CN116997757A

CN116997757A - Air conditioning system, method for estimating amount of refrigerant in air conditioning system, air conditioner, and method for estimating amount of refrigerant in air conditioner

Info

Publication number: CN116997757A
Application number: CN202280021759.0A
Authority: CN
Inventors: 佐佐木慎司
Original assignee: Fujitsu General Ltd
Current assignee: Fujitsu General Ltd
Priority date: 2021-03-31
Filing date: 2022-02-24
Publication date: 2023-11-03
Also published as: US20240175595A1; EP4317820A1; WO2022209444A1; JP2022157825A; JP7147909B1; AU2022247651A1

Abstract

An air conditioning system includes an air conditioner having a refrigerant circuit configured by connecting an outdoor unit and at least one indoor unit through refrigerant pipes, the refrigerant circuit being filled with a predetermined amount of refrigerant, and a server communicatively connected to the air conditioner. The air conditioner has a first communication unit that detects a state quantity related to control of the air conditioner, acquires a detected value, and transmits the acquired detected value to a server. The server has: a second communication unit for receiving the detection value from the air conditioner; an estimating unit that uses a detected value of a first characteristic amount to estimate a remaining refrigerant amount remaining in the refrigerant circuit when a state amount related to the refrigerant amount filled in the refrigerant circuit is taken as the first characteristic amount; and a determination unit for determining whether or not the detected value of the first feature amount is a detected value to be used for estimating the amount of remaining refrigerant. As a result, even when the feature quantity for estimating the remaining refrigerant quantity is affected by another failure, it is possible to achieve an improvement in the estimation accuracy of the remaining refrigerant quantity.

Description

Air conditioning system, method for estimating amount of refrigerant in air conditioning system, air conditioner, and method for estimating amount of refrigerant in air conditioner

Technical Field

The present invention relates to an air conditioning system, a method for estimating the amount of refrigerant in the air conditioning system, an air conditioner, and a method for estimating the amount of refrigerant in the air conditioner.

Background

In recent years, various methods for detecting the amount of refrigerant charged in a refrigerant circuit have been proposed in a multi-connected air conditioner in which a plurality of indoor units are connected to an outdoor unit. In patent document 1, for example, the amount of refrigerant is determined using the degree of supercooling of the condenser outlet, with the refrigerant circuit as a predetermined condition.

Further, the applicant filed patent document 2, in which a model for estimating the amount of refrigerant remaining in the refrigerant circuit is generated by multiple regression analysis using the characteristic amount of the refrigerant circuit related to the amount of refrigerant, and the amount of remaining refrigerant is estimated using the model.

Patent document 1: japanese patent laid-open No. 2006-23072

Patent document 2: japanese patent laid-open No. 2021-156528

Disclosure of Invention

In the model for estimating the amount of residual refrigerant, the amount of residual refrigerant is estimated using a feature quantity having a correlation with the amount of residual refrigerant among a plurality of feature quantities related to the refrigerant circuit. However, these characteristic amounts may also be related to abnormal states other than the decrease in the amount of remaining refrigerant due to refrigerant leakage, for example, a failure of the compressor. Therefore, if any of the characteristic amounts related to the remaining refrigerant amount changes due to factors other than refrigerant leakage, for example, failure of the device constituting the refrigerant circuit, there is a risk that an erroneous estimation result of the remaining refrigerant amount is obtained. In addition, when the characteristic amount other than the characteristic amount related to the remaining refrigerant amount is changed by a factor other than the refrigerant leakage, there is a possibility that the characteristic amount that appears normal, which has a relationship to the refrigerant leakage, is also affected by the factor other than the refrigerant leakage.

The present invention has been made in view of the above-described problems, and an object of the present invention is to provide an air conditioning system, a refrigerant amount estimating method for an air conditioning system, an air conditioner, and a refrigerant amount estimating method for an air conditioner, which can achieve an improvement in accuracy of estimating the amount of remaining refrigerant even when a feature amount for estimating the amount of remaining refrigerant is affected by another failure.

An air conditioning system according to one aspect includes an air conditioner having a refrigerant circuit configured by connecting an outdoor unit and at least one indoor unit through a refrigerant pipe, the refrigerant circuit being filled with a predetermined amount of refrigerant, and a server communicatively connected to the air conditioner. The air conditioner comprises: a detection unit configured to detect a state quantity related to control of the air conditioner; an acquisition unit configured to acquire a detection value detected by the detection unit; and a first communication unit that transmits the detection value acquired by the acquisition unit to the server. The server has: a second communication unit configured to receive the detection value from the air conditioner; an estimating unit that estimates an amount of remaining refrigerant remaining in the refrigerant circuit using a detected value of a first characteristic amount when the first characteristic amount is a state amount related to an amount of refrigerant filled in the refrigerant circuit; and a determination unit configured to determine whether or not the detected value of the first feature amount is a detected value to be used for estimating the amount of remaining refrigerant by the estimation unit.

On the other hand, even when the feature amount for estimating the remaining refrigerant amount is affected by another failure, it is possible to achieve an improvement in the accuracy of estimating the remaining refrigerant amount.

Drawings

Fig. 1 is an explanatory diagram showing an example of an air conditioner according to the present embodiment.

Fig. 2 is an explanatory diagram showing an example of the outdoor unit and the indoor unit.

Fig. 3 is a block diagram showing an example of a control circuit of the outdoor unit.

Fig. 4 is a mollier chart showing a state in which the refrigerant of the air conditioner is changed.

Fig. 5 is an explanatory diagram showing an example of the first feature amount used in the first to third estimated models for cooling and the second feature amount used in the judgment model for cooling.

Fig. 6 is an explanatory diagram showing an example of the first feature quantity used in the first to third estimated heating models and the second feature quantity used in the judgment model at the time of heating.

Fig. 7A is an explanatory diagram showing an example of a case where the estimation result of the first cooling estimation model and the estimation result of the second cooling estimation model are not interpolated by an S-type (Sigmoid) curve.

Fig. 7B is an explanatory diagram showing an example of a case where the estimation result of the first cooling estimation model and the estimation result of the second cooling estimation model are interpolated by an S-shaped curve.

Fig. 8A is an explanatory diagram showing an example of a case where the estimation result of the first heating estimation model and the estimation result of the second heating estimation model are not interpolated by the S-shaped curve.

Fig. 8B is an explanatory diagram showing an example of a case where the estimation result of the first heating estimation model and the estimation result of the second heating estimation model are interpolated by an S-shaped curve.

Fig. 9 is an explanatory diagram showing an example of a distribution method of the detection values of the second feature quantity of the judgment model.

Fig. 10 is an explanatory diagram showing an example of abnormality detection from outliers.

Fig. 11 is a flowchart showing an example of a processing operation of the control circuit related to the estimation processing.

Fig. 12 is a flowchart showing an example of a processing operation of the control circuit related to the multiple regression analysis processing.

Fig. 13 is an explanatory diagram showing an example of the failure determination table in the control unit.

Fig. 14 is an explanatory diagram showing an example of the air conditioning system of embodiment 2.

Detailed Description

Embodiments of an air conditioning system, a method for estimating a refrigerant amount in an air conditioning system, an air conditioner, and a method for estimating a refrigerant amount in an air conditioner according to the present application will be described in detail below with reference to the accompanying drawings. However, the disclosed technology is not limited to the present embodiment. In addition, the respective embodiments shown below can be appropriately modified within a reasonable range.

Example 1

Structure of air conditioner

Fig. 1 is an explanatory diagram showing an example of an air conditioner 1 according to the present embodiment. The air conditioner 1 shown in fig. 1 includes one outdoor unit 2 and N indoor units 3, where N is a natural number of 2 or more. The outdoor unit 2 is connected in parallel to each indoor unit 3 via a liquid pipe 4 and an air pipe 5. The outdoor unit 2 and the indoor unit 3 are connected by refrigerant piping such as a liquid pipe 4 and an air pipe 5, thereby forming a refrigerant circuit 6 of the air conditioner 1.

Structure of outdoor unit

Fig. 2 is an explanatory diagram showing an example of the outdoor unit 2 and the N indoor units 3. The outdoor unit 2 includes: a compressor 11, a four-way valve 12, an outdoor heat exchanger 13, an outdoor expansion valve 14, a first stop valve 15, a second stop valve 16, a liquid reservoir 17, an outdoor fan 18, and a control circuit 19. The compressor 11, the four-way valve 12, the outdoor heat exchanger 13, the outdoor unit expansion valve 14, the first shutoff valve 15, the second shutoff valve 16, and the accumulator 17 are connected to each other by refrigerant pipes described in detail below, so that an outdoor side refrigerant circuit as a part of the refrigerant circuit 6 is formed.

The compressor 11 is a high-pressure-container-type variable-capacity compressor capable of changing the operating capacity according to driving of an electric motor, not shown, whose rotation speed is controlled by an inverter, for example. The refrigerant discharge side of the compressor 11 is connected to the first port 12A of the four-way valve 12 via a discharge pipe 21. The refrigerant suction side of the compressor 11 and the refrigerant outflow side of the accumulator 17 are connected by a suction pipe 22.

The four-way valve 12 is a valve for switching the flow direction of the refrigerant in the refrigerant circuit 6, and includes first to fourth valve ports 12A to 12D. The first valve port 12A is connected to the refrigerant discharge side of the compressor 11 via a discharge pipe 21. The second valve port 12B is connected to a side refrigerant inlet/outlet of the outdoor heat exchanger 13 through an outdoor refrigerant pipe 23. The third valve port 12C is connected to the refrigerant inflow side of the accumulator 17 through an outdoor refrigerant pipe 26. The fourth valve port 12D and the second shutoff valve 16 are connected by an outdoor air pipe 24.

The outdoor heat exchanger 13 exchanges heat between the refrigerant and the outside air sucked into the outdoor unit 2 by the rotation of the outdoor unit fan 18. The refrigerant inlet/outlet on one side of the outdoor heat exchanger 13 is connected to the second port 12B of the four-way valve 12 through an outdoor refrigerant pipe 26. The other side refrigerant inlet and outlet of the outdoor heat exchanger 13 is connected to the first shutoff valve 15 through an outdoor liquid pipe 25. The outdoor heat exchanger 13 functions as a condenser when the air conditioner 1 performs a cooling operation, and functions as an evaporator when the air conditioner 1 performs a heating operation.

The outdoor unit expansion valve 14 is provided in the outdoor liquid pipe 25, and is an electronic expansion valve driven by a pulse motor, not shown. The outdoor expansion valve 14 adjusts the opening degree according to the number of pulses supplied to the pulse motor, thereby adjusting the amount of refrigerant flowing into the outdoor heat exchanger 13 or the amount of refrigerant flowing out of the outdoor heat exchanger 13. When the air conditioner 1 is performing a heating operation, the opening degree of the outdoor unit expansion valve 14 is adjusted so that the degree of superheat of the refrigerant on the refrigerant suction side of the compressor 11 reaches the target degree of superheat of suction. The opening degree of the outdoor unit expansion valve 14 is fully opened when the air conditioner 1 performs a cooling operation.

The refrigerant inflow side of the accumulator 17 is connected to the third valve port 12C of the four-way valve 12 through an outdoor refrigerant pipe 26. Further, the refrigerant outflow side of the accumulator 17 and the refrigerant inflow side of the compressor 11 are connected by a suction pipe 22. The accumulator 17 separates the refrigerant flowing into the interior of the accumulator 17 from the outdoor refrigerant pipe 26 into a gas refrigerant and a liquid refrigerant, so that only the gas refrigerant is sucked into the compressor 11.

The outdoor fan 18 is formed of a resin material and is disposed in the vicinity of the outdoor heat exchanger 13. The outdoor unit fan 18 sucks outside air from an intake port, not shown, into the interior of the outdoor unit 2 based on rotation of a fan motor, not shown, and discharges outside air, which has exchanged heat with the refrigerant in the outdoor heat exchanger 13, from an exhaust port, not shown, to the outside of the outdoor unit 2.

Further, a plurality of sensors are disposed in the outdoor unit 2. The discharge pipe 21 is provided with a discharge pressure sensor 31 for detecting a discharge pressure, which is a pressure of the refrigerant discharged from the compressor 11, and a discharge temperature sensor 32 for detecting a discharge temperature, which is a temperature of the refrigerant discharged from the compressor 11. A suction pressure sensor 33 and a suction temperature sensor 34 are disposed near the refrigerant inflow port of the accumulator 17 of the outdoor refrigerant pipe 26, the suction pressure sensor 33 detecting the suction pressure, which is the pressure of the refrigerant sucked into the compressor 11, and the suction temperature sensor 34 detecting the temperature of the refrigerant sucked into the compressor 11.

