AIR-CONDITIONER BASED ON PARAMETER LEARNING USING ARTIFICIAL INTELLIGENCE, CLOUD SERVER, AND METHOD OF OPERATING AND CONTROLLING THEREOF

• US 20190309970A1
• Filed: 04/10/2019
• Published: 10/10/2019
• Est. Priority Date: 04/10/2018
• Status: Active Grant

• Current Assignee: LG Electronics, Inc. (LG Corporation)
• Original Assignee: LG Electronics, Inc. (LG Corporation)
• Inventors: PARK, Yun Sik, KWON, Young Tag, HAN, Dongwoo
• Granted Patent: US 11,353,226 B2
• Time in Patent Office: Days
• CPC Class Codes: F24F 1/06 Separate outdoor units, e.g. outdoor unit to be linked to a separate room comprising a compressor and a heat exchanger
• CPC Class Codes: F24F 11/0008 for air-humidification F24F11/30 takes precedence
• CPC Class Codes: F24F 11/30 for purposes related to the operation of the system, e.g. for safety or monitoring
• CPC Class Codes: F24F 11/56 Remote control
• CPC Class Codes: F24F 11/58 using Internet communication
• CPC Class Codes: F24F 11/61 using timers
• CPC Class Codes: F24F 11/64 using pre-stored data
• CPC Class Codes: F24F 11/65 for selecting an operating mode
• CPC Class Codes: F24F 11/77 by controlling the speed of ventilators
• CPC Class Codes: F24F 11/79 for controlling the direction of the supplied air
• CPC Class Codes: F24F 11/80 for controlling the temperature of the supplied air
• CPC Class Codes: F24F 11/86 by controlling compressors within refrigeration or heat pump circuits
• CPC Class Codes: F24F 2110/10 Temperature
• CPC Class Codes: F24F 2110/20 Humidity
• CPC Class Codes: F24F 2120/12 Position of occupants
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: Y02B 30/70 Efficient control or regulation technologies, e.g. for control of refrigerant flow, motor or heating

Abstract

An air conditioner includes: an indoor unit including a blowing unit to discharge air; an outdoor unit supplying compressed refrigerant to the indoor unit, to exchange heat with the air before discharge. The air conditioner also includes at least one computer memory operably connectable to at least one processor and storing instructions that, when executed, perform operations that include: generating at least one parameter in a first operation mode, which is operated only within a preset time range; controlling, based on an operation mode information, at least one of (i) the blowing unit to change a wind direction and an air volume, or (ii) the outdoor unit; and switching to a second operation mode after a time period of operation in the first operation mode. The operation mode information includes at least one result factor obtained from at least one machine-learning network.

Claims

1. An air conditioner, comprising:

an indoor unit comprising a blowing unit configured to discharge air;

an outdoor unit configured to supply compressed refrigerant to the indoor unit, such that the supplied refrigerant exchanges heat with the air before the air is discharged from the blowing unit;

at least one processor; and

at least one computer memory operably connectable to the at least one processor and storing instructions that, when executed by the at least one processor, perform operations comprising;

generating at least one parameter in a first operation mode of the air conditioner, wherein the first operation mode is operated only within a preset time range;

controlling, based on an operation mode information, at least one of (i) the blowing unit of the indoor unit to change a wind direction and an air volume of the discharged air, or (ii) the outdoor unit, wherein the operation mode information comprises at least one result factor that is obtained from at least one machine-learning network; and

performing an operation of switching to a second operation mode after a time period of operation in the first operation mode,

wherein an absolute value of a first temperature change rate during the first operation mode is greater than an absolute value of a second temperature change rate during the second operation mode.

2. The air conditioner of claim 1, further comprising a communication unit configured to communicate with a cloud server,

wherein the at least one result factor obtained from the at least one machine-learning network is obtained from at least one of (i) the cloud server after the communication unit transmits the at least one parameter to the cloud server, or (ii) an embedded learning unit.

3. The air conditioner of claim 1, wherein the operations further comprise:

based on the air conditioner operating for cooling, maintaining a temperature during a first period from a time point when the first operation mode ends, and subsequently increasing the temperature compared to the temperature of the first period during a second period; and

based on the air conditioner operating for heating, maintaining a temperature during a first period from a time point when the first operation mode ends, and subsequently reducing the temperature compared to the temperature of the first period during a second period,

wherein the first period and the second period form a time interval of the second operation mode.

4. The air conditioner of claim 3, wherein the operations further comprise:

receiving, through an interface unit, a control signal related to an increase or a decrease in a temperature of the air conditioner,

wherein, based on the air conditioner operating for cooling in the second operation mode, increasing or decreasing a length of the first period or the second period in inverse proportion to the increase or the decrease of the temperature by the control signal.

5. The air conditioner of claim 1, wherein the operations further comprise:

based on a difference between a current temperature and a target set temperature being within a temperature band, or based on the difference being maintained for a time period in the second operation mode, performing an operation to switch to the first operation mode.

6. The air conditioner of claim 1, wherein the operations further comprise:

performing a humidity control operation mode based on a humidity level, which is sensed at a time of switching from the first operation mode to the second operation mode, exceeding a first reference threshold criteria, and

releasing the humidity control operation mode based on the humidity level reaching a second reference threshold criteria and being maintained for a period of time.

7. The air conditioner of claim 1, wherein the operations further comprise:

storing position information related to a residing area, based on an occupant repeatedly being detected in a space where the air conditioner is installed; and

setting a left/right wind direction of the blowing unit in the first operation mode to be directed to the residing area.

