JP3296582B2 - Equipment control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for an air conditioner such as an air conditioner or a hot air heater, and controls a device provided with a saving operation control means for controlling a saving operation of the device based on control data. Related to the device.

[0002]

2. Description of the Related Art Conventionally, in this type of equipment, a microcomputer is used as a control device so that the operation of the equipment can be controlled with high accuracy and complexity. However, since the control data used for controlling the operation of the device is set in advance by the manufacturer of the device or arbitrarily set by the user of the device, the former is suitable for the usability of the user. In the latter case, there was a problem that there was a concern that the setting was not appropriate.

[0003]

Therefore, if a neural network having a learning function as described in, for example, Japanese Patent Application Laid-Open No. 2-158860 is used as the control data generating means, the usability of the user is improved. It is possible to generate control data. However, the backpropagation method is used for learning the neural network. However, this backpropagation method is slow in convergence, and does not converge to the minimum value. There was a worry that it could not occur.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned facts. In the present invention, in which the control data of a control data generating means having a learning function is used to control the saving operation of the equipment, the control data is transmitted as much as possible. It is an object of the present invention to quickly converge to an ideal value and to improve user convenience.

[0005]

According to the present invention, a saving operation is performed.
Saving operation instruction means for instructing execution of a mode,
Control data of the operation of air-conditioning equipment in the operation mode
Operation control means for controlling based on the temperature and a room temperature sensor
The temperature data detected at is set by the temperature setting means.
Temperature of the set temperature data and the conditioned room to the set temperature.
The time required for air conditioning and air conditioning.
Control data suitable for updating control data of devices
And the temperature data, the set temperature data, the time data,
And new air conditioning based on the last control data output last time
Calculates the new control data
Generation of control data of neural network output as output
Means, and the saving operation control means includes the control data
Based on updated new control data obtained from
And configured to control the operation of air-conditioning equipment.
And the control data generating means performs the updated new control.
The previous control data before generating the control data
It is configured to be input together with other input data including time data .

Further, according to the present invention, the realization of the saving operation mode is performed.
Saving operation instruction means for instructing a line, and the saving operation mode
Operation of air-conditioning equipment based on control data
Controlled by the saving operation control means and the room temperature sensor
Temperature data, set temperature set by temperature setting means
Data and required to adjust the temperature of the conditioned room to the set temperature
Time data and control air-conditioning equipment.
Learning the control data suitable for updating the control data,
Data, set temperature data, time data, and previous output
New air conditioning control data based on the previous control data
Is calculated and output as updated new control data.
Neural network using backpropagation method
And control data generating means having a configuration, wherein the saving operation control means is provided.
Is the updated new data obtained from the control data generating means.
Controls the operation of air-conditioning equipment based on
And the control data generating means comprises:
Previous control data before generating new updated control data
Data is input together with other input data including temperature data and time data .

[0007]

[Function] The control data generating means outputs the temperature data and the set temperature data.
While inputting various data including data and time data ,
Based on these input data, air conditioning in the saving operation mode
Learning control data suitable Runoni, wherein the various input data
Data and a new blank based on the previous control data output last time.
Calculated air conditioning control data and updated new control
Output as data . The saving operation control means controls the saving operation of the device based on the control data. The control data generating unit, the previous control data is any other
Since the data is input together with the seed input data, the convergence of the control data is faster than that in which only data such as time and temperature is input, and the control of the saving operation which is convenient for the user of the device. Data can be obtained quickly. In particular, it is most suitable for a device having a neural network configuration using a back propagation method, such as the one described in claim 2.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to an oil fan heater will be described below with reference to the drawings.

FIG. 5 shows an oil fan heater 1 to which the present invention is applied. A burner 4 is provided in an outer case 3 having a hot air outlet 2 at the center of the front and an air inlet (not shown) at the back. , A hot air fan 5 and refueling tanks 6 and 7 which can be freely taken in and out. An operation panel 8 is provided at the front of the top surface of the outer case 3.