The outdoor liquid pipe 25 between the outdoor heat exchanger 13 and the outdoor expansion valve 14 is provided with a refrigerant temperature sensor 35 for detecting the temperature of the refrigerant flowing into the outdoor heat exchanger 13 or the temperature of the refrigerant flowing out of the outdoor heat exchanger 13. An outside air temperature sensor 36 for detecting the outside air temperature, which is the temperature of the outside air flowing into the outdoor unit 2, is disposed near the intake port, not shown, of the outdoor unit 2.

The control circuit 19 is used to control the entire air conditioner 1. Fig. 3 is a block diagram showing an example of the control circuit 19 of the outdoor unit 2. The control circuit 19 includes: an acquisition unit 41, a communication unit 42, a storage unit 43, a control unit 44, an estimation unit 45, and a judgment unit 46. The acquisition unit 41 is configured to acquire the sensor values of the detection unit, which is the various sensors described above. The communication unit 42 is a communication interface for communicating with the communication unit of each indoor unit 3. The storage unit 43 is, for example, a flash memory, and is configured to store: the control program of the outdoor unit 2, the operation state amounts such as detection values corresponding to detection signals from various sensors, the driving states of the compressor 11 and the outdoor unit fan 18, the operation information (including, for example, operation/stop information, operation modes such as cooling/heating, etc.) transmitted from each indoor unit 3, the rated capacity of the outdoor unit 2, the required capacity of each indoor unit 3, and the like. Further, the storage unit 43 includes an abnormal record storage unit 43A for storing an abnormal record described later.

The control unit 44 acquires the detection values of the various sensors periodically (for example, every 30 seconds) via the communication unit 42, and a signal including the operation state amount transmitted from each indoor unit 3 is input to the control unit 44 via the communication unit 42. The control unit 44 adjusts the opening degree of the outdoor expansion valve 14 or controls the driving of the compressor 11 based on these various pieces of input information.

The estimating unit 45 includes an estimating model 45A that estimates the refrigerant shortage rate of the refrigerant circuit 6 using the detected value of the first characteristic amount when the operation state amount related to the refrigerant amount of the refrigerant circuit 6 is the first characteristic amount. In the present embodiment, as the amount of refrigerant remaining in the refrigerant circuit 6, for example, a relative amount of refrigerant is used. Specifically, the estimation model 45A is a model for estimating the refrigerant shortage ratio of the refrigerant circuit 6 (the amount of reduction relative to a predetermined amount when the amount of refrigerant filled is 100%, and the same applies hereinafter). The estimation model 45A includes: the first cooling estimation model 45A1, the second cooling estimation model 45A2, the third cooling estimation model 45A3, the first heating estimation model 45A4, the second heating estimation model 45A5, and the third heating estimation model 45A6. These respective estimation models will be described in detail later.

The determination unit 46 includes a determination model 46A that uses a second feature value detection value described later in the operation state quantity to determine whether or not the first feature value detection value is a detection value to be used for estimating the refrigerant shortage rate by the estimation unit 45. The judgment model 46A includes: a cooling time determination model 46B used when the air conditioner 1 performs a cooling operation, and a heating time determination model 46C used when the air conditioner 1 performs a heating operation. These respective judgment models will be described in detail later.

Structure of indoor unit

As shown in fig. 2, the indoor unit 3 includes: an indoor heat exchanger 51, an indoor unit expansion valve 52, a liquid pipe connection portion 53, an air pipe connection portion 54, and an indoor unit fan 55. The indoor heat exchanger 51, the indoor unit expansion valve 52, the liquid pipe connection portion 53, and the gas pipe connection portion 54 are connected to each other by refrigerant pipes described later, so that an indoor unit refrigerant circuit as a part of the refrigerant circuit 6 is configured.

The indoor heat exchanger 51 exchanges heat between the refrigerant and indoor air sucked into the indoor unit 3 from an intake port, not shown, by rotation of the indoor unit fan 55. The refrigerant inlet/outlet on one side of the indoor heat exchanger 51 is connected to the liquid pipe connection portion 53 via an indoor liquid pipe 56. The other side refrigerant inlet and outlet of the indoor heat exchanger 51 is connected to the gas pipe connection portion 54 via an indoor gas pipe 57. The indoor heat exchanger 51 functions as a condenser when the air conditioner 1 performs a heating operation. In contrast, when the air conditioner 1 performs a cooling operation, the indoor heat exchanger 51 functions as an evaporator.

The indoor unit expansion valve 52 is an electronic expansion valve, and is provided in the indoor liquid pipe 56. When the indoor heat exchanger 51 functions as an evaporator, that is, when the indoor unit 3 performs a cooling operation, the opening degree of the indoor unit expansion valve 52 is adjusted so that the degree of superheat of the refrigerant at the refrigerant outlet (the side of the gas pipe connection portion 54) of the indoor heat exchanger 51 becomes the target degree of superheat of the refrigerant. When the indoor heat exchanger 51 functions as a condenser, that is, when the indoor unit 3 performs a heating operation, the opening degree of the indoor unit expansion valve 52 is adjusted so that the degree of supercooling of the refrigerant at the refrigerant outlet (the liquid pipe connection portion 53 side) of the indoor heat exchanger 51 becomes the target degree of supercooling of the refrigerant. The target refrigerant superheat degree or the target refrigerant supercooling degree refers to a refrigerant superheat degree and a refrigerant supercooling degree required for the indoor unit 3 to exhibit sufficient cooling capacity or heating capacity.

The indoor unit fan 55 is formed of a resin material and is disposed in the vicinity of the indoor heat exchanger 51. The indoor unit fan 55 is driven to rotate by a fan motor, not shown, and sucks indoor air from an intake port, not shown, into the interior of the indoor unit 3, and discharges indoor air, which has exchanged heat with the refrigerant in the indoor heat exchanger 51, from an exhaust port, not shown, into the room.

The indoor unit 3 is provided with various sensors. The indoor liquid pipe 56 is provided with a liquid-side refrigerant temperature sensor 61 for detecting the temperature of the refrigerant flowing into the indoor heat exchanger 51 (indoor-side heat exchange inlet temperature during cooling operation) or the temperature of the refrigerant flowing out of the indoor heat exchanger 51 (indoor-side heat exchange outlet temperature during heating operation) between the indoor heat exchanger 51 and the indoor-side expansion valve 52. The indoor air pipe 57 is provided with an air-side temperature sensor 62 for detecting the temperature of the refrigerant flowing out of the indoor heat exchanger 51 (indoor-side heat exchange outlet temperature during cooling operation) or the temperature of the refrigerant flowing into the indoor heat exchanger 51 (indoor-side heat exchange inlet temperature during heating operation). A suction temperature sensor 63 for detecting a temperature of indoor air flowing into the indoor unit 3, that is, a suction temperature is disposed near an intake port, not shown, of the indoor unit 3.

Operation of refrigerant circuit

Next, the flow of the refrigerant in the refrigerant circuit 6 and the operation of each portion during the air conditioning operation of the air conditioner 1 according to the present embodiment will be described. The arrows in fig. 1 indicate the flow direction of the refrigerant during the heating operation.

When the air conditioner 1 is performing a heating operation, the four-way valve 12 is switched such that the first port 12A communicates with the fourth port 12D, and the second port 12B communicates with the third port 12C. Thus, the refrigerant circuit 6 is formed as a heating cycle in which each indoor heat exchanger 51 functions as a condenser and the outdoor heat exchanger 13 functions as an evaporator. For convenience of explanation, the flow direction of the refrigerant during the heating operation is indicated by solid arrows shown in fig. 2.

When the compressor 11 is driven while the refrigerant circuit 6 is in the above state, the refrigerant discharged from the compressor 11 flows through the discharge pipe 21, flows into the four-way valve 12, flows through the outdoor air pipe 24 from the four-way valve 12, and flows into the air pipe 5 via the second shutoff valve 16. The refrigerant flowing through the gas pipe 5 is branched to each indoor unit 3 via each gas pipe connection portion 54. The refrigerant flowing into each indoor unit 3 flows through each indoor air pipe 57 and then flows into each indoor heat exchanger 51. The refrigerant flowing into each indoor heat exchanger 51 exchanges heat with the indoor air sucked into each indoor unit 3 by the rotation of each indoor unit fan 55, and is condensed. That is, each of the indoor heat exchangers 51 functions as a condenser, and the indoor air heated by the refrigerant in each of the indoor heat exchangers 51 is blown out from an exhaust port, not shown, into the room, thereby heating the room in which each of the indoor units 3 is provided.

The refrigerant flowing from each indoor heat exchanger 51 into each indoor liquid pipe 56 flows through each indoor unit expansion valve 52 to be depressurized, wherein the opening degree of each indoor unit expansion valve 52 is adjusted so that the degree of supercooling of the refrigerant at the refrigerant outlet side of each indoor heat exchanger 51 becomes the target degree of supercooling of the refrigerant. Wherein the target refrigerant supercooling degree is determined based on the cooling capacity required in each indoor unit 3.

The refrigerant decompressed by the indoor unit expansion valves 52 flows out from the indoor liquid pipes 56 to the liquid pipe 4 via the liquid pipe connection portions 53. The refrigerant merged in the liquid pipe 4 flows into the outdoor unit 2 through the first shutoff valve 15. The refrigerant flowing into the first shutoff valve 15 of the outdoor unit 2 flows through the outdoor liquid pipe 25, and is depressurized by the outdoor unit expansion valve 14. The refrigerant decompressed by the outdoor expansion valve 14 flows through the outdoor liquid pipe 25, flows into the outdoor heat exchanger 13, exchanges heat with the outside air flowing in through the intake port, not shown, of the outdoor unit 2 by the rotation of the outdoor fan 18, and is evaporated. The refrigerant flowing out from the outdoor heat exchanger 13 to the outdoor refrigerant pipe 26 flows into the four-way valve 12, the outdoor refrigerant pipe 26, the accumulator 17, and the suction pipe 22 in this order, is sucked into the compressor 11, is compressed again, and flows out to the outdoor air pipe 24 through the first port 12A and the fourth port 12D of the four-way valve 12.

When the air conditioner 1 is performing the cooling operation, the four-way valve 12 is switched such that the first port 12A communicates with the second port 12B, and the third port 12C communicates with the fourth port 12D. Thus, the refrigerant circuit 6 is formed as a refrigeration cycle in which each indoor heat exchanger 51 functions as an evaporator, and the outdoor heat exchanger 13 functions as a condenser. For convenience of explanation, the flow of the refrigerant during the cooling operation is indicated by a broken-line arrow shown in fig. 2.

When the compressor 11 is driven in the state of the refrigerant circuit 6, the refrigerant discharged from the compressor 11 flows through the discharge pipe 21, flows into the four-way valve 12, flows through the outdoor refrigerant pipe 26 from the four-way valve 12, and flows into the outdoor heat exchanger 13. The refrigerant flowing into the outdoor heat exchanger 13 exchanges heat with the outdoor air sucked into the outdoor unit 2 by the rotation of the outdoor fan 18, and is condensed. That is, the outdoor heat exchanger 13 functions as a condenser, and the indoor air heated by the refrigerant in the outdoor heat exchanger 13 is blown out from an exhaust port, not shown, to the outside.

The refrigerant flowing from the outdoor heat exchanger 13 into the outdoor liquid pipe 25 flows through the outdoor expansion valve 14 whose opening degree is adjusted to be fully opened, and is depressurized. The refrigerant decompressed by the outdoor unit expansion valve 14 flows through the liquid pipe 4 via the first shutoff valve 15, and is split into the indoor units 3. The refrigerant flowing into each indoor unit 3 flows through each liquid pipe connection 53 from the indoor liquid pipe 56 and through the indoor unit expansion valve 52 to be depressurized, wherein the opening degree of the indoor unit expansion valve 52 is adjusted so that the degree of supercooling of the refrigerant at the refrigerant outlet of the indoor heat exchanger 51 becomes the target degree of supercooling of the refrigerant. The refrigerant decompressed by the indoor unit expansion valve 52 flows through the indoor liquid pipe 56 and flows into the indoor heat exchanger 51, exchanges heat with indoor air flowing in from a suction port, not shown, of the indoor unit 3 by rotation of the indoor unit fan 55, and is evaporated. That is, each of the indoor heat exchangers 51 functions as an evaporator, and the indoor air cooled by the refrigerant in each of the indoor heat exchangers 51 is blown out from an exhaust port, not shown, into the room, thereby cooling the room in which each of the indoor units 3 is provided.

The refrigerant flowing into the air pipe 5 from the indoor heat exchanger 51 via the air pipe connection portion 54 flows through the outdoor air pipe 24 via the second shutoff valve 16 of the outdoor unit 2, and flows into the fourth valve port 12D of the four-way valve 12. The refrigerant flowing into the fourth port 12D of the four-way valve 12 flows from the third port 12C into the refrigerant inflow side of the accumulator 17. The refrigerant flowing in from the refrigerant inflow side of the accumulator 17 is sucked into the compressor 11 through the suction pipe 22 and compressed again.