8. The air conditioner of claim 7, wherein the operations further comprise:

storing position information related to an activity area that comprises the residing area, based on the occupant being intermittently detected; and

setting the left/right wind direction of the blowing unit in the second operation mode to the activity area.

9. The air conditioner of claim 1, wherein the operations further comprise:

based on (i) the air conditioner starting to operate and the difference between a current indoor temperature and a target set temperature being greater than a predefined value, or (ii) the air conditioner receiving an input signal that instructs a driving of the first operation mode;

performing an operation of switching to the first operation mode.

10. The air conditioner of claim 1, wherein the at least one parameter comprises:

(i) an indoor initial temperature related to a time at which the first operation mode is started, (ii) a target set temperature for the first operation mode, (iii) a first temperature change rate during an initial time interval of the first operation mode, (iv) a second temperature change rate during the first operation mode, and (v) a time difference between a start time and an end time of the first operation mode.

11. The air conditioner of claim 7, wherein the at least one machine-learning network comprises:

at least one input layer that has the at least one parameter as at least one input node;

at least one output layer that has the operation mode information as at least one output node; and

at least one hidden layer arranged between the at least one input layer and the at least one output layer,

wherein weights of at least one node and at least one edge between the at least one input node and the at least one output node are updated by a learning process of the at least one machine-learning network,

12. A method of controlling an air conditioner based on machine-learning, the air conditioner comprising (i) an indoor unit having a blowing unit configured to discharge air, and (ii) an outdoor unit configured to supply compressed refrigerant to the indoor unit, the method comprising:

operating the blowing unit to discharge air in a first operation mode of the air conditioner within a preset time range, wherein the first operation mode is operated only within a preset time range;

generating at least one parameter in the first operation mode of the air conditioner;

controlling, based on an operation mode information, at least one of (i) the blowing unit of the indoor unit to change a wind direction and an air volume of the discharged air, or (ii) the outdoor unit, wherein the operation mode information comprises at least one result factor that is obtained from at least one machine-learning network; and

performing an operation of switching to a second operation mode after a time period of operation in the first operation mode,

wherein an absolute value of a first temperature change rate during the first operation mode is greater than an absolute value of a second temperature change rate during the second operation mode.

13. The method of claim 12, further comprising:

based on the air conditioner operating for cooling, maintaining a temperature during a first period from a time point when the first operation mode ends, and subsequently increasing the temperature compared to the temperature of the first period during a second period; and

based on the air conditioner operating for heating, maintaining a temperature during a first period from a time point when the first operation mode ends, and subsequently reducing the temperature compared to the temperature of the first period during a second period,

wherein the first period and the second period form a time interval of the second operation mode.

14. The method of claim 13, wherein the at least one parameter comprises (i) an indoor initial temperature related to a time at which the first operation mode is started, (ii) a target set temperature for the first operation mode, (iii) a first temperature change rate during an initial time interval of the first operation mode, (iv) a second temperature change rate during the first operation mode, and (v) a time difference between a start time and an end time of the first operation mode.

15. The method of claim 12, further comprising:

based on a difference between a current temperature and a target set temperature being within a temperature band, or based on the difference being maintained for a time period in the second operation mode, performing an operation to switch to the first operation mode.

16. The method of claim 12, further comprising:

performing a humidity control operation mode based on a humidity level, which is sensed at a time of switching from the first operation mode to the second operation mode, exceeding a first reference threshold criteria, and releasing the humidity control operation mode based on the humidity, level reaching a second reference threshold criteria and being maintained for a period of time.

17. A cloud server, comprising:

a communication unit configured to communicate with a plurality of air conditioners;

at least one processor; and

at least one computer memory operably connectable to the at least one processor and storing instructions that, when executed by the at least one processor, perform operations comprising;

receiving, from each of the plurality of air conditioners, at least one parameter generated in a first operation mode of respective air conditioner, wherein the at least one parameter is related to at least one temperature individually set in the respective air conditioner;

for each of the plurality of air conditioners, using at least one machine-learning network to process the at least one parameter received from the respective air conditioner, and generating, through a learning process, operation mode information for the respective air conditioner; and

transmitting, to each of the plurality of air conditioners, the operation mode information for the respective air conditioner,

wherein the operation mode information comprises data to set a second operation mode of the respective air conditioner after a period of time of operating the first operation mode, and

wherein an absolute value of a first temperature change rate during the first operation mode is greater than an absolute value of a second temperature change rate during the second operation mode of the respective air conditioner.

18. The cloud server of claim 17, wherein, for each of the plurality of air conditioners, the at least one parameter comprises (i) an indoor initial temperature related to a time at which the first operation mode is started for the respective air conditioner, (ii) a target set temperature for the first operation mode of the respective air conditioner, (iii) a first temperature change rate during an initial time interval of the first operation mode of the respective air conditioner, (iv) a second temperature change rate during the first operation mode of the respective air conditioner, and (v) a time difference between a start time and an end time of the first operation mode of the respective air conditioner.

19. The cloud server of claim 18, wherein the at least one machine-learning network comprises:

at least one input layer that has the at least one parameter as at least one input node;

at least one output layer that has the operation mode information as at least one output node; and

at least one hidden layer arranged between the at least one input layer and the at least one output layer,

wherein weights of at least one node and at least one edge of the at least one input node and the at least one output node are updated by a learning process of the at least one machine-learning network.