in the output layer of the neural network 2. The drive control unit 4 is
The ventilation fan 5 is controlled to the ventilation amount determined by the determination unit 3. Further, the air supply opening / closing unit 6 for opening / closing the air supply opening may be provided, and the opening / closing control may be performed according to the increase / decrease of the ventilation amount. The switch 7 includes operation keys 8 and a display panel 9.
It is used for designating a learning / use mode, designating a ventilation volume at the time of learning, and displaying a learning / use mode when a resident learns.

When the elements indicating the environmental condition in the room are humidity, the concentration of miscellaneous gas, and the amount of dust, and the ventilation amount to be judged is the strong operation or the weak operation, the sensors 1a to 1n, the neural network 2, and the judging section 3 are used. May be configured as shown in FIG. That is, the humidity sensor 11, the gas sensor 12, and the dust amount sensor 13 are provided, and the units 21, 22, and 23 of the input layer of the neural network 2 are assigned to the respective sensors. Further, the unit 24 in the output layer is assigned to the weak operation, and the unit 25 is assigned to the strong operation.

First, the operation during learning will be described. Humidity, miscellaneous gas concentration, and dust amount in the room are detected by a humidity sensor 11, a gas sensor 12, and a dust amount sensor 13, respectively. Each sensor sequentially detects at 1-second intervals, and the output values averaged every 5 minutes are input to the units 21, 22, and 23, respectively. If the occupant's experience determines the amount of ventilation in the environmental state at this time, and if it is felt that a large amount of ventilation is required, the output of the unit 25 to which strong driving is assigned is “1”, and weak driving is Assigned unit 2
The output of 4 is set to "0". If it is determined that a small ventilation amount is sufficient, the output of the unit 25 is set to "0" and the output of the unit 24 is set to "1". By repeating this process for about 30 minutes, the neural network 2 is learned by the existing learning algorithm. When the resident performs learning, the switch 7 is provided, and the operation key 8 is used to specify the learning mode and the ventilation amount.
At this time, the display panel 9 displays "learning" by designating the learning mode, and turns off the display after the learning is completed.

Next, the operation during use will be described. Humidity, miscellaneous gas concentration, and dust amount in the room are detected by a humidity sensor 11, a gas sensor 12, and a dust amount sensor 13, respectively. Each sensor sequentially detects at 1-second intervals, and the output values averaged every 5 minutes are input to the units 21, 22, and 23, respectively. For this input, the units 24, 2
The determination unit 3 selects the ventilation amount according to the magnitude of the output value output from the device 5. That is, as shown in FIG. 3A, when the output value of the unit 25 is larger than the output value of the unit 24, the strong operation is selected. In addition, in the determination, as shown in FIGS. 3B and 3C, one-step or two-step thresholds are set, and when one exceeds the threshold and the other is below the threshold, the output is performed. A method of selecting the driving with the larger value may be adopted. When the ventilation amount is selected by the determination unit 3, the drive control unit 4 controls the ventilation fan 5 based on the selection result. In the case of FIG. 3A, since the strong operation is selected, when the ventilation fan 5 is in the weak operation, it is switched to the strong operation, and in the case of the strong operation, the current state is maintained until the next determination. The drive control unit 3 can also be configured to control the air supply port opening / closing unit 6 in conjunction with the switching of the operation. That is, the opening degree is controlled to be large during the strong driving and to be decreased during the weak driving.

In the above embodiment, the sensors 11 to 1 are used.
3 detects the environmental condition at an interval of 1 second and averages it every 5 minutes, but this can be arbitrarily set without departing from the object of the present invention. Needless to say, if one of the ventilation volumes assigned to the units in the output layer of the neural network 2 is set to "0", the ventilation fan 5 is turned on and off.

[0011]

As described above in detail, according to the present invention,
The ventilation volume is controlled in consideration of the overall environmental conditions in the room, including the concentration of miscellaneous gases, humidity, dust amount, etc., so that extremely efficient ventilation can be performed without causing excessive ventilation or insufficient ventilation. . In addition, since a neural network is used, complicated control is not required to adjust each element that indicates the environmental condition in the room, and the ventilation volume can be controlled according to the sensation of the occupants with an extremely simple configuration. it can. further,
Since the opening degree of the air supply port is controlled in conjunction with the control of the ventilation volume, there is also an effect that the room temperature change due to ventilation can be minimized.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

[Fig. 2] The elements showing the environmental condition are humidity, miscellaneous gas concentration,
It is explanatory drawing which shows a structure when it is set as a dust amount.

FIG. 3 is an explanatory diagram showing a processing operation of a determination unit.

[Explanation of symbols]

 1a to 1n Sensor 2 Neural network 3 Judgment part 4 Drive control part 5 Ventilation fan 6 Air supply opening / closing part 7 Switch 8 Operation key 9 Display panel

Claims (4)

[Claims] 1. A ventilation fan for ventilating a room, a plurality of sensors for detecting an environmental condition in the room, and an output value of the sensor as an input to each unit of an input layer, and a ventilation amount for each unit of the output layer. Is assigned to the neural network, a determination unit that selects a ventilation amount from the output value of the output layer of the neural network, and a drive control unit that controls the operation of the ventilation fan based on the selection result of the determination unit. An automatic ventilation control device characterized by the above. 2. The automatic ventilation control device according to claim 1, wherein the determination unit selects a ventilation volume corresponding to a unit having a large output value among the units in the output layer of the neural network. 3. The determination unit determines that one of the units in the output layer of the neural network has one output value when the output value exceeds a predetermined level and the other output values are below the predetermined level. The automatic ventilation control device according to claim 1, wherein a corresponding ventilation amount is selected. 4. An air inlet opening / closing unit for controlling opening / closing of an air inlet for supplying outside air into the room is provided, and the drive controller interlocks with the control of the operation of the ventilation fan. The automatic ventilation control device according to claim 1, wherein the opening degree of the control device is controlled.