CN115095953B - Training method and device for fault diagnosis model - Google Patents

Training method and device for fault diagnosis model Download PDF

Info

Publication number
CN115095953B
CN115095953B CN202210682549.8A CN202210682549A CN115095953B CN 115095953 B CN115095953 B CN 115095953B CN 202210682549 A CN202210682549 A CN 202210682549A CN 115095953 B CN115095953 B CN 115095953B
Authority
CN
China
Prior art keywords
air conditioning
conditioning system
split air
fault
value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202210682549.8A
Other languages
Chinese (zh)
Other versions
CN115095953A (en
Inventor
张佳舒
陈焕新
苟伟
石靖峰
阮岱玮
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huazhong University of Science and Technology
Qingdao Hisense Hitachi Air Conditioning System Co Ltd
Original Assignee
Huazhong University of Science and Technology
Qingdao Hisense Hitachi Air Conditioning System Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huazhong University of Science and Technology, Qingdao Hisense Hitachi Air Conditioning System Co Ltd filed Critical Huazhong University of Science and Technology
Priority to CN202210682549.8A priority Critical patent/CN115095953B/en
Publication of CN115095953A publication Critical patent/CN115095953A/en
Priority to PCT/CN2023/071431 priority patent/WO2023197711A1/en
Application granted granted Critical
Publication of CN115095953B publication Critical patent/CN115095953B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
    • F24F11/32—Responding to malfunctions or emergencies
    • F24F11/38—Failure diagnosis
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/88—Electrical aspects, e.g. circuits
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/89—Arrangement or mounting of control or safety devices
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F2110/00—Control inputs relating to air properties
    • F24F2110/10—Temperature
    • F24F2110/12—Temperature of the outside air

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Air Conditioning Control Device (AREA)

Abstract

The embodiment of the application provides a training method and device of a fault diagnosis model, relates to the technical field of air conditioners, and is used for improving the efficiency of fault diagnosis of a multi-split air conditioning system. The method comprises the following steps: acquiring normal values and fault values of M characteristic parameters of a first multi-split air conditioning system, and determining a characteristic offset space of the first multi-split air conditioning system based on the normal values and the fault values of the M characteristic parameters; correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of the second multi-split air conditioning system, wherein the model of the first multi-split air conditioning system is different from that of the second multi-split air conditioning system; determining fault values of M characteristic parameters of the second multi-split air conditioning system based on the characteristic offset space of the second multi-split air conditioning system and normal values of the M characteristic parameters; and training the fault diagnosis model to be trained based on normal values and fault values of M characteristic parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model.

Description

Training method and device for fault diagnosis model
Technical Field
The application relates to the technical field of air conditioners, in particular to a training method and device for a fault diagnosis model.
Background
Along with the development of economy and society, multi-split air conditioning systems commonly called one-to-many are increasingly widely used in various places such as entertainment, home, work and the like.
At present, fault diagnosis for the multi-split air conditioning system comprises the step of carrying out fault diagnosis on the multi-split air conditioning system by combining operation data of the multi-split air conditioning system through a fault diagnosis model. However, at present, multiple online air conditioning systems have more machine types and wider operating condition range, and in the process of establishing a fault diagnosis model, the operating data of the multiple online air conditioning systems of different machine types in a normal operating state are easy to collect under normal conditions, and the operating data of the multiple online air conditioning systems of different machine types in the fault operating state are difficult to collect, so that the establishment period of the fault diagnosis model is longer, and further, the efficiency of fault diagnosis on the multiple online air conditioning systems is lower.
Disclosure of Invention
The embodiment of the application provides a training method and device of a fault diagnosis model, which are used for improving the efficiency of fault diagnosis of a multi-split air conditioning system.
In order to achieve the above purpose, the following technical scheme is adopted in the application.