[0010] As shown in FIG.
An operation switch 9, a good morning timer switch 10,
The sleep timer switch 11 and the saving operation mode
A saving switch 12 (saving saving instruction means) for instructing a line ;
Up key 13 and down key 14 for room temperature adjustment, plus key 15 and minus key 1 for timer time setting
6, a plus key 17 and a minus key 18 for setting the current time, and a digital display 19 for displaying temperature and time. As the operation switch 9, an on / off switch that is turned on when pressed for the first time and is turned off when pressed for the second time is used, and a tact switch or the like that is turned on only when pressed is used for other switches and keys. I have.

FIG. 2 shows a control device for the oil fan heater. In FIG. 2, reference numeral 20 denotes a central processing unit (CP
U), program memory (ROM), data memory (R
AM) and a microcomputer having a clock function, a timer function, etc. (hereinafter referred to as a microcomputer)
The input side of the microcomputer 20 has the above-described switch,
In addition to the keys, a room temperature sensor 21, a burner thermistor 22, a frame rod 23 and the like are provided. The output side of the microcomputer 20 is provided with a warm air fan 5, a burner motor 24, a spark plug 25, a fuel pump 26, a vaporization heater 27, a display 19, and the like.

When the operation switch 9 is turned on, the microcomputer 20 executes the program stored in the program memory, starts the burner 4 on the basis of a predetermined sequence, and operates the hot air fan 5. Further, data of the temperature (for example, 10 ° C. to 30 ° C.) set by the up key 13 and the down key 14 for temperature setting is stored in the data memory, and the temperature and the room temperature detected by the room temperature sensor 21 are stored. And the amount of combustion in the burner 4 is adjusted according to the temperature difference. When the good morning timer switch 10 is pressed, the heating operation is temporarily stopped and then a plus key 15 for setting the timer time is set.
And resumes at the time preset by the minus key 16 and stored in the data memory. When the good night timer switch 11 is pressed, the heating operation is continued as it is and stops after a predetermined time.

When the saving switch 12 is pressed together with the pushing operation of the operation switch 9, the microcomputer 20 enters a saving operation mode. That is, as shown in FIG. 1, the microcomputer 20 has control data generating means 28 for generating control data (saving temperature data) TD in the saving operation mode.
When, therefore it is calculated in the control data generating unit 28, a new
A saving operation control unit 29 for controlling the saving heating operation based on new control data output as air conditioning control data is built in.

The control data generating means 28 has four input layers I
1 to I4, the five intermediate layers H1 to H5, the output layer S, the teacher data generating means 30, and the teacher data from the teacher data generating means 30 and the control data of the output layer S. output layer according to the error between the control data - the coupling coefficient of the intermediate layers, and an intermediate layer - consists errors to update respectively the coupling coefficient of the input interlayer correction means 31., has a neural network configuration having a learning function.

The input layer I1 is connected to a temperature difference signal generating means 32 for generating temperature difference data ΔT between a set temperature TS (for example, 20 ° C.) of the saving operation and an initial room temperature T0 of the saving operation.
The input layer I2 is connected to time data generating means 33 which generates time data t1 from the start of the saving operation to the time when the room temperature reaches the set temperature TS of the saving operation. Also, the input layer I
3 is an up key 13 for setting the temperature during the previous saving operation.
And a user satisfaction data generating means 34 for generating user satisfaction data N determined from the number of times the down key 14 is operated. The satisfaction data N is a value indicating the satisfaction during the previous saving operation. The initial value is 0, and 1 is subtracted each time the temperature setting up key 13 is pressed once, and the down key 1
Each time 4 is pressed once, 1 is added, and when the up key 13 and the down key 14 for temperature setting are never operated, it becomes 0. The input layer I4 also includes a previous data generating means 3 for generating the previous control data TD 'of the control data generating means.
5 is connected.