The acquisition unit 41 in the control circuit 19 is configured to acquire sensor values of the discharge pressure sensor 31, the discharge temperature sensor 32, the suction pressure sensor 33, the suction temperature sensor 63, the refrigerant temperature sensor 35, and the outside air temperature sensor 36 in the outdoor unit 2. Further, the acquisition unit 41 acquires sensor values of the liquid-side refrigerant temperature sensor 61, the gas-side temperature sensor 62, and the suction temperature sensor 63 of each indoor unit 3.

Fig. 4 is a mollier chart showing a refrigeration cycle of the air conditioner 1. In the cooling operation of the air conditioner 1, the outdoor heat exchanger 13 functions as a condenser, and the indoor heat exchanger 51 functions as an evaporator. In the heating operation of the air conditioner 1, the outdoor heat exchanger 13 functions as an evaporator, and the indoor heat exchanger 51 functions as a condenser.

The compressor 11 compresses the low-temperature low-pressure gas refrigerant flowing in from the evaporator into a high-temperature high-pressure gas refrigerant (refrigerant in the state of point B in fig. 4), and discharges the gas refrigerant. The temperature of the gas refrigerant discharged from the compressor 11 is a discharge temperature, and the discharge temperature is detected by a discharge temperature sensor 32.

The condenser exchanges heat between the high-temperature and high-pressure gas refrigerant from the compressor 11 and air, and condenses the gas refrigerant. At this time, in the condenser, the gas refrigerant is changed to a liquid refrigerant entirely by a latent heat change, and then the temperature of the liquid refrigerant is reduced by a sensible heat change, so that the liquid refrigerant is in a supercooled state (a state at point C in fig. 4). The temperature at which the gas refrigerant changes to the liquid refrigerant by the latent heat change is a high-pressure saturation temperature, and the temperature of the refrigerant in a supercooled state at the outlet of the condenser is a heat exchange outlet temperature. The high-pressure saturation temperature is a temperature corresponding to the pressure value (pressure value P2 indicated by "HPS" in fig. 4) detected by the discharge pressure sensor 31. The heat exchange outlet temperature is the temperature of the refrigerant flowing through the outdoor liquid pipe 25 detected by the refrigerant temperature sensor 35.

The expansion valve decompresses the low-temperature high-pressure refrigerant flowing out of the condenser to become a gas-liquid two-phase refrigerant (refrigerant in the state of point D in fig. 4) in which the gas and the liquid are mixed.

The evaporator exchanges heat between the gas-liquid two-phase refrigerant and air, and evaporates the gas-liquid two-phase refrigerant. At this time, in the evaporator, the gas-liquid two-phase refrigerant is changed to a gas refrigerant entirely by a latent heat change, and then the temperature of the gas refrigerant is raised to a superheated state (a state of a point a in fig. 4) by a sensible heat change, and is then sucked into the compressor 11. The temperature at which the liquid refrigerant changes to the gas refrigerant by the latent heat change is the low-pressure saturation temperature. The low pressure saturation temperature is a temperature corresponding to a pressure value (pressure value P1 indicated by "LPS" in fig. 4) detected by the suction pressure sensor 33. The temperature of the refrigerant that is superheated in the evaporator and then sucked into the compressor 11 is a suction temperature. The suction temperature is detected by a suction temperature sensor 34.

The degree of supercooling of the refrigerant in a supercooled state when it flows out from the condenser can be calculated by subtracting the refrigerant temperature at the refrigerant outlet of the heat exchanger functioning as the condenser (the heat exchange outlet temperature described above) from the high-pressure saturated temperature. The suction superheat of the refrigerant in a superheated state when it flows out from the evaporator can be calculated by subtracting the suction temperature from the low-pressure saturation temperature.

First characteristic quantity

Fig. 5 is an explanatory diagram showing an example of the first feature amount used in the first to third estimation models 45A1, 45A2, 45A3 for cooling and the second feature amount used in the judgment model 46B for cooling. As the operation state quantity used in the estimation model 45A, there is a first feature quantity. Examples of the first feature amounts used in the first to third refrigeration estimation models 45A1, 45A2, 45A3 include: the rotation speed of the compressor 11, the high-pressure saturation temperature, the suction temperature, the low-pressure refrigerant temperature, the refrigerant supercooling degree (outdoor heat exchange supercooling), and the outside air temperature. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11. The high pressure saturation temperature is a value obtained by converting the pressure value detected by the discharge pressure sensor 31 into a temperature value. The suction temperature is detected by a suction temperature sensor 34. The low-pressure refrigerant temperature is the temperature of the refrigerant sucked into the compressor 11 after being superheated in the evaporator. The refrigerant supercooling degree is, for example, a value calculated by (high pressure saturation temperature-outdoor heat exchange outlet temperature). The outside air temperature is detected by an outside air temperature sensor 36. The outdoor heat exchange outlet temperature is detected by the refrigerant temperature sensor 35. Further, for example, the operation state amounts including the first characteristic amounts used in the first to third cooling estimation models 45A1, 45A2, and 45A3 are periodically detected by detection units such as the rotation speed sensor, the discharge pressure sensor 31, the suction temperature sensor 34, the outside air temperature sensor 36, and the refrigerant temperature sensor 35. Further, while the air conditioner 1 is in operation, the control unit 44 instructs the detection unit to acquire the operation state quantity periodically (for example, every 10 minutes). The instructed detection unit detects the operation state quantity from various sensors provided in the air conditioner 1. The regularly acquired operation state quantity is also marked with acquisition time information.

Fig. 6 is an explanatory diagram showing an example of the first feature amount used in the first to third estimation models 45A4, 45A5, 45A6 for heating and the second feature amount used in the judgment model 46C for heating. Examples of the first feature amounts used in the first to third heating estimation models 45A4, 45A5, and 45A6 include: the opening degree of the outdoor unit expansion valve 14, the rotation speed of the compressor 11, the suction superheat degree, and the outside air temperature. The opening degree of the outdoor expansion valve 14 is the number of pulses of a stepping motor, not shown, applied to the outdoor expansion valve 14 by the control unit 44. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11. The suction superheat is, for example, a value calculated by (suction temperature-low pressure saturation temperature). The outside air temperature is detected by an outside air temperature sensor 36. The suction temperature is a value detected by the suction temperature sensor 34, and the low-pressure saturation temperature is a value obtained by converting the pressure value detected by the suction pressure sensor 33 into a temperature value. Further, for example, the operation state amounts including the first characteristic amounts used in the first to third heating estimation models 45A4, 45A5, and 45A6 are periodically detected by the detection units such as the rotation speed sensor, the intake temperature sensor 34, and the outside air temperature sensor 36.

Second characteristic quantity

As the operation state quantity used in the judgment model 46A, there is a second feature quantity. The second feature amount used for generating the determination model 46A is, for example, a value obtained when the refrigerant circuit 6 is reproduced on a computer and a numerical analysis (hereinafter, numerical analysis will also be referred to as simulation) is performed, and the operation of the refrigerant circuit 6 is normal and only the remaining refrigerant amount is changed. The second feature quantity used in the generation of the judgment model 46A is expressed as a simulation value (sometimes referred to as a "value"). The second feature quantity includes: at least one operation state quantity included in the first feature quantity, and at least one operation state quantity not included in the first feature quantity. As described later, the generated judgment model 46A is applied to the value of the second feature quantity detected by the detection section (hereinafter, also referred to as a detected value of the second feature quantity). The judgment model 46A is used to calculate an outlier of the detection value of the second feature quantity. The determination unit 46, which will be described later, determines whether or not the detection value of the first feature amount acquired by the detection unit at the same time as the detection value of the second feature amount is a detection value to be used for estimating the refrigerant shortage rate by the estimation unit 45, based on the value of the outlier.

The second feature quantity is detected by the detection unit at the same timing as the first feature quantity. Specifically, the control unit 44 instructs the detection unit to include the second feature amount in the operation state amounts acquired periodically (for example, every 10 minutes). As shown in fig. 5, examples of the second feature amount used in the cooling time determination model 46B include: the rotation speed of the compressor 11, the high pressure saturation temperature, the suction temperature, the low pressure refrigerant temperature, the outside air temperature, the discharge pressure, and the heat exchange outlet temperature. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11. The high pressure saturation temperature is a value obtained by converting the pressure value detected by the discharge pressure sensor 31 into a temperature value. The suction temperature is detected by a suction temperature sensor 34. The low-pressure refrigerant temperature is the temperature of the refrigerant sucked into the compressor 11 after being superheated in the evaporator. The outside air temperature is detected by an outside air temperature sensor 36. The discharge pressure is a pressure value detected by the discharge pressure sensor 31. The heat exchange outlet temperature is detected by the refrigerant temperature sensor 35.

As shown in fig. 6, the second feature amount used in the heating time determination model 46C is, for example: the outdoor unit expansion valve 14, the rotation speed of the compressor 11, the outside air temperature, the discharge temperature, the suction temperature, the low pressure saturation temperature, and the suction pressure (LPS). The opening degree of the outdoor unit expansion valve 14 is detected by a sensor not shown. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11. The outside air temperature is detected by an outside air temperature sensor 36. The discharge temperature is detected by a discharge temperature sensor 32. The suction temperature is detected by a suction temperature sensor 34. The low pressure saturation temperature is a value obtained by converting the pressure value detected by the suction pressure sensor 33 into a temperature value. The suction pressure is a pressure value detected by the suction pressure sensor 33. The operation state quantity including the second characteristic quantity used in the heating time determination model 46C is periodically detected by detection units such as the rotation speed sensor, the intake temperature sensor 34, the outside air temperature sensor 36, and the intake pressure sensor 33, for example.

The second characteristic amounts that are common to the cooling time determination model 46B and the heating time determination model 46C may be the rotation speed and the suction temperature of the compressor 11, which are the operation state amounts of the outdoor unit 2.

The second feature values that are common to the cooling time determination model 46B and the heating time determination model 46C may be, for example, operation state values on the indoor unit 3 side: the indoor-side heat exchange inlet temperature (in the cooling operation: detected by the liquid-side refrigerant temperature sensor 61/detected by the air-side temperature sensor 62), the indoor-side heat exchange outlet temperature (in the cooling operation: detected by the air-side temperature sensor 62/detected by the liquid-side refrigerant temperature sensor 61/detected by the air-side temperature sensor 62/detected by the air-side temperature sensor 61), and the opening degree of the indoor-unit expansion valve 52. As examples of the second characteristic amount on the indoor unit 3 side, for example, the indoor-unit-side heat exchange inlet temperature, the indoor-unit-side heat exchange outlet temperature, and the opening degree of the indoor-unit expansion valve 52 are illustrated, and the characteristic amounts that can be obtained in common even when the indoor units 3 are different in type such as duct type and ceiling-embedded type.

Structure of estimation model

The estimation model 45A is generated using the detected value of the first feature quantity. The estimating unit 45 applies the detection value of the first characteristic amount acquired at a timing different from the timing at which the estimation model 45A is generated to the estimation model 45A, thereby estimating the refrigerant shortage ratio of the refrigerant circuit 6.

The estimation model 45A is generated by one of regression analysis methods, that is, a multiple regression analysis method, using any of a plurality of operation state amounts (detection values of the first feature amounts). The multiple regression analysis method is to select a regression equation having the smallest P value (a preset weight parameter) indicating the degree of influence of the operation state quantity on the accuracy of the generated estimation model) and the largest possible correction value R2 (a value indicating the accuracy of the generated estimation model 45A) from among regression equations obtained based on a plurality of simulation results (a result obtained by numerically calculating the refrigerant circuit 6 and calculating what the operation state quantity is with respect to the remaining refrigerant quantity) to generate the estimation model 45A. The P value and the correction value R2 are values related to the accuracy of the estimation model 45A when the estimation model 45A is generated by the multiple regression analysis method, and the smaller the P value is, the closer the correction value R2 is to 1.0, the higher the accuracy of the generated estimation model 45A is. As a result, when the refrigerant shortage rate at the time of cooling is 0 to 30%, for example, the operation state amounts such as the degree of supercooling of the refrigerant, the outside air temperature, the high-pressure saturation temperature, and the rotation speed of the compressor 11 are set as the first characteristic amounts. When the refrigerant shortage ratio at the time of cooling is 40 to 70%, for example, the suction temperature, the outside air temperature, the rotation speed of the compressor 11, and other operation state amounts are used as the first characteristic amount. When the refrigerant shortage rate at the time of heating is 0 to 20%, for example, the opening degree of the outdoor unit expansion valve 14 is used as a characteristic amount as an operation state amount. In the case where the refrigerant shortage ratio at the time of heating is 30% to 70%, for example, the suction superheat (suction temperature-low pressure saturation temperature), the outside air temperature, the rotation speed of the compressor 11, the operation state quantity of the outdoor unit expansion valve 14, and the like are used as the first characteristic quantity.