In a first aspect, an embodiment of the present application provides a training method of a fault diagnosis model, where the method includes: acquiring normal values and fault values of M characteristic parameters of a first multi-split air conditioning system, wherein M is an integer greater than 1; determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein the characteristic offset space comprises a difference value between the fault value and the normal value of each characteristic parameter; correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of a second multi-split air conditioning system, wherein the model of the second multi-split air conditioning system is different from that of the first multi-split air conditioning system; determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and a characteristic offset space of the second multi-split air conditioning system; and training the fault diagnosis model to be trained based on normal values and fault values of M characteristic parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
The technical scheme provided by the application at least brings the following beneficial effects: according to the technical scheme, the characteristic offset space of the second multi-split air conditioning system is obtained by correcting the characteristic offset space of the first multi-split air conditioning system according to the model of the second multi-split air conditioning system. It can be understood that the values of the characteristic parameters of the multi-split air conditioning systems of different models have certain differences, but the multi-split air conditioning systems of different models should have similar characteristic offset spaces because the thermophysical mechanisms causing faults of the multi-split air conditioning systems of different models are the same. Therefore, the characteristic offset space of the first multi-split air conditioning system can be corrected according to the model of the second multi-split air conditioning system so as to obtain the characteristic offset space of the second multi-split air conditioning system. And then, the fault values of the M characteristic parameters of the second multi-split air conditioning system can be obtained according to the characteristic offset space of the second multi-split air conditioning system and the normal values of the M characteristic parameters of the second multi-split air conditioning system. And further training the fault diagnosis model to be trained by using the fault values and normal values of the M characteristic parameters of the second multi-split air conditioning system. Therefore, the fault diagnosis model can be built without collecting fault values of M characteristic parameters when the second multi-split air conditioning system breaks down in the actual working condition, namely without collecting operation data of the multi-split air conditioning systems of different machine types in the fault operation state in the actual working condition, the building period of the fault diagnosis model can be shortened, and further the efficiency of fault diagnosis of the multi-split air conditioning systems is improved.
In some embodiments, training the fault diagnosis model to be trained based on normal values and fault values of M feature parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model, including: determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system; and training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model, including: training a self-encoder (AE) model to be trained based on a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained self-encoder model; inputting the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system into the trained self-encoder model to obtain the normal value and the fault value of at least one target characteristic parameter after data reconstruction; and training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after data reconstruction to obtain a trained fault diagnosis model.
In some embodiments, the M characteristic parameters include a compressor suction temperature value, a compressor discharge temperature value, a condenser pressure value, an evaporator pressure value, an expansion valve opening value, a compressor current value, an ambient temperature value in which the outdoor unit is located, a compressor suction superheat, a compressor discharge superheat, an evaporator temperature value, and a condenser temperature value.
In some embodiments, obtaining normal values and fault values of M feature parameters of the first multi-split air conditioning system includes: acquiring historical operation data of a first multi-split air conditioning system; analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs.
In a second aspect, embodiments of the present application provide a training device, including: the acquisition unit is used for acquiring normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein M is an integer greater than 1; a processing unit for: determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein the characteristic offset space comprises a difference value between the fault value and the normal value of each characteristic parameter; correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of a second multi-split air conditioning system, wherein the model of the second multi-split air conditioning system is different from that of the first multi-split air conditioning system; determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and a characteristic offset space of the second multi-split air conditioning system; and training the fault diagnosis model to be trained based on normal values and fault values of M characteristic parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, the processing unit is specifically configured to: determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system; and training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, the processing unit is specifically configured to: training the self-encoder model to be trained based on the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a self-encoder model after training; inputting the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system into the trained self-encoder model to obtain the normal value and the fault value of at least one target characteristic parameter after data reconstruction; and training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after data reconstruction to obtain a trained fault diagnosis model.
In some embodiments, the M characteristic parameters include a compressor suction temperature value, a compressor discharge temperature value, a condenser pressure value, an evaporator pressure value, an expansion valve opening value, a compressor current value, an ambient temperature value in which the outdoor unit is located, a compressor suction superheat, a compressor discharge superheat, an evaporator temperature value, and a condenser temperature value.