The data input to the input layers I1 to I4 is adjusted by the coupling coefficient and transmitted to the intermediate layers H1 to H5. Intermediate layers H1 to H5
Calculates the sum of the inputs from the input layers I1 to I4,
An output is calculated (for example, a Sigmond function) based on the sum of these inputs. The outputs of the intermediate layers H1 to H5 are transmitted to the output layer S after being adjusted by other coupling coefficients. In the output layer S, the sum of all inputs is obtained, the output is calculated (for example, a Sigmond function) based on the sum of these inputs, and is output as newly updated control data (saving temperature data) TD.

The learning of the control data generating means 28 is performed in a factory or in the field. The factory learning means that the microcomputer 20 stores the above-described temperature difference data ΔT, time data t1, satisfaction degree data N, previous control data TD ′, and preferable control data (teacher data).
Are repeatedly input to generate control data TD suitable for each data. On the other hand, on-site learning refers to inputting data based on actual use, finding ideal teacher data in the microcomputer 20, and performing learning according to actual use conditions. It may be performed, or both factory learning and on-site learning may be performed.

Table 1 shows the control data generating means 28 described above.
3 shows an example of learning. For example, in the learning example,
Temperature difference data ΔT is larger by 2 ° C. than that of
Since it is considered that the temperature has become relatively colder than the previous time, the control data TD decreases from 2.5 ° C. to 1.75 ° C.
In the learning example of, the time data t1 required for the room temperature to reach the set temperature TS is 25 minutes shorter than in the case of, and is considered to be relatively warm. Therefore, the control data TD is changed from 1.75 ° C. to 2.15 ° C. ° C. In the learning example of, the satisfaction data increases from 0 to +1 as compared with, and it is considered that the user felt hot in the previous economizing driving. Therefore, the control data TD was changed from 2.15 ° C. to 2.85 ° C.
To become larger.

[0019]

[Table 1]

In any of the learning examples, the previous control data TD 'is input to the control data generating means 28 together with other data, so that the control data generating means 28 learns to match the life pattern of the user. , And outputs control data (saving temperature data) TD. That is, by repeatedly inputting the previous control data TD 'together with other data to the control data generating means 28, the control data TD is quickly converged to an ideal value suitable for the life pattern of the user. Become.

FIGS. 3 and 4 show specific examples of the saving operation. When the microcomputer 20 enters the saving operation mode,
The temperature difference between the set temperature TS and the initial room temperature T0 is determined. Further, a time t1 required for the room temperature to reach the set temperature TS is obtained. These data ΔT and t1 are important factors in estimating the heating load. When the room temperature reaches the set temperature TS, the data ΔT and t1 and the previous satisfaction data N and control data TD ′ are input to the control data generating means 28, and the new control data T
D is output. When the room temperature reaches the set temperature TS, the saving operation control means 29 maintains the room temperature at the set temperature TS for a predetermined time (t2), and then changes the room temperature from TS to (TS3) for a predetermined time (t3) commensurate with the control data. -TD) to perform a conservative heating operation that gradually decreases to TD).

According to this embodiment, the control data generating means 2
8, the temperature difference data ΔT relating to the heating load
Since the previous control data (saving temperature data) TD 'is inputted together with the time data t1 and the previous satisfaction data N, learning suitable for the heating load and the user's satisfaction (comfortable feeling) is performed. Not only is learning performed, but learning adapted to the life pattern of the user is performed, and the control data (saving temperature data) TD quickly converges to an ideal value.
Comfortable and economical heating operation can be realized.

[0023] In the embodiment described above, control is data generated the saving temperature data during the save operation in the generating means 28 Therefore as the control temperature data, setting at saving operating temperature T
Various control data such as S and room temperature holding time t2 may be generated. In addition, the present invention can be applied as a control device for an air conditioning device that controls operation based on control data, such as an air conditioner, in addition to the above-described hot air heater such as an oil fan heater.