As described above, the estimation model 45A includes: the first cooling estimation model 45A1, the second cooling estimation model 45A2, the third cooling estimation model 45A3, the first heating estimation model 45A4, the second heating estimation model 45A5, and the third heating estimation model 45A6. In the present embodiment, each of the above-described estimation models is generated using a simulation result described later, and is stored in advance in the estimation unit 45 in the control circuit 19 of the air conditioner 1.

The first estimation model 45A1 for cooling is an estimation model 45A effective when the refrigerant shortage ratio is 0% to 30% (first range), and is a first regression equation that can estimate the refrigerant shortage ratio with high accuracy. The first regression equation is, for example, (α1×refrigerant supercooling degree) + (α2×outside air temperature) + (α3×high-pressure saturation temperature) + (α4×rotation speed of the compressor 11) +α5. The coefficients α1 to α5 are determined when the estimation model is generated. The estimating unit 45 calculates the current refrigerant shortage ratio of the refrigerant circuit 6 by substituting the current refrigerant supercooling degree, the outside air temperature, the high-pressure saturation temperature, and the rotation speed of the compressor 11 acquired by the acquiring unit 41 into the first regression equation. The reasons for substituting the refrigerant supercooling degree, the outside air temperature, the high-pressure saturation temperature, and the rotation speed of the compressor 11 are the first feature amount used for generating the first cooling estimation model 45 A1. The degree of supercooling of the refrigerant can be calculated, for example, by (high pressure saturation temperature-heat exchange outlet temperature). The outside air temperature is detected by an outside air temperature sensor 36. The high pressure saturation temperature is a value obtained by converting the pressure value detected by the discharge pressure sensor 31 into a temperature value. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11.

The second estimation model 45A2 for cooling is an estimation model 45A effective when the refrigerant shortage ratio is 40% to 70% (second range), and is a second regression equation that can estimate the refrigerant shortage ratio with high accuracy. The second regression equation is, for example, (α11×suction temperature) + (α12×outside air temperature) + (α13×rotation speed of the compressor 11) +α14. The coefficients α11 to α14 are determined when the estimation model is generated. The estimating unit 45 calculates the current refrigerant shortage ratio of the refrigerant circuit 6 by substituting the current suction temperature, the outside air temperature, and the rotation speed of the compressor 11 acquired by the acquiring unit 41 into the second regression equation. The reason why the suction temperature, the outside air temperature, and the rotation speed of the compressor 11 are substituted is to use the feature amount used in the generation of the second cooling estimation model 45 A2. The suction temperature is detected by a suction temperature sensor 34. The outside air temperature is detected by an outside air temperature sensor 36. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11.

On the other hand, as described above, the refrigerant shortage ratio that can be found by the first regression equation is 0% to 30%, and the refrigerant shortage ratio that can be found by the second regression equation is 40% to 70%. In this case, when the refrigerant shortage ratio is 30% to 40%, the refrigerant shortage ratio is calculated to be 30% using the first regression equation, and 40% using the second regression equation. That is, when the refrigerant shortage ratio is 30% to 40%, the degree of supercooling of the refrigerant, which is high in contribution when the refrigerant shortage ratio is 30% or less, and the suction temperature, which is high in contribution when the refrigerant shortage ratio is 40% or more, are each less changed, and therefore an effective estimated model cannot be generated. Therefore, if the first regression equation or the second regression equation is used, the difference in the refrigerant deficiency is large depending on which model is used, as shown in fig. 7A.

The third estimation model 45A3 for cooling is a cooling-time refrigerant shortage ratio calculation formula that can cover a range of 0% to 70% of the refrigerant shortage ratio, and the coverage includes a range in which the refrigerant shortage ratio cannot be estimated using either the first regression equation or the second regression equation described above. As shown in fig. 7B, the refrigerant shortage ratio calculation formula at the time of cooling is a calculation formula for continuously connecting the refrigerant shortage ratio as the estimation result of the first regression equation and the refrigerant shortage ratio as the estimation result of the second regression equation by an S-type curve using an S-type coefficient. Specifically, the refrigerant shortage ratio calculation formula at the time of cooling is: (refrigerant shortage ratio obtained by the S-type coefficient×first regression equation) + (refrigerant shortage ratio obtained by the (1-S-type coefficient) ×second regression equation). The estimating unit 45 calculates the refrigerant shortage ratio by substituting the current operation state quantity acquired by the acquiring unit 41 into the first regression equation and the second regression equation, and calculates the current refrigerant shortage ratio of the refrigerant circuit 6 by substituting the refrigerant shortage ratio into the refrigerant shortage ratio calculation formula during cooling.

Wherein the calculation of the S-type coefficient uses any one of the running state amounts. In the present embodiment, considering that the result based on the first regression equation is almost unchanged when the supercooling degree is 0, a calculation formula in which the S-type coefficient is 0.5 when the supercooling degree is 5 ℃ is set.

p＝1/(1+exp(-(sc-5)))

And p: s-shaped coefficient

sc: supercooling value

Thus, by determining the S-type coefficient and using it for the third cooling estimation model 45A3, the estimated value of the first cooling estimation model 45A1 is dominant among the estimated values based on the third cooling estimation model 45A3 when the refrigerant shortage rate is 0% to 30%, that is, when the refrigerant shortage rate is in the first range, and the estimated value of the second cooling estimation model 45A2 is dominant among the estimated values based on the third cooling estimation model 45A3 when the refrigerant shortage rate is 40% to 70%, that is, when the refrigerant shortage rate is in the second range.

The calculation of the S-type coefficient is not limited to the above method, and the S-type coefficient may be determined so that the estimated value of the second cooling estimation model 45A2 out of the estimated values based on the third cooling estimation model 45A3 is dominant when the actual refrigerant shortage ratio is 30% or more, that is, when the actual refrigerant shortage ratio is not within the first range, and the estimated value of the first cooling estimation model 45A1 out of the estimated values based on the third cooling estimation model 45A3 is dominant when the actual refrigerant shortage ratio is 40% or less, that is, when the actual refrigerant shortage ratio is not within the second range.

The first estimation model 45A4 for heating is an estimation model 45A effective when the refrigerant shortage ratio is 0% to 20% (third range), and is a fourth regression equation that can estimate the refrigerant shortage ratio with high accuracy. The fourth regression equation is, for example, (α31×the opening degree of the outdoor unit expansion valve 14) +α32. The estimating unit 45 calculates the refrigerant shortage ratio by substituting the current opening degree of the outdoor unit expansion valve 14 acquired by the acquiring unit 41 into a fourth regression equation. The reason for substituting the opening degree of the outdoor unit expansion valve 14 is to use the feature amount used in the generation of the first heating estimation model 45 A4.

The second estimation model 45A5 for heating is an estimation model 45A effective when the refrigerant shortage ratio is 30% to 70% (fourth range), and is a fifth regression equation that can estimate the refrigerant shortage ratio with high accuracy. The fifth regression equation is, for example, (α41×suction superheat) + (α42×outside air temperature) + (α43×rotation speed of the compressor 11) + (α44×opening degree of the outdoor unit expansion valve 14) +α45. The coefficients α41 to α45 are determined when the estimation model is generated. The estimating unit 45 calculates the current refrigerant shortage ratio of the refrigerant circuit 6 by substituting the current suction superheat, the outside air temperature, the rotation speed of the compressor 11, and the opening degree of the main-side expansion valve, which are acquired by the acquiring unit 41, into a fifth regression equation. The reason why the suction superheat degree, the outside air temperature, the rotation speed of the compressor 11, and the opening degree of the outdoor unit expansion valve 14 are substituted is to use the feature amount used in the generation of the second heating estimation model 45 A5. The suction superheat degree can be calculated, for example, by (suction temperature-low pressure saturation temperature). The outside air temperature is detected by an outside air temperature sensor 36. The rotation speed of the compressor 11 is detected by a rotation speed sensor, not shown, of the compressor 11. The opening degree of the outdoor unit expansion valve 14 is detected by a sensor not shown.

Further, as described above, the refrigerant shortage ratio that can be found by the fourth regression equation is 0% to 20%, and the refrigerant shortage ratio that can be found by the fifth regression equation is 30% to 70%. In this case, when the refrigerant shortage ratio is 20% to 30%, the refrigerant shortage ratio is calculated to be 20% using the fourth regression equation, and the refrigerant shortage ratio is calculated to be 30% using the fifth regression equation. That is, when the refrigerant shortage ratio is 20% to 30%, the opening degree of the outdoor unit expansion valve 14 that is high in contribution when the refrigerant shortage ratio is 20% or less and the suction superheat degree that is high in contribution when the refrigerant shortage ratio is 30% or more are both small in change, and an effective estimated model cannot be generated. Therefore, if the fourth regression equation or the fifth regression equation is used, the difference in the refrigerant deficiency is large depending on which model is used, as shown in fig. 8A.

The third estimation model 45A6 for heating is a calculation formula of the refrigerant shortage ratio at the time of heating, which can cover the range of 0% to 70% of the refrigerant shortage ratio, and the coverage range includes a range in which the refrigerant shortage ratio cannot be estimated by using either one of the fourth regression equation and the fifth regression equation described above. As shown in fig. 8B, the heating-time refrigerant shortage ratio calculation formula is a calculation formula for continuously connecting the refrigerant shortage ratio as the estimation result of the fourth regression equation and the refrigerant shortage ratio as the estimation result of the fifth regression equation by an S-type curve using an S-type coefficient. Specifically, the refrigerant shortage ratio calculation formula at the time of heating is: (refrigerant shortage ratio obtained by the S-type coefficient×fifth regression equation) + (refrigerant shortage ratio obtained by the (1-S-type coefficient) ×fourth regression equation). The estimating unit 45 calculates the current refrigerant shortage ratio of the refrigerant circuit 6 by substituting the current operation state quantity acquired by the acquiring unit 41 into the fourth regression equation and the fifth regression equation, respectively, and substituting the refrigerant shortage ratio into the heating-time refrigerant shortage ratio calculation formula.

Here, as in the case of the cooling operation, any one of the operation state amounts is used for the calculation of the S-type coefficient. In the present embodiment, it is considered that, when the opening degree of the outdoor unit expansion valve 14 is made to be completely closed: 0/fully open: when 100 is set, the result of the fourth regression equation is almost unchanged if the opening degree of the outdoor expansion valve 14 is fully opened, and therefore, a calculation formula is set such that the S-type coefficient is 0.5 when the opening degree of the outdoor expansion valve 14 is 90.

p＝1/(1+exp(-(D/10-45)))

And p: s-shaped coefficient

D: opening degree of outdoor unit expansion valve 14

Thus, by determining the S-type coefficient and using it for the third estimated model 45A6, the estimated value of the first estimated model 45A4 for heating is dominant among the estimated values based on the third estimated model 45A6 for heating when the refrigerant shortage rate is 0% to 20%, that is, when the refrigerant shortage rate is in the third range, and the estimated value of the second estimated model 45A5 for heating is dominant among the estimated values based on the third estimated model 45A6 when the refrigerant shortage rate is 30% to 70%, that is, when the refrigerant shortage rate is in the fourth range.

The calculation of the S-type coefficient is not limited to the above method, and the S-type coefficient may be determined so that the estimated value of the second estimated heating model 45A5 among the estimated values based on the third estimated heating model 45A6 is dominant when the actual refrigerant shortage ratio is 20% or more, that is, when the actual refrigerant shortage ratio is not within the third range, and the estimated value of the first estimated heating model 45A4 among the estimated values based on the third estimated heating model 45A6 is dominant when the actual refrigerant shortage ratio is 30% or less, that is, when the actual refrigerant shortage ratio is not within the fourth range.

As described above, during the cooling operation, the refrigerant shortage ratio is estimated using the first regression equation, the second regression equation, and the refrigerant shortage ratio calculation formula during cooling. When the degree of supercooling of the refrigerant at the time of cooling is a value greater than the first threshold (Tv 1 of fig. 7), the selection of the first regression equation enables the refrigerant shortage ratio to be estimated with higher accuracy than the selection of the second regression equation. Further, in the case where the degree of supercooling of the refrigerant at the time of cooling is a value smaller than the first threshold value, the selection of the second regression equation enables the refrigerant shortage ratio to be estimated with higher accuracy than the selection of the first regression equation. When the degree of supercooling of the refrigerant during cooling is a value in the vicinity of the first threshold value, the estimated value of the refrigerant shortage ratio is greatly changed according to which regression equation is used. Therefore, at the time of cooling, a cooling-time refrigerant shortage ratio calculation formula including the first regression equation and the second regression equation is selected. This makes it possible to estimate the refrigerant shortage rate at the time of cooling with high accuracy.