In some embodiments, the obtaining unit is specifically configured to: acquiring historical operation data of a first multi-split air conditioning system; analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs.
In a third aspect, embodiments of the present application provide a training device, including: one or more processors; one or more memories; wherein the one or more memories are configured to store computer program code comprising computer instructions that, when executed by the one or more processors, cause the training apparatus to perform the training method of any of the fault diagnosis models provided in the first aspect described above.
In a fourth aspect, embodiments of the present application provide a computer-readable storage medium comprising computer instructions that, when run on a computer, cause the computer to perform the training method of any one of the fault diagnosis models provided in the first aspect.
In a fifth aspect, embodiments of the present invention provide a computer program product directly loadable into a memory and comprising software code, the computer program product being capable of implementing, upon loading and execution via a computer, a training method of any one of the fault diagnosis models as provided in the first aspect.
The beneficial effects of the second aspect to the fifth aspect of the present application may refer to the beneficial effect analysis of the first aspect, and are not described herein.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate and do not limit the invention.
Fig. 1 is a schematic structural diagram of a multi-split air conditioning system according to an embodiment of the present application;
fig. 2 is a schematic diagram of an electronic expansion valve setting position according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of another multi-split air conditioning system according to an embodiment of the present application;
fig. 4 is a schematic diagram of a refrigeration cycle principle of a multi-split air conditioning system according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of a controller according to an embodiment of the present application;
Fig. 6 is a hardware configuration block diagram of a multi-split air conditioning system provided in an embodiment of the present application;
fig. 7 is an interaction schematic diagram of a controller and a terminal device of a multi-split air conditioning system provided in an embodiment of the present application;
FIG. 8 is a flowchart of a training method of a fault diagnosis model according to an embodiment of the present application;
FIG. 9 is a flowchart of another training method of a fault diagnosis model according to an embodiment of the present application;
fig. 10 is a schematic structural diagram of a fault diagnosis model according to an embodiment of the present application;
FIG. 11 is a flowchart of another training method of a fault diagnosis model according to an embodiment of the present application;
FIG. 12 is a flowchart of another training method of a fault diagnosis model according to an embodiment of the present application;
FIG. 13 is a schematic diagram of a self-encoder model according to an embodiment of the present disclosure;
FIG. 14 is a flowchart of an application process of a fault diagnosis model according to an embodiment of the present application;
FIG. 15 is a schematic diagram of a training device according to an embodiment of the present disclosure;
fig. 16 is a schematic hardware structure of a training device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
In the description of the present invention, it should be understood that the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate orientations or positional relationships based on the orientation or positional relationships shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention.
The terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present application, unless otherwise indicated, the meaning of "a plurality" is two or more.
In the description of the present application, it should be noted that, unless explicitly stated and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected. The specific meaning of the terms in this application will be understood by those of ordinary skill in the art in a specific context. In addition, when describing a pipeline, the terms "connected" and "connected" as used herein have the meaning of conducting. The specific meaning is to be understood in conjunction with the context.
For ease of understanding, the basic concepts of some terms or techniques involved in embodiments of the present invention are first briefly described and illustrated.
Cooling mode: the compressor of the air conditioning system sucks the low-temperature low-pressure gaseous refrigerant evaporated by the evaporator into a compressor cavity, compresses the low-temperature low-pressure gaseous refrigerant into a high-temperature high-pressure gaseous refrigerant, and enters the condenser. The high-temperature high-pressure gas refrigerant is condensed into a high-temperature high-pressure liquid refrigerant in the condenser, then the high-temperature high-pressure liquid refrigerant is throttled by a throttling element such as a capillary tube, and then the high-temperature high-pressure gas refrigerant becomes a low-temperature low-pressure liquid refrigerant, and finally the low-pressure liquid refrigerant returns to the compressor after entering the evaporator to evaporate, so that the whole refrigeration cycle is completed. The outdoor heat exchanger in the refrigerating mode is used as a condenser, and the indoor heat exchanger is used as an evaporator.