[0024]

According to the present invention as the invention has been described above, in that so as to control the saving operation operation of the equipment to air conditioning by using the control data of the control data generating means of the neural network structure having a learning function, the control In the data generation means, the last control data is the temperature data, the set temperature data,
As it is input together with other data including time data, it is calculated and updated based on these various data
To converge to the ideal of value that was in the life pattern of as soon as possible to the user a new control data, heater and air conditioner
The control data generating means has a neural network configuration using a back propagation method, such as the one described in claim 2, wherein the user can improve the usability of the air-conditioning device such as the air conditioner. Suitable for things.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of an internal configuration of a microcomputer that also serves as a control data generating unit and a control unit.

FIG. 2 is a block diagram showing a basic configuration of a control device according to one embodiment of the present invention.

FIG. 3 is a flowchart for explaining the operation of the control device.

FIG. 4 is an explanatory diagram of operation characteristics of the control device.

FIG. 5 is a perspective view of an oil fan heater to which the present invention is applied.

FIG. 6 is a plan view of an operation panel of the oil fan heater.

[Explanation of symbols]

Reference Signs List 1 Oil fan heater (equipment) 12 Saving switch (Saving operation instructing means) 13 Up key for temperature setting (Temperature setting means) 14 Down key for temperature setting (Temperature setting means) 20 Microcomputer 21 Room temperature sensor 28 Control data generation Means 29 Saving operation control means 35 Previous data generation means

──────────────────────────────────────────────────続 き Continuation of front page (72) Inventor Masafumi Okada 2-18-18 Keihanhondori, Moriguchi-shi, Osaka Sanyo Electric Co., Ltd. (56) References JP-A-3-77101 (JP, A) JP-A-Hei JP-A-3-255504 (JP, A) JP-A-3-14002 (JP, A) JP-A-5-240973 (JP, A) JP-A-5-240974 (JP, A) JP-A-60-107103 (JP, A) A) JP-A-3-134706 (JP, A) JP-A-2-81161 (JP, A) JP-A-3-201160 (JP, A) JP-A-3-28627 (JP, A) JP-A-2 −289863 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G05B 13/02

Claims (2)

(57) [Claims] 1. A saving operation instructing execution of a saving operation mode.
Change instruction means and air-conditioning in the saving operation mode
Saving operation system that controls the operation of equipment based on control data
Control means, temperature data detected by the room temperature sensor, temperature
Temperature data set by the temperature setting means and the conditioned room
Enter the time data required to adjust the temperature to the set temperature
At the same time as updating control data for air-conditioning equipment.
The control data learned is learned and the temperature data and set temperature data are learned.
Data, time data, and previous control data output last time
Calculates new air conditioning control data based on the
Neural network to output as new control data
And control data generating means having a configuration, wherein the saving operation control means is provided.
Is the updated new data obtained from the control data generating means.
Controls the operation of air-conditioning equipment based on
And the control data generating means comprises:
Previous control data before generating new updated control data
A controller for inputting data together with other input data including temperature data and time data . 2. A saving operation instructing execution of a saving operation mode.
Change instruction means and air-conditioning in the saving operation mode
Saving operation system that controls the operation of equipment based on control data
Control means, temperature data detected by the room temperature sensor, temperature
Temperature data set by the temperature setting means and the conditioned room
Enter the time data required to adjust the temperature to the set temperature
At the same time as updating control data for air-conditioning equipment.
The control data learned is learned and the temperature data and set temperature data are learned.
Data, time data, and previous control data output last time
Calculates new air conditioning control data based on the
Backproperge that outputs as new control data
Control Data Generation of Neural Network Configuration Using Motion Method
And a saving operation control means, the control data
Based on the updated new control data obtained from the generator
To control the operation of air-conditioning equipment
And the control data generating means performs the updated new control.
The previous control data before generating the control data
A device control device configured to be input together with other input data including time data .