In addition, at the time of heating operation, the refrigerant shortage ratio is estimated using the fourth regression equation, the fifth regression equation, and the calculation formula of the refrigerant shortage ratio at the time of heating. When the opening degree of the outdoor unit expansion valve 14 at the time of heating is smaller than the second threshold value (Tv 2 in fig. 8), the refrigerant shortage ratio can be estimated with higher accuracy by selecting the fourth regression equation than by selecting the fifth regression equation. When the opening degree of the outdoor unit expansion valve 14 during heating is equal to or greater than the second threshold value, the refrigerant shortage ratio can be estimated with higher accuracy by selecting the fifth regression equation than by selecting the fourth regression equation. When the opening degree of the outdoor unit expansion valve 14 at the time of heating is a value in the vicinity of the first threshold value, the estimated value of the refrigerant shortage ratio is greatly changed according to which regression equation is used. Therefore, at the time of heating, a heating-time refrigerant shortage ratio calculation formula including the fourth regression equation and the fifth regression equation is selected. This makes it possible to estimate the refrigerant shortage rate at the time of heating with high accuracy.

Structure of judgment model

The determination model 46A is generated using a simulation value that is a value of a second characteristic amount, which is a value obtained based on a result of simulation of the operation of the refrigerant circuit 6 when the operation of the refrigerant circuit 6 is normal and only the remaining refrigerant amount is changed. The determination unit 46 calculates an outlier by applying the detection value of the second feature amount acquired from the operating air conditioner 1 to the determination model 46A. The determination unit 46 determines, based on the value of the outlier, whether or not the detection value of the first feature amount acquired by the detection unit at the same time as the detection value of the second feature amount is a detection value to be used for the estimation of the refrigerant shortage ratio by the estimation unit 45.

For example, a nuclear density estimation method is used for generating the judgment model 46A. The kernel density estimation method is a method of estimating a density function of the whole distribution based on a limited number of sample points. The judgment model 46A calculates the degree of deviation (hereinafter, also referred to as outlier) from the maximum value of the density function (the center of the Cluster (set of data having similarity)) based on the density function of the whole distribution estimated from the limited sample points. Then, the judgment model 46A calculates an outlier of the data to be judged after the data to be judged is input, and judges whether or not the outlier is within a predetermined range (whether or not the data to be judged is included in the cluster).

Fig. 9 is an explanatory diagram showing an example of the distribution of the detection values of the second feature quantity. As shown in fig. 9, the judgment model 46A classifies a set of values of the second feature quantity (hereinafter, also referred to as "simulated values of the second feature quantity") obtained by the simulation as one cluster and classifies the set as normal. The simulation conditions are that the refrigerant circuit 6 is in a steady state (a state in which the amount of refrigerant charged is a predetermined amount) or a state in which the amount of refrigerant charged is reduced (a refrigerant leakage state). The simulation value of the second characteristic amount in the steady state is a value of the second characteristic amount obtained when the simulation is performed in a state where the elements (the refrigerant circuit 6, the compressor, the expansion valve, and the like) constituting the air conditioner 1 are normally operated. The simulated value of the second characteristic amount in the refrigerant leakage state is a value of the second characteristic amount obtained when the simulation is performed in a state where each element (refrigerant circuit 6, compressor, expansion valve, etc.) constituting the air conditioner is normally operated and only the amount of refrigerant remaining in refrigerant circuit 6 is changed (reduced). When a detection value of the second feature quantity that does not belong to the cluster classified as normal by the judgment model 46A is input, the detection value is classified as abnormal. In addition, as shown in fig. 9, when the detection values are plotted on the graph, the detection values classified as abnormal are detection values that deviate from the clusters classified as normal. The abnormality is a state in which the possibility of failure of the device constituting the refrigerant circuit 6 is high.

The judgment model 46A applies the detection value of the second feature quantity acquired from the operating air conditioner 1 to calculate an outlier. Specifically, the judgment model 46A calculates an outlier indicating the degree of deviation from the detected value of the second feature amount acquired by the acquisition unit 41 of the air conditioner 1 during operation, using the value of the second feature amount used when the judgment model 46A is generated as a normal sample value (classified as a normal cluster). An outlier is a value obtained by converting a distance, which is a distance from the center of a cluster classified as normal, into a numerical value, and the degree of deviation is higher as the absolute value of the numerical value is larger. As a result, as the degree of deviation increases, the probability that the detected value of the second feature amount is abnormal increases.

Fig. 10 is an explanatory diagram showing an example of abnormality detection from outliers. The determination unit 46 classifies the detected value of the second feature as normal when the outlier of the detected value of the second feature is, for example, greater than "-150" (when the absolute value of the outlier is less than 150), and classifies the detected value of the second feature as abnormal when the outlier of the detected value of the second feature is, for example, equal to or less than "-150" (when the absolute value of the outlier is equal to or greater than 150). The deviation threshold X can be determined, for example, based on the result of verifying the value actually determined to be abnormal after collecting the failure history of the air conditioner 1, to a value that does not misdetermine the normal data as being abnormal. Therefore, when the detected value of the second feature amount is classified as abnormal, the determination unit 46 does not execute the operation of estimating the refrigerant shortage ratio by the estimation unit 45 using the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount. Further, the determination unit 46 stores the detected value of the second feature amount classified as abnormal as an abnormal record in the abnormal record storage unit 43A.

The determination unit 46 classifies the detected value of the second feature amount as normal when the absolute value of the estimated outlier is smaller than the deviation threshold X, for example, smaller than 150. In this case, the determination unit 46 uses the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount to execute the operation of estimating the refrigerant shortage rate by the estimation unit 45.

For convenience of explanation, the deviation threshold value X is set to "-150", for example, but may be appropriately adjusted based on the result of verifying the value actually determined to be abnormal after collecting the failure history.

Action of estimation processing

Fig. 11 is a flowchart showing an example of the processing operation of the control circuit 19 relating to the estimation process. The estimation unit 45 in the control circuit 19 is also configured to hold: the first cooling estimation model 45A1, the second cooling estimation model 45A2, the third cooling estimation model 45A3, the first heating estimation model 45A4, the second heating estimation model 45A5, and the third heating estimation model 45A6, which are generated in advance. Further, the judgment unit 46 in the control circuit 19 holds a pre-generated cooling judgment model 46B and heating judgment model 46C. For example, the estimation process is periodically performed on the operation state amounts of 24 hours in total every 10 minutes sequentially detected by the detection unit in a preset period (for example, at night) once a day. Further, the night time is shown as an example of the preset period, and for example, the operation state quantity of one day may be acquired after the operation of the air conditioner 1 is stopped at the night time of the period in which the operation frequency of the air conditioner 1 is small. Further, as the preset time period, the preset time period for the non-operation may be determined not at night but, for example, according to the operation state of the air conditioner 1 for one month.

In fig. 11, the control unit 44 in the control circuit 19 collects the operation state quantity as operation data by the acquisition unit 41 (step S11). The control unit 44 performs data filtering processing of extracting an arbitrary operation state quantity from the collected operation data (step S12). The control unit 44 executes data cleansing processing (step S13). Further, the judgment section 46 performs a judgment process of classifying the detected value of the second feature quantity after the data cleaning process is performed as normal or abnormal using the judgment model 46A (step S14).

The control unit 44 determines whether the detected value of the second feature quantity is classified as normal or abnormal (step S15). The estimating unit 45 performs the remaining refrigerant amount estimating process in the case where the detected value of the second feature amount is classified as normal in step S15, that is, applies the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount classified as normal to each of the estimation models (step S16). Then, the estimating unit 45 calculates the refrigerant shortage ratio of the refrigerant circuit 6 (step S17), and ends the processing operation shown in fig. 11.

In addition, when the detected value of the second feature amount is classified as abnormal in step S15, the determination unit 46 performs an abnormal output process of storing the detected value of the second feature amount classified as abnormal in the abnormal record storage unit 43A and outputting an alarm (step S18), and ends the processing operation shown in fig. 11.

The data filtering process extracts only a part of the operation state amounts (the detected value of the first feature amount and the detected value of the second feature amount) necessary for performing the judgment process of the second feature amount and calculating the refrigerant shortage rate, from among the plurality of operation state amounts, based on the preset filtering condition, instead of using all of the plurality of operation state amounts. By substituting the detected values of the first feature amount and the second feature amount (from which the abnormal value or the escape value is removed) after the data filtering process described later into the generated estimation model 45A or the judgment model 46A, it is possible to more accurately perform the judgment using the second feature amount or the estimation of the refrigerant shortage ratio using the first feature amount.

The preset filtering conditions comprise a first filtering condition, a second filtering condition and a third filtering condition. The first filter condition is, for example, a filter condition of data extracted in all operation modes of the air conditioner 1. The second filtering condition is a filtering condition of data extracted during the cooling operation. The third filtering condition is the filtering condition of the data extracted during the heating operation.

The first filtering condition is, for example: the driving state of the compressor 11, the identification of the operation mode, the elimination of the special operation, the elimination of the defective value among the obtained values, the selection of the value having a small variation of the operation state quantity which has a large influence in generating each regression equation, and the like. The driving state of the compressor 11 is a condition that needs to be judged because the refrigerant shortage rate cannot be estimated unless the compressor is stably operated so that the refrigerant circulates in the refrigerant circuit 6, and the driving state of the compressor 11 is a filter condition set for removing the operation state quantity detected at the transient period at the time of starting the compressor 11 or the like.

The operation mode is identified as a filter condition for extracting only the operation state amounts acquired during the cooling operation and the heating operation. Therefore, the operation state quantity acquired at the time of the dehumidification operation or the air blowing operation is removed. The exclusion of the special operation is a filter condition for removing the operation state quantity acquired during the special operation, and the special operation is an operation in which the state of the refrigerant circuit 6 is greatly different from that during the cooling operation or the heating operation, such as the refrigerating oil recovery operation or the defrosting operation. The elimination of the defect value is a filter condition for eliminating the operation state quantity including the defect value, because the accuracy may be lowered by generating each regression equation using the operation state quantity when the defect value exists in the operation state quantity for determining the refrigerant shortage rate.

The value having a small amount of change in the operation state quantity substituted into each regression equation or each refrigerant shortage ratio calculation equation is a necessary condition for extracting only the filter condition of the operation state quantity when the operation state of the air conditioner 1 is in the steady state in order to improve the estimation accuracy based on each regression equation and each refrigerant shortage ratio calculation equation. Further, the operation state amounts greatly affected are, for example: the refrigerant supercooling degree used when the refrigerant shortage rate in the cooling operation is 0 to 30%, the suction temperature used when the refrigerant shortage rate in the cooling operation is 40 to 70%, the suction superheat degree in the heating operation, and the like.

The second filtering conditions include, for example, elimination of the heat exchange outlet temperature, abnormality of the supercooling degree, abnormality of the discharge temperature, and the like.

The exclusion of the heat exchange outlet temperature is a filtering condition for removing the heat exchange outlet temperature lower than the outside air temperature, considering that the heat exchange outlet temperature detected by the heat exchange outlet temperature sensor 35 is not lower than the outside air temperature detected by the outside air temperature sensor 36 at the time of the cooling operation because the outside air temperature sensor 36 and the heat exchange outlet temperature sensor 35 are disposed in a close position.

The supercooling degree abnormality is a filtering condition in which a particularly high or low supercooling degree of a refrigerant is removed when it is detected due to a maximum or minimum refrigeration load. The abnormality in the discharge temperature is a filtering condition for removing the discharge temperature detected in a so-called gas shortage state in which the amount of refrigerant sucked into the compressor 11 is reduced due to a small refrigeration load.

The third filtering condition is, for example, abnormality of the discharge temperature. When the discharge temperature protection control is executed because the discharge temperature increases due to a large heating load during the heating operation, the discharge temperature is lowered by lowering the rotation speed of the compressor 11, for example, so that the detected discharge temperature is removed from the filter condition.

The data cleaning process is a process for removing the detected values of the first feature quantity that are at risk of erroneous estimation, and not using all the detected values of the first feature quantity that are acquired for estimating the refrigerant shortage rate. The data cleaning process is a process for removing the detected values of the second feature amounts that are at risk of erroneous determination, and not using all the detected values of the second feature amounts that are acquired for the determination process. Specifically, smoothing the acquired operation state quantity to suppress interference (noise), limiting the data amount, and the like are included. The smoothing of the data to suppress the disturbance is a process of calculating an average value of the section to obtain a moving average line of, for example, the degree of supercooling of the refrigerant, the suction temperature, and the degree of suction superheat in each model, thereby suppressing the disturbance. Limiting the amount of data refers to, for example, processing of removing data having a smaller amount of data due to its lower reliability. For example, if the number of data left after the filtering process is X or more for one-day input data, the data is used for the estimation of the refrigerant shortage ratio or the judgment process of the second characteristic amount, and if the number is less than X, all the data on that day is not used. That is, in the data cleaning process, by substituting the operation state quantity from which the abnormal value and the escape value are removed into the estimation model 45A, the refrigerant shortage ratio can be estimated more accurately; by substituting the operation state quantity from which the abnormal value and the escape value are removed into the judgment model 46A, more accurate judgment of the second feature quantity can be made.

The judgment processing is processing for calculating a degree of deviation (outlier) from a maximum value (center of cluster) of the density function based on the density function of the overall distribution estimated from the analog value of the second feature quantity, and judging whether or not the outlier is within a predetermined range (whether or not data to be judged is included in the cluster). The judgment model 46A applies the detection value of the second feature quantity acquired from the operating air conditioner 1 to calculate an outlier. In the judgment process, the value of the second feature amount used when the judgment model 46A is generated is set to a normal sample value, whereby an outlier of the detection value of the second feature amount is calculated. Further, in the judgment processing, when the absolute value of the calculated outlier is equal to or more than the absolute value of the deviation threshold X, the detected value of the second feature quantity is classified as abnormal. Further, in the judgment process, when the absolute value of the calculated outlier is smaller than the absolute value of the deviation threshold X, the detected value of the second feature quantity is classified as normal.

Fig. 12 is a flowchart showing an example of the processing operation of the control circuit 19 relating to the remaining refrigerant amount estimation processing. The estimation of the remaining refrigerant amount is as follows: the detected value of the first feature quantity obtained simultaneously with the detected value of the second feature quantity classified as normal in the judgment process, out of the current operation state quantity (sensor value) after the data filtering process and the data cleaning process, is substituted into each regression equation or each refrigerant shortage ratio calculation formula of the estimation model 45A to calculate the current refrigerant shortage ratio of the refrigerant circuit 6. In fig. 12, the estimating unit 45 in the control circuit 19 determines whether or not the acquired first feature amount is the feature amount acquired during the cooling operation (step S21). When the acquired first feature quantity is the feature quantity acquired in the cooling operation (yes in step S21), the estimating unit 45 applies the first feature quantity to each of the first to third cooling estimation models 45A1 to 45A3 (step S22).

When the acquired first feature quantity is not the feature quantity acquired in the cooling operation (no in step S21), that is, when the acquired first feature quantity is the feature quantity acquired in the heating operation, the estimating unit 45 applies the first feature quantity to each of the first to third heating estimation models 45A4 to 45A6 (step S23). Then, the estimating unit 45 calculates the current refrigerant shortage ratio by integrating the results obtained by applying the first feature amounts to the first to third cooling estimation models 45A1 to 45A3 and the results obtained by applying the first feature amounts to the first to third heating estimation models 45A4 to 45A6, respectively (step S24), and ends the processing operation shown in fig. 12.

The abnormality output process stores the detected value of the second feature amount classified as abnormal in the judgment process as an abnormality record in the abnormality record storage unit 43A and outputs an alarm. As a result, an abnormality in the detection value of the second feature amount can be identified.

Fault judging method

The estimating unit 45 calculates the current refrigerant shortage ratio of the refrigerant circuit 6, and notifies the control unit 44 of the calculated refrigerant shortage ratio. Further, the judgment section 46 classifies the current second feature quantity as normal or abnormal, and notifies the classification result to the control section 44. The control unit 44 determines whether the refrigerant amount is abnormal or normal based on the refrigerant shortage ratio calculated by the estimating unit 45, and outputs the determination result as a refrigerant amount determination result. The control unit 44 outputs the state of the air conditioner 1 as a determination result based on the refrigerant amount determination result and the classification result of the determination unit 46. Fig. 13 is an explanatory diagram showing an example of the failure determination table 44A in the control unit 44.

The control unit 44 refers to the failure determination table 44A, determines that the detected refrigerant leakage is caused by another failure when the refrigerant amount determination result is abnormal and the classification result of the determination unit 46 is abnormal, and outputs an alarm indicating the determination. The control unit 44 refers to the failure determination table 44A, determines that the refrigerant leakage is detected when the refrigerant amount determination result is abnormal and the classification result of the determination unit 46 is normal, and outputs an alarm indicating the determination. Further, the control unit 44 refers to the failure determination table 44A, determines that a failure other than the refrigerant leakage is detected when the refrigerant amount determination result is normal and the classification result of the determination unit 46 is abnormal, and outputs an alarm indicating the determination. The control unit 44 refers to the failure determination table 44A, and determines that the state is stable when the braking dose determination result is normal and the classification result by the determination unit 46 is normal.

Effect of example 1

In the air conditioner 1 of embodiment 1, the outlier of the detected value of the second feature amount is calculated by setting the value of the second feature amount used when the determination model 46A is generated as the normal sample value. Further, the air conditioner 1 classifies the detected value of the second feature amount as abnormal when the absolute value of the calculated outlier is equal to or greater than the absolute value of the deviation threshold X. Further, the air conditioner 1 does not use the detection value of the first feature quantity acquired simultaneously with the second feature quantity classified as abnormal for the estimation model 45A. As a result, the erroneous refrigerant shortage rate can be prevented from being estimated.

For example, when the refrigerant shortage rate is estimated by the estimation model 45A generated by the linear analysis of the multiple regression analysis, a failure other than the refrigerant leakage occurs simultaneously in addition to the refrigerant leakage, and thus the first characteristic amount changes, and in this case, depending on the degree of change of each characteristic amount, there is a possibility that the case where the refrigerant shortage rate is supposed to be increased (=abnormal) may be estimated as a small value of the refrigerant shortage rate. For example, there may be a case where the rotation speed and the suction temperature of the compressor change due to a failure other than the leakage of the refrigerant, and the refrigerant shortage rate is estimated to be a small value (=normal) as a result of the offset of the amounts of the changes of the respective values. However, in the air conditioner 1 of the present embodiment, the detection value of the first feature quantity acquired simultaneously with the detection value of the second feature quantity classified as abnormal by the judgment model 46A is not used for the estimation model 45A, wherein the judgment model 46A is generated by nonlinear analysis by a kernel density estimation method or the like. As a result, the erroneous refrigerant shortage rate can be prevented from being estimated.

Further, in the case of using the estimation model 45A generated by the linear analysis, a value (=normal) that is originally small in the refrigerant shortage ratio may be estimated as a large refrigerant shortage ratio (=abnormal). For example, the rotation speed of the compressor changes due to a failure other than refrigerant leakage, and as a result, it is estimated that the refrigerant shortage ratio increases. However, in the air conditioner 1 of the present embodiment 1, the detection value of the first feature quantity acquired simultaneously with the detection value of the second feature quantity classified as abnormal by the judgment model 46A is not used for the estimation model 45A, wherein the judgment model 46A is generated by nonlinear analysis. As a result, the erroneous refrigerant shortage rate can be prevented from being estimated.

In the judgment model 46A of the air conditioner 1, when the absolute value of the calculated outlier is smaller than the absolute value of the deviation threshold, the detected value of the second feature amount is classified as normal. In the air conditioner 1, the refrigerant shortage ratio of the refrigerant circuit 6 is calculated by performing multiple regression analysis on the detected value of the first characteristic quantity acquired simultaneously with the detected value of the second characteristic quantity classified as normal. As a result, the refrigerant shortage ratio of the refrigerant circuit 6 can be accurately estimated.

The determination model 46A mounted in the air conditioner 1 is generated by nonlinear analysis such as a nuclear density estimation method using a part of the detected value of the first characteristic amount and a value including the second characteristic amount that affects the operation state amount greatly in the refrigeration cycle operation, which are used in the estimation model 45A. In the judgment model 46A, the detected value of the second feature quantity is classified as normal or abnormal. Further, the estimation model 45A is generated by using not all the operation state amounts but the detection value of the first feature amount acquired simultaneously with the detection value of the second feature amount classified as normal, for the estimation model 45A. As a result, the highly accurate estimation model 45A can be generated.

In the present embodiment, each regression equation of the estimation model 45A is generated using the feature quantity obtained by the simulation, and the feature quantity obtained by the simulation contains no abnormal value or a large or small value that stands out as compared with other values. The detection values of the operation state amounts from which the abnormal values and the escape values have been removed by the data filtering process and the data cleaning process are substituted into the regression equations or the refrigerant shortage ratio calculation formulas of the estimation model 45A generated using the feature amounts obtained by the simulation. At this time, by substituting only the detected value of the first feature quantity acquired at the same time as the detected value of the second feature quantity classified as normal by the judgment model 46A, the refrigerant shortage ratio can be estimated more accurately.

The determination model 46A is generated using the feature quantity obtained by the simulation, and the feature quantity obtained by the simulation does not contain an abnormal value or a large or small value that stands out from other values. By applying the detection value of the second feature quantity from which the abnormal value and the escape value have been removed by the data filtering process and the data cleaning process to the judgment model 46A generated using the feature quantity that does not include the abnormal value or the escape value, the detection value of the second feature quantity can be accurately judged. Further, since the control circuit 19 can reduce the amount of data used when calculating the outlier by the judgment model 46A by performing the data filtering process and the data cleaning process, the time taken to calculate the outlier by the judgment model 46A can be reduced, and the load on the control circuit 19 can be reduced.

In the present embodiment, the operation state amounts described in fig. 5 and 6 are used as the second feature amounts, but if a large number of sensors are mounted in the refrigerant circuit to detect a large number of operation state amounts and a determination model 46A is generated, the possibility of detecting various faults can be increased by using the determination model 46A. Further, by limiting the first characteristic amount to a state amount having a correlation with the decrease in the amount of refrigerant, the amount of remaining refrigerant can be estimated with high accuracy.

In example 1 described above, the simulation results of the respective operation state amounts are obtained in the design stage of the air conditioner 1, and the simulation results are learned by an information processing device such as a server having a learning function, so that the estimation model 45A and the judgment model 46A are obtained and held in the control circuit 19. Instead of the above example, there may be a server 120 connected to the air conditioner 1 via the communication network 110, and the server 120 may generate the estimation model 45A and the determination model 46A and transmit the estimation result of the estimation model 45A to the air conditioner 1. For this embodiment, description will be made hereinafter.

Example 2

Structure of air conditioning system

Fig. 14 is an explanatory diagram showing an example of the air conditioning system 100 of embodiment 2. Note that the same components as those of the air conditioner 1 of embodiment 1 are denoted by the same reference numerals, and description of the repetitive components and operations is omitted. The air conditioning system 100 shown in fig. 14 includes: an air conditioner 1, a communication network 110, and a server 120. The air conditioner 1 includes an outdoor unit 2, an indoor unit 3, and a control circuit 19A, the outdoor unit 2 includes a compressor 11, an outdoor heat exchanger 13, and an outdoor expansion valve 14, and the indoor unit 3 includes an indoor heat exchanger 51. The air conditioner 1 includes a refrigerant circuit 6 configured by connecting the outdoor unit 2 and the indoor unit 3 through refrigerant pipes such as a liquid pipe 4 and an air pipe 5, and the refrigerant circuit 6 is filled with a predetermined amount of refrigerant. The control circuit 19A includes: an acquisition unit 41, a communication unit 42 as a first communication unit, a storage unit 43, and a control unit 44. The control circuit 19A does not include the estimation unit 45, the determination unit 46, and the abnormality record storage unit 43A.

The server 120 has: the generation unit 121, a communication unit 121A as a second communication unit, an estimation unit 122, a determination unit 123, and a storage unit 124. The storage unit 124 includes an abnormal record storage unit 124A. The generating unit 121 generates the estimation model 45A by a multiple regression analysis method using a detected value or a simulated value of the first characteristic amount related to the estimation of the refrigerant shortage ratio of the refrigerant filled in the refrigerant circuit 6. Further, the estimation model 45A includes, for example, the one described in the first embodiment: the first cooling estimation model 45A1, the second cooling estimation model 45A2, the third cooling estimation model 45A3, the first heating estimation model 45A4, the second heating estimation model 45A5, and the third heating estimation model 45A6. The estimating unit 122 stores the estimation model 45A generated by the generating unit 121. Further, the generating unit 121 generates the judgment model 46A by the kernel density estimation method using the second feature quantity. The determination model 46A includes, for example, the cooling determination model 46B and the heating determination model 46C described in embodiment 1.

The judgment unit 123 stores the judgment model 46A generated by the generation unit 121. The judgment section 123 classifies the detected value of the second feature quantity as normal or abnormal using the judgment model 46A. When the detected value of the second feature quantity is classified as abnormal, the judgment unit 123 stores the detected value of the second feature quantity classified as abnormal in the abnormal record storage unit 124A as an abnormal record.

Further, the estimating unit 122 calculates the refrigerant shortage ratio in the refrigerant circuit 6 of the air conditioner 1 using the detected value of the first characteristic amount acquired simultaneously with the detected value of the normal second characteristic amount classified by the judgment model 46A, and the received estimation model 45A. The communication unit 121A transmits the refrigerant shortage ratio calculated by the estimation unit 122 to the air conditioner 1 via the communication network 110.

The generating unit 121 generates or updates the cooling time determination model 46B using the simulated values of the second characteristic amounts of the steady state and the refrigerant leak state during cooling in the normal state of the refrigerant circuit 6.

The generating unit 121 periodically collects the operation state amounts at the time of cooling operation from a standard machine (provided in a laboratory or the like of a manufacturer) of the air conditioner 1 that can actually measure the steady state and the refrigerant leakage state at the time of cooling in the normal state of the refrigerant circuit 6, and generates or updates the cooling time determination model 46B using the comparison result between the normal or abnormal classification result and the actually measured classification result of the cooling time determination model 46B and the collected operation state amounts. As a result, the cooling time determination model 46B can be generated with higher accuracy.

The generating unit 121 periodically collects operation state amounts at the time of cooling operation from a standard machine (provided in a laboratory or the like of a manufacturer) of the air conditioner 1 capable of actually measuring the refrigerant shortage ratio in the refrigerant circuit 6, and generates or updates the first, second, and third cooling estimation models 45A1, 45A2, 45A3 using the result of comparison between the refrigerant shortage ratio estimated by each estimation model 45A and the actually measured refrigerant shortage ratio, and the collected operation state amounts. Further, as in example 1, the operation state quantity for generating each estimation model may be obtained by simulation, and the generation unit 121 may generate each estimation model 45A using the operation state quantity obtained by simulation.

The generating unit 121 generates or updates the heating time determination model 46C using the simulated values of the second characteristic amounts of the steady state and the refrigerant leak state at the time of heating in the normal state of the refrigerant circuit 6.

The generating unit 121 periodically collects the operation state quantity at the time of heating operation from a standard machine (provided in a laboratory or the like of a manufacturer) of the air conditioner 1 that can actually measure the steady state and the refrigerant leakage state at the time of heating in the normal state of the refrigerant circuit 6, and generates or updates the heating time determination model 46C using the comparison result between the normal or abnormal classification result and the actually measured classification result of the heating time determination model 46C and the collected operation state quantity. As a result, the heating time determination model 46C can be generated with higher accuracy.

The generating unit 121 periodically collects the operation state amounts at the time of heating operation from the standard machine of the air conditioner 1, and generates or updates the first, second, and third heating estimation models 45A4, 45A5, and 45A6 using the comparison result between the refrigerant shortage ratio estimated by each estimation model 45A and the actually measured refrigerant shortage ratio, and the collected operation state amounts. Further, as in example 1, the operation state quantity for generating each estimation model 45A may be obtained by simulation, and the generation unit 121 may generate each estimation model 45A using the operation state quantity obtained by simulation.

The determination model 46A generated by the generation unit 121 uses the feature quantity obtained by the simulation, and the value of the feature quantity obtained by the simulation does not include an abnormal value or a large or small value that stands out from other values. By applying the detection value of the second feature quantity from which the abnormal value and the escape value have been removed by the data filtering process and the data cleaning process to the judgment model 46A generated using the value of the feature quantity that does not include the abnormal value and the escape value, more accurate judgment of the detection value of the second feature quantity can be realized. Further, by performing the data filtering process and the data cleaning process of the second feature amount described in the first embodiment, the generation unit 121 can reduce the amount of data used when calculating the outlier from the determination model 46A. Thus, the time taken to calculate the outlier from the determination model 46A can be shortened and the utilization of the server 120 can be reduced, so that the cost of calculating the outlier can be suppressed in the case of a fee-based system such as a fee-based system that charges a fee based on the usage of the server 120.

Effect of example 2

The server 120 of embodiment 2 generates the judgment model 46A using the values of the second characteristic amounts of the steady state and the refrigerant leakage state in the normal state of the refrigerant circuit 6 obtained by the simulation, and stores the generated judgment model 46A in the judgment section 123. The judgment unit 123 in the server 120 can classify whether the detected value of the second feature amount acquired at different timings is normal or abnormal, using the stored judgment model 46A.

The server 120 generates the estimation model 45A using the value of the first feature amount acquired from the air conditioner 1, and stores the generated estimation model 45A in the estimation unit 122. The server 120 estimates the refrigerant shortage rate using the stored estimation model 45A, and transmits the estimation result to the air conditioner 1 via the communication network 110. As a result, the air conditioner 1 can recognize the refrigerant shortage rate of the refrigerant circuit 6.

In the air conditioner 1 of embodiments 1 and 2, an example of the estimation model 45A and the determination model 46A for estimating the refrigerant shortage ratio in the case where N indoor units 3 are connected to one outdoor unit 2 is shown. On the other hand, the air conditioner 1 in which one outdoor unit 2 and one indoor unit 3 are connected can estimate the refrigerant shortage ratio by the same method as in embodiment 1 or embodiment 2. The air conditioner 1 described above will be described below as example 3.

Example 3

In the outdoor unit: the indoor unit is 1:1, the control circuit has: a fourth estimation model for cooling for estimating a current refrigerant shortage ratio at the time of cooling operation, and a fifth estimation model for heating for estimating a current refrigerant shortage ratio at the time of heating operation. For convenience of explanation, the same components as those of the air conditioner 1 of embodiment 1 are denoted by the same reference numerals, and redundant explanation of the components and operations is omitted. The air conditioner 1 of embodiment 1 differs from the air conditioner 1 of embodiment 3 in that: the indoor unit 3 is one unit, and uses a fourth cooling estimation model and a fourth heating estimation model, wherein the fourth cooling estimation model is generated using different operation state amounts from the first to third cooling estimation models 45A1, 45A2, and 45A3, and the fourth heating estimation model is generated using different operation state amounts from the first to third heating estimation models 45A4, 45A5, and 45 A6.

The fourth refrigeration estimation model is a seventh regression equation generated by a multiple regression analysis method. The seventh regression equation is, for example, (α71×outdoor heat exchange temperature) - (α72×outside air temperature) - (α73×discharge temperature) + (α74×rotation speed of the compressor 11) - (α75×opening degree of the expansion valve) +α76. The coefficients α71 to α75 are determined when the estimation model is generated. The estimating unit 45 calculates the current refrigerant shortage ratio by substituting the detected value of the first characteristic amount, which is obtained at the same time as the detected value of the normal second characteristic amount classified by the judgment model 46A, of the current operation state amount after the data cleaning, for example, the outdoor heat exchange temperature, the outside air temperature, the discharge temperature, the rotation speed of the compressor 11, and the opening degree of the expansion valve, into a seventh regression equation. The reason why the outdoor heat exchange temperature, the outside air temperature, the discharge temperature, the rotation speed of the compressor 11, and the opening degree of the expansion valve are substituted is to use the feature amount used in the generation of the fourth estimated model for cooling. Further, the outdoor heat exchange temperature is detected by the refrigerant temperature sensor 35.

The fourth heating estimation model is an eighth regression equation generated by a multiple regression analysis method. The eighth regression equation is, for example, (α81×indoor heat exchange temperature) + (α82×rotation speed of the compressor 11) + (α83×outside air temperature) - (α84×outdoor heat exchange temperature) - (α85×opening degree of the expansion valve) +α86. The coefficients α81 to α85 are determined when the estimation model is generated. The estimating unit 45 calculates the current refrigerant shortage ratio by substituting the detected value of the first characteristic amount, which is obtained at the same time as the detected value of the normal second characteristic amount classified by the judgment model 46A, of the current operation state amount after the data washing, for example, the indoor heat exchange temperature, the rotation speed of the compressor 11, the outside air temperature, the outdoor heat exchange temperature, the outside air temperature, the discharge temperature, and the opening degree of the expansion valve, into the eighth regression equation. The reason why the indoor heat exchange temperature, the rotation speed of the compressor 11, the outside air temperature, the outdoor heat exchange temperature, the outside air temperature, the discharge temperature, and the opening degree of the expansion valve are substituted is to use the feature amount used in the generation of the fourth estimation model for heating. The indoor heat exchange temperature during heating can be converted from the pressure value detected by the discharge pressure sensor 31.

In the present embodiment, a case where the relative amount of refrigerant is estimated to represent the amount of refrigerant remaining in the refrigerant circuit 6 is described. Specifically, a case is described in which the refrigerant shortage ratio, which is the ratio of the amount of refrigerant leaking from the refrigerant circuit 6 to the outside to the amount of refrigerant charged (initial value) when the refrigerant circuit 6 is charged, is estimated and provided. However, the present invention is not limited to this, and the estimated refrigerant shortage ratio may be multiplied by an initial value to provide the amount of refrigerant leaking from the refrigerant circuit 6 to the outside. Further, it is also possible to generate an estimation model for estimating the absolute amount of refrigerant leaking to the outside from the refrigerant circuit 6 or the absolute amount of refrigerant remaining in the refrigerant circuit 6, and to provide an estimation result according to the estimation model. In the case of generating an estimation model for estimating the absolute amount of refrigerant leaking from the refrigerant circuit 6 to the outside or the absolute amount of refrigerant remaining in the refrigerant circuit 6, the volumes of the outdoor heat exchanger 13 and the indoor heat exchangers 51 and the volume of the liquid pipe 4 may be considered in addition to the respective operation state amounts described so far.

Modification examples

In the present embodiment, for example, the case where the interpolation is performed by the S-type coefficient between the estimation result of the first cooling estimation model 45A1 and the estimation result of the second cooling estimation model 45A2 is shown, but the interpolation method is not limited to the S-type coefficient, and for example, an interpolation method such as linear interpolation may be used, and suitable modification may be performed.

In the present embodiment, not all of the plurality of simulation results but a partial simulation result is used. For example, the models such as the first estimated model 45A1 for cooling used when the refrigerant shortage ratio is 0 to 30% at the time of cooling operation, the second estimated model 45A2 for cooling used when the refrigerant shortage ratio is 40 to 70%, and the third estimated model 45A3 for cooling used when the refrigerant shortage ratio is 30 to 40% are separately generated. Therefore, since the operation state quantity is prepared by simulation, the operation state quantity of a required quantity can be easily collected as compared with the case where the operation state quantity is collected by operating the air conditioner 1.

In the present embodiment, the case where the estimation model 45A and the judgment model 46A are generated by the server 120 or the control circuit 19 is shown, but the estimation model 45A and the judgment model 46A may be calculated by the user based on the simulation result. In the present embodiment, the case where each estimation model is generated using the multiple regression analysis method is exemplified, but the estimation model may be generated using a machine learning algorithm capable of performing a general regression analysis method, that is, SVR (support vector regression: support Vector Regression), NN (Neural Network), or the like. In this case, instead of the P value and the correction value R used in the multiple regression analysis method, a general method (forward feature selection (Forward Feature Selection) method, backward feature elimination (Backward feature Elimination), etc.) may be used to select the feature amount so as to improve the accuracy of the estimation model.

In the present embodiment, the determination model 46A is generated by using the nuclear density estimation method, but the present invention is not limited to the nuclear density estimation method, and may be appropriately modified as long as the method is a nonlinear analysis method.

In the present embodiment, the example of the air conditioner 1 in which one outdoor unit 2 is connected to one or more indoor units 3 is shown, but the present invention is applicable to an air conditioner 1 in which two or more outdoor units 2 are connected to one or more indoor units 3.

In example 1, a case is shown in which the simulation results of the respective operation state amounts are obtained in the design stage of the air conditioner 1, and the simulation results are learned by an information processing apparatus such as a server having a learning function, whereby the estimation model 45A and the judgment model 46A are obtained and held in the control circuit 19. However, a server connected to the air conditioner 1 via a communication network may be provided, and the estimation model 45A and the determination model 46A may be generated by the server and transmitted to the air conditioner 1. The air conditioner 1 may also hold the estimation model 45A and the determination model 46A received from the server in the control circuit 19.

In the refrigerant circuit 6, at least one or more indoor units 3 connected to at least one or more outdoor units 2 are connected by refrigerant piping. Therefore, the estimation model 45A can estimate the refrigerant shortage ratio using the outdoor unit 2 represented by one of the at least one or more outdoor units 2 and the detected value of the first characteristic amount of the indoor unit 3 represented by one of the at least one or more indoor units 3. The representative outdoor unit 2 may be selected from at least one or more outdoor units 2 in operation by an arbitrary rule, and the representative indoor unit 3 may be selected from at least one or more indoor units 3 in operation by an arbitrary rule. An arbitrary rule is, for example, a sequence in which the identification numbers are given to the respective machines in a short time.

In addition, each constituent element of each portion shown in the drawings is not necessarily physically constituted as shown in the drawings. That is, the specific form of the dispersion/combination of the respective portions is not limited to that shown in the drawings, and the entire or a part thereof may be functionally or physically dispersed or combined in any unit according to various loads, use conditions, and the like.

Further, the various processing functions performed by each device may be executed in all or any of the above-described processing functions on a CPU (central processing unit ) (or a microcomputer such as an MPU (micro processing unit ), an MCU (micro control unit, micro Controller Unit)). It is apparent that all or any of the various processing functions may be executed on a program analyzed and executed by a CPU (or a microcomputer such as an MPU or an MCU) or on hardware using wired logic.

In each of the embodiments described above, the refrigerant shortage is defined as the amount reduced from the predetermined amount when the refrigerant is 100% of the predetermined amount of refrigerant. Alternatively, immediately after the refrigerant circuit 6 is filled with a predetermined amount of the refrigerant, the refrigerant shortage ratio may be estimated by the method described in this embodiment, and the result of the estimation may be taken as 100%. For example, when the estimated refrigerant shortage rate immediately after the predetermined amount of refrigerant is filled into the refrigerant circuit 6 is 90%, that is, when the amount of refrigerant currently filled into the refrigerant circuit 6 is estimated to be 10% smaller than the predetermined amount, the amount of refrigerant 10% smaller than the predetermined amount may be set to 100%. By matching the amount of refrigerant determined to be 100% in this way with the estimation result, the subsequent refrigerant shortage rate can be estimated more accurately.

Symbol description

1. Air conditioner

2. Outdoor unit

3. Indoor machine

41. Acquisition unit

44. Control unit

45. Estimation unit

45A presumption model

46. Judgment part

46A judgment model

46B judgment model during refrigeration

46C heating time judging model

100. Air conditioning system

120. Server device

121. Generating part

121A communication unit

122. Estimation unit

123. Judgment part

Claims

1. An air conditioning system having an air conditioner including a refrigerant circuit configured by connecting an outdoor unit and at least one indoor unit via a refrigerant pipe, the refrigerant circuit being filled with a predetermined amount of refrigerant, and a server communicably connected to the air conditioner,

the air conditioner comprises:

a detection unit configured to detect a state quantity related to control of the air conditioner;

an acquisition unit configured to acquire a detection value detected by the detection unit; and

a first communication unit configured to transmit the detection value acquired by the acquisition unit to the server,

the server has:

a second communication unit configured to receive the detection value from the air conditioner;

an estimating unit that estimates an amount of remaining refrigerant remaining in the refrigerant circuit using a detected value of a first characteristic amount when the first characteristic amount is a state amount related to an amount of refrigerant filled in the refrigerant circuit; and

And a determination unit configured to determine whether or not the detected value of the first feature amount is a detected value to be used for estimating the amount of remaining refrigerant by the estimation unit.

2. An air conditioning system according to claim 1, wherein,

when a state quantity including at least one state quantity included in the first feature quantity and at least one state quantity not included in the first feature quantity among the state quantities is taken as a second feature quantity,

the determination unit determines whether or not the detected value of the first feature amount is a detected value to be used for estimating the remaining refrigerant amount, using the detected value of the second feature amount.

3. An air conditioning system according to claim 2, wherein,

the number of state amounts contained in the second feature amount is greater than the number of state amounts contained in the first feature amount.

4. An air conditioning system according to claim 2 or 3, characterized in that,

the estimating unit has an estimation model generated using the first feature quantity,

the estimating unit applies the detection value of the first feature quantity to the estimation model to estimate the remaining refrigerant quantity,

the judgment section has a judgment model generated by the second feature quantity,

The determination unit applies the detection value of the second feature to the determination model, and determines whether or not to use the detection value of the first feature for estimation of the remaining refrigerant amount by the estimation unit.

5. The air conditioning system of claim 4, wherein the air conditioning system comprises,

the judgment model calculates an outlier representing a degree of deviation of a detection value of the second feature quantity from the normal sample value among the detection values acquired by the acquisition unit, by using the second feature quantity used in the generation of the judgment model as the normal sample value,

the determination unit does not perform estimation of the amount of remaining refrigerant by the estimation unit using the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount when the calculated absolute value of the outlier is equal to or greater than a predetermined threshold value,

and estimating the amount of remaining refrigerant by the estimating unit using the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount when the calculated absolute value of the outlier is smaller than a predetermined threshold.

6. An air conditioning system according to claim 1, 2, 3 or 5, characterized in that,

before the estimation of the remaining refrigerant amount by the estimation unit, the determination unit determines whether or not the detection value of the first feature amount can be used.

7. The air conditioning system of claim 4, wherein the air conditioning system comprises,

8. The air conditioning system of claim 4, wherein the air conditioning system comprises,

the second feature amount used in the generation of the determination model is a value obtained based on a result of simulation of the operation of the refrigerant circuit, which is an operation when the operation of the refrigerant circuit is normal and only the remaining refrigerant amount is changed.

9. The air conditioning system of claim 4, wherein the air conditioning system comprises,

the outdoor unit is provided with a compressor and,

the detection unit includes:

a suction temperature sensor for detecting a temperature of the refrigerant sucked into the compressor, i.e., a suction temperature,

The second feature quantity includes:

the rotational speed of the compressor, and the suction temperature.

10. An air conditioning system according to claim 9, wherein,

the judgment model has:

a heating time judgment model used when the air conditioner performs heating operation; and

a cooling time judgment model used when the air conditioner performs cooling operation,

the second feature amount used in the heating time determination model and the cooling time determination model is different from each other except for the rotation speed of the compressor and the suction temperature.

11. An air conditioning system according to claim 10, wherein,

the outdoor unit is also provided with an outdoor heat exchanger and an outdoor unit expansion valve,

the detection unit further includes:

a discharge temperature sensor for detecting a temperature of the refrigerant discharged from the compressor, i.e., a discharge temperature;

a discharge pressure sensor for detecting a pressure of the refrigerant discharged from the compressor, i.e., a discharge pressure; and

a suction pressure sensor for detecting a pressure of the refrigerant sucked into the compressor, i.e., a suction pressure,

the second feature amount used in the heating time determination model includes: the discharge temperature, the suction pressure, a low pressure saturation temperature calculated using the suction pressure, and an opening degree of the outdoor unit expansion valve,

The second feature amount used in the cooling time determination model includes: a heat exchange outlet temperature of the outdoor heat exchanger, a high pressure saturation temperature calculated using the discharge pressure, and the discharge pressure.

12. The air conditioning system of claim 4, wherein the air conditioning system comprises,

the indoor unit is provided with an indoor heat exchanger and an indoor unit expansion valve,

the detection unit includes:

a refrigerant temperature sensor for detecting an indoor-side heat exchange inlet temperature, which is a temperature of refrigerant flowing into the indoor heat exchanger at a heating operation, and an indoor-side heat exchange outlet temperature, which is a temperature of refrigerant flowing out of the indoor heat exchanger at the heating operation,

the second feature quantity includes:

the indoor unit side heat exchange inlet temperature, the indoor unit side heat exchange outlet temperature and the opening degree of the indoor unit expansion valve.

13. The air conditioning system of claim 4, wherein the air conditioning system comprises,

the estimation model is generated using linear analysis,

the judgment model is generated using nonlinear analysis.

14. An air conditioning system according to claim 5, 8, 9, 10, 11 or 12, characterized in that,

The estimation model is generated using linear analysis,

the judgment model is generated using nonlinear analysis.

15. A method for estimating the amount of refrigerant in an air conditioning system having an air conditioner and a server, wherein the air conditioner has a refrigerant circuit formed by connecting an outdoor unit and at least one indoor unit through a refrigerant pipe, the refrigerant circuit is filled with a predetermined amount of refrigerant, the server is communicably connected to the air conditioner, the method for estimating the amount of refrigerant in the air conditioning system is characterized in that,

the air conditioner performs the steps of:

detecting, by a detection unit, a state quantity related to control of the air conditioner;

acquiring, by an acquisition section, a detection value of the detected state quantity; and

the acquired detection value is transmitted to the server by the first communication section,

the server performs the steps of:

receiving the detection value from the air conditioner by a second communication part;

a determination step of determining, when a state quantity related to the amount of refrigerant charged in the refrigerant circuit is a first feature quantity, whether or not a detection value of the first feature quantity is a detection value to be used for estimating the remaining amount of refrigerant, by a determination unit; and

The estimating unit estimates the amount of remaining refrigerant remaining in the refrigerant circuit using the detected value of the first characteristic amount.

16. The refrigerant quantity presumption method of an air conditioning system according to claim 15, wherein,

when a state quantity including at least one state quantity included in the first feature quantity and at least one state quantity not included in the first feature quantity is taken as a second feature quantity, the determination unit uses the detected value of the second feature quantity to determine whether or not the detected value of the first feature quantity is a detected value to be used for estimating the remaining refrigerant quantity.

17. The refrigerant quantity presumption method of an air conditioning system according to claim 16, wherein,

the judging section further performs the steps of:

inputting the detection value of the second feature quantity into a judgment model generated by the second feature quantity;

taking a second feature quantity used in the generation of the judgment model as a normal sample value to calculate an outlier representing the degree of deviation of the detection value of the second feature quantity from the normal sample value among the detection values acquired by the acquisition section; and

When the absolute value of the outlier is equal to or greater than a predetermined threshold, it is determined that the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount should not be used for estimating the remaining refrigerant amount.

18. The refrigerant quantity presumption method of an air conditioning system according to claim 17, wherein,

19. The refrigerant quantity presumption method of an air conditioning system according to claim 17, wherein,

the judging section further performs the steps of: when the absolute value of the outlier is smaller than a predetermined threshold, it is determined that the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount should be used for estimation of the remaining refrigerant amount.

20. An air conditioner having a refrigerant circuit configured by connecting an outdoor unit and at least one indoor unit through refrigerant piping, the refrigerant circuit being filled with a predetermined amount of refrigerant, the air conditioner comprising:

an acquisition unit configured to acquire a detection value detected by the detection unit;

21. The air conditioner according to claim 20, wherein,

the determination unit uses the detected value of the second feature quantity to determine whether or not the detected value of the first feature quantity is a detected value to be used for estimating the remaining refrigerant quantity.

22. The air conditioner according to claim 21, wherein,

23. An air conditioner according to claim 21 or 22, wherein,

24. The air conditioner according to claim 23, wherein,

25. An air conditioner according to claim 20, 21, 22 or 24, wherein,

26. The air conditioner according to claim 23, wherein,

27. The air conditioner according to claim 23, wherein,

28. The air conditioner according to claim 23, wherein,

the outdoor unit is provided with a compressor and,

the detection unit includes:

the second feature quantity includes:

the rotational speed of the compressor, and the suction temperature.

29. The air conditioner according to claim 28, wherein,

the judgment model has:

30. The air conditioner according to claim 29, wherein,

the detection unit includes:

31. The air conditioner according to claim 23, wherein,

the detection unit includes:

the second feature quantity includes:

32. The air conditioner according to claim 23, wherein,

the estimation model is generated using linear analysis,

the judgment model is generated using nonlinear analysis.

33. An air conditioner according to claim 24, 27, 28, 29, 30 or 31,

the estimation model is generated using linear analysis,

the judgment model is generated using nonlinear analysis.

34. A refrigerant quantity estimating method for estimating a quantity of remaining refrigerant in an air conditioner or an air conditioning system including the air conditioner, the air conditioner having a refrigerant circuit configured by connecting an outdoor unit and at least one indoor unit through a refrigerant pipe, the refrigerant circuit being filled with a predetermined quantity of refrigerant, the refrigerant quantity estimating method comprising:

detecting a state quantity related to control of the air conditioner;

acquiring a detection value of the detected state quantity;

when a state quantity related to the amount of refrigerant filled in the refrigerant circuit is set as a first feature quantity, it is determined whether or not a detected value of the first feature quantity is a detected value to be used for estimating the remaining refrigerant amount; and

The amount of remaining refrigerant remaining in the refrigerant circuit is estimated using the detected value of the first characteristic amount.

35. The refrigerant quantity estimation method of an air conditioner according to claim 34, wherein,

in the determining step, when a state quantity including at least one state quantity included in the first feature quantity and at least one state quantity not included in the first feature quantity is taken as a second feature quantity, a detection value of the second feature quantity is used to determine whether or not the detection value of the first feature quantity is a detection value to be used for estimating the remaining refrigerant quantity.

36. The refrigerant quantity estimation method of an air conditioner according to claim 35, wherein,

the step of determining further includes:

taking the second characteristic quantity as a normal sample value to calculate an outlier, wherein the outlier represents the deviation degree of the detection value of the second characteristic quantity relative to the normal sample value in the acquired detection values; and

37. The refrigerant quantity estimation method of an air conditioner according to claim 35 or 36, characterized in that,

the second characteristic amount used in the determining step is a value obtained based on a result of simulating an operation of the refrigerant circuit, the operation of the refrigerant circuit being an operation when the operation of the refrigerant circuit is normal and only the remaining refrigerant amount is changed.

38. The refrigerant quantity estimation method of an air conditioner according to claim 36, further comprising the steps of:

when the absolute value of the outlier is smaller than a predetermined threshold, it is determined that the detected value of the first feature amount acquired simultaneously with the detected value of the second feature amount should be used for estimation of the remaining refrigerant amount.