Refrigerant: a substance which is easily evaporated into gas by heat absorption and easily condensed into liquid by heat release. In an air conditioning system, heat energy is transferred by evaporation and condensation of a refrigerant, thereby generating a freezing effect.
Degree of superheat: refers to the difference between the actual temperature of the refrigerant at the outlet of the evaporator and the saturation temperature corresponding to the refrigerant pressure, namely the difference between the outlet temperature of the evaporator and the evaporation temperature.
Supercooling degree: refers to the difference between the saturation temperature corresponding to the refrigerant pressure at a certain point of the condenser outlet and the actual temperature of the refrigerant.
Expansion valve: the valve consists of a valve body and a coil, and is used for throttling, reducing pressure and regulating flow. The expansion valve in the air conditioning system can throttle the medium-temperature high-pressure liquid refrigerant into low-temperature low-pressure wet steam, then the refrigerant absorbs heat in the evaporator to achieve the refrigerating effect, and the valve flow is controlled through the superheat degree change of the outlet of the evaporator.
In the prior art, aiming at fault diagnosis of the multi-split air conditioning system, a fault diagnosis model is established by combining operation data of the multi-split air conditioning system under a wide working condition, and then the multi-split air conditioning system is subjected to fault diagnosis through the fault diagnosis model. However, in the process of establishing the fault diagnosis model, the operation data of the multi-split air conditioning systems of various types in the normal operation state and the operation data of the multi-split air conditioning systems in the fault operation state are required to be collected as sample sets to train the fault diagnosis model to be trained, however, the operation data of the multi-split air conditioning systems of various types in the fault operation state are not easy to collect, so that the establishment period of the fault diagnosis model is overlong and the cost is high, and further the problem of low fault diagnosis efficiency of the multi-split air conditioning system is caused.
Based on the above, the embodiment of the application provides a training method of a fault diagnosis model, which constructs a feature offset space of a first multi-split air conditioning system according to a normal value of a feature parameter of the first multi-split air conditioning system in a normal operation state and a fault value of the feature parameter when a target fault occurs, corrects the feature offset space of the first multi-split air conditioning system according to a model of a second multi-split air conditioning system based on the same characteristics of thermophysical mechanisms of faults generated by the multi-split air conditioning systems of different models, and further combines the normal value of the feature parameter of the second multi-split air conditioning system in the normal operation state and the feature offset space of the second multi-split air conditioning system to obtain the fault value of the feature parameter when the target fault occurs in the second multi-split air conditioning system. Therefore, the fault diagnosis model can be built without collecting the operation data of the second multi-split air conditioning system in the fault operation state under the actual working condition, the building period of the fault diagnosis model is shortened, and the fault diagnosis efficiency of the multi-split air conditioning system is further improved.
Fig. 1 is a schematic structural diagram of a multi-split air conditioning system according to an exemplary embodiment of the present application. It should be noted that, the multi-split air conditioning systems of different models according to the embodiments of the present application are illustrated by taking the schematic structural diagram of the multi-split air conditioning system shown in fig. 1 as an example.
As shown in fig. 1, the multi-split air conditioning system 10 includes an outdoor unit 11, a throttle device 12, a plurality of indoor units 13, and a controller 14 (not shown in fig. 1).
The throttle device 12 includes a plurality of electronic expansion valves 121, and each electronic expansion valve 121 corresponds to one indoor unit 13. There is a pipe connection between the outdoor unit 11 and the plurality of indoor units 13, and an electronic expansion valve 121 is provided on a pipe between each indoor unit 13 and the outdoor unit 11. The conduit, also known as a gas-liquid tube, comprises: a gas pipe 15 for transporting a gaseous refrigerant, and a liquid pipe 16 for transporting a two-phase refrigerant.
For example, as shown in fig. 2, for a schematic view of an electronic expansion valve setting position provided in the present application according to an exemplary embodiment, an electronic expansion valve 121 may be disposed on a liquid pipe 16, a throttle valve may be further disposed on the liquid pipe 16, one end of the liquid pipe 16 may be connected to an indoor heat exchanger 131 described below, and similarly, one end of an air pipe 15 may also be connected to the indoor heat exchanger 131 described below.
In some embodiments, the controller 14 is coupled to each electronic expansion valve in the throttle device 12, and the opening value of each electronic expansion valve may be obtained.
Further, the outdoor unit 11, the throttle device 12, and the plurality of indoor units 13 are all communicatively connected to a controller (not shown in fig. 1), and perform related operations according to instructions of the controller.
The outdoor unit 11 is typically disposed outdoors to assist in heat exchange in an indoor environment.
The throttle device 12 is used for adjusting the flow rate of fluid in the air-conditioning gas-liquid pipe and adjusting the flow rate of refrigerant. The electronic expansion valves 121 are used for adjusting the supply amount of the refrigerant in the pipeline, and the electronic expansion valves 121 may be independent of the outdoor unit 11 (as shown in fig. 1), or may belong to a part of the outdoor unit 11 (as shown in fig. 3), and fig. 3 is a schematic structural diagram of another multi-split air conditioning system according to an exemplary embodiment of the present application. The plurality of indoor units 13 may be indoor hanging units or indoor cabinet units, which is not limited in this embodiment. The number of electronic expansion valves and the number of indoor units shown in fig. 1 or 3 are merely examples, and do not limit the embodiments of the present application.
Taking the example that the plurality of electronic expansion valves are independent of the plurality of indoor units 13, fig. 4 shows a schematic diagram of a refrigeration cycle principle of a multi-split air conditioning system.
As shown in fig. 4, the multi-split air conditioning system includes an outdoor unit 11, a throttle device 12, a plurality of indoor units 13, and a controller 14 (not shown in fig. 4).
The outdoor unit 11 includes: a compressor 111, an outdoor heat exchanger 112, a reservoir 113, and a four-way valve 114. In some embodiments, the outdoor unit 11 further includes one or more of the following: an outdoor fan, and an outdoor fan motor.
The throttling device 12 is used for adjusting the fluid flow rate in the air pipe 15 and the liquid pipe 16 in the multi-split air conditioning system.
The indoor unit 13 includes: an indoor heat exchanger 131, a display 132, and an indoor fan 133. In some embodiments, the indoor unit 13 further includes an indoor fan motor.
In some embodiments, the compressor 111 is configured to compress refrigerant delivered by the accumulator 113 and deliver the compressed refrigerant to the throttling device 12 via the four-way valve 114. The compressor 111 may be an inverter compressor of variable capacity that performs rotational speed control based on an inverter.
In some embodiments, controller 14 may obtain an operating current value (which may also be referred to as a compressor current value) for compressor 111 at each time instance.
In some embodiments, the outdoor heat exchanger 112 is connected at one end to the reservoir 113 via a four-way valve 114 and at the other end to the restriction device 12. The outdoor heat exchanger 112 has a first inlet and outlet for allowing the refrigerant to flow between the outdoor heat exchanger 112 and the suction port of the compressor 111 via the accumulator 113, and has a second inlet and outlet for allowing the refrigerant to flow between the outdoor heat exchanger 112 and the throttle device 12. The outdoor heat exchanger 112 exchanges heat between the outdoor air and the hot and cold air flowing through the heat transfer pipe connected between the first inlet and the second inlet, and the outdoor heat exchanger 112 operates as a condenser in the cold cycle. For ease of description, the outdoor heat exchanger 112 is exemplified as a condenser.
In some embodiments, a reservoir 113 is connected to the compressor 111 at one end and to the outdoor heat exchanger 112 at the other end via a four-way valve 114. In the accumulator 113, the refrigerant flowing from the outdoor heat exchanger 112 to the compressor 111 via the four-way valve 114 is separated into a gas refrigerant and a liquid refrigerant. The gas refrigerant is mainly supplied from the accumulator 113 to the suction port of the compressor 111.
In some embodiments, four ports of four-way valve 114 are connected to compressor 111, outdoor heat exchanger 112, reservoir 113, and plurality of electronic expansion valves 121, respectively. The four-way valve 114 is used to switch between cooling and heating by changing the flow direction of the refrigerant in the system piping.
In some embodiments, the outdoor fan facilitates heat exchange with the outdoor air by generating an airflow of the outdoor air through the outdoor heat exchanger 112 to promote heat exchange with the outdoor air by the refrigerant flowing in the heat transfer tube between the first and second inlets and outlets.
The flow direction of the refrigerant in the principle of the refrigeration cycle is shown by the arrow flow direction shown in fig. 4:
the compressor 111 discharges high temperature and high pressure gas, the four-way valve 114, the plurality of indoor units 13, the outdoor heat exchanger 112, the four-way valve 114, the liquid storage 113 and the suction inlet of the compressor 111, and the circulation process of the refrigerant is completed.
In some embodiments, the outdoor fan motor is used to drive or alter the rotational speed of the outdoor fan.
In some embodiments, the electronic expansion valve 121 has a function of expanding and decompressing the refrigerant flowing through the electronic expansion valve 121, and may be used to adjust the supply amount of the refrigerant in the pipe. When the electronic expansion valve 121 decreases in opening degree, the flow path resistance of the refrigerant passing through the electronic expansion valve 121 increases. When the electronic expansion valve 121 increases in opening degree, the flow path resistance of the refrigerant passing through the electronic expansion valve 121 decreases. In this way, even if the state of other devices in the circuit does not change, when the opening degree of the electronic expansion valve 121 changes, the flow rate of the refrigerant flowing to the indoor unit 13 changes.
In some embodiments, the indoor heat exchanger 131 has a third inlet and outlet for flowing liquid refrigerant between it and the electronic expansion valve 121, and has a fourth inlet and outlet for flowing gaseous refrigerant between it and the discharge of the compressor 111. The indoor heat exchanger 131 exchanges heat between the indoor air and the refrigerant flowing through the heat transfer pipe connected between the third inlet and the fourth inlet. In the cold cycle, the indoor heat exchanger 131 operates as an evaporator. For convenience of description, the indoor heat exchanger 131 is taken as an evaporator for example for distance illustration.
In some embodiments, the indoor fan 133 generates an airflow of the indoor air passing through the indoor heat exchanger 131 to promote heat exchange of the refrigerant flowing in the heat transfer pipe between the third and fourth inlets and outlets with the indoor air.
In some embodiments, an indoor fan motor is used to drive or alter the rotational speed of the indoor fan 133.
In some embodiments, the display 132 is used to display the indoor temperature or the current mode of operation.
In the embodiment shown in the present application, the controller 14 refers to a device that can generate an operation control signal according to the instruction operation code and the timing signal, and instruct the multi-split air conditioning system to execute a control instruction. By way of example, the controller may be a central processing unit (central processing unit, CPU), a general purpose processor network processor (network processor, NP), a digital signal processor (digital signal processing, DSP), a microprocessor, a microcontroller, a programmable logic device (programmable logic device, PLD), or any combination thereof. The controller may also be any other device having a processing function, such as a circuit, a device, or a software module, which is not limited in any way by the embodiments of the present application.
In addition, the controller 14 may be used to control the operation of various components within the multi-split air conditioning system 10 such that the various components of the multi-split air conditioning system 10 operate to perform various predetermined functions of the multi-split air conditioning system.
In some embodiments, the multi-split air conditioning system 10 is also attached with a remote control having functionality to communicate with the controller 14, for example, using infrared or other communication means. The remote controller is used for various controls of the multi-split air conditioning system by a user, and interaction between the user and the multi-split air conditioning system 10 is realized.
Referring to fig. 5, a schematic structural diagram of a controller according to an exemplary embodiment is provided in the present application. As shown in fig. 5, the controller 14 includes an outdoor control module 141 and an indoor control module 142. The outdoor control module 141 includes a first memory 1411, and the indoor control module 142 includes a second memory 1421. The indoor control module 142 is connected with the outdoor control module 141 through a wired or wireless communication form. The outdoor control module 141 may be installed in the outdoor unit 11 or may be independent of the outdoor unit 11 to control the outdoor unit 11 to perform related operations. The indoor control module 142 may be installed in the indoor unit 13, or may be independent of the indoor unit 13, and may be used to control components of the indoor unit 13 and the throttle device 12 to perform related operations. It should be understood that the above division of the modules is only a functional division, and the outdoor control module 141 and the indoor control module 142 may be integrated in one module. The first memory 1411 and the second memory 1421 can also be integrated as one memory.
In some embodiments, the first memory 1411 is used to store applications and data related to the outdoor unit 11, and the outdoor control module 141 performs various functions and data processing of the multi-split air conditioning system by running the applications and data stored in the memory 1411. The first memory 1411 mainly includes a program storage area and a data storage area, wherein the program storage area can store an operating system and application programs (such as an outdoor fan on function, an outdoor temperature measurement function, etc.) required by at least one function; the storage data area may store data (such as outdoor temperature, opening degree of each electronic expansion valve, etc.) created according to the use of the multi-split air conditioning system. In addition, the first memory 1411 may include high-speed random access memory, and may also include nonvolatile memory, such as magnetic disk storage devices, flash memory devices, or other volatile solid state storage devices, and the like.
In some embodiments, the second memory 1421 is used for storing application programs and data related to the plurality of indoor units 13 and the plurality of electronic expansion valves 121, and the indoor control module 1421 performs various functions and data processing of the multi-split air conditioning system by running the application programs and data stored in the memory 1421. The second memory 1421 mainly includes a program storage area and a data storage area, where the program storage area can store an operating system and at least one application program required by a function (such as an indoor unit fan on function, an indoor temperature measurement function, etc.); the storage data area may store data (e.g., indoor temperature, etc.) created according to the use of the multi-split air conditioning system. In some examples, the second memory 1421 is also used to store a correspondence between an address of the indoor unit 13 and an address of the electronic expansion valve 121.
In some embodiments, the outdoor control module 141 is in communication with the outdoor unit 11, and is configured to control the outdoor unit to perform related operations according to a user command or a default command of the system. Alternatively, the outdoor control module 141 may control the rotation speed of the outdoor fan according to an air conditioner operation mode selected by a user. Alternatively, the outdoor control module 141 may also acquire an outdoor temperature according to a user instruction or a system instruction and store the acquired outdoor temperature to the first memory 1411. Optionally, the outdoor control module 141 may further control the four-way valve 114 in the outdoor unit 11 to rotate according to the air conditioning operation mode selected by the user, so as to realize the selection of the cooling or heating mode. Alternatively, the outdoor control module 141 may also control the operation mode of the outdoor unit 11, the compressor frequency, etc. during the address correction.
In some embodiments, a communication link exists between the indoor control module 142 and the indoor unit 13 for controlling the indoor unit 13 to perform related operations according to user instructions or system default instructions. Alternatively, the indoor control module 142 controls the indoor unit 13 to turn on the indoor fan and the fan motor according to a user instruction. Optionally, the indoor control module 142 may also control the indoor unit to turn on or off the compressor in the indoor unit according to a user instruction. Optionally, the indoor control module 142 may also control the indoor unit to turn on the indoor temperature sensor according to a user instruction, so as to detect the indoor temperature.
In some embodiments, a communication link exists between the indoor control module 142 and the plurality of electronic expansion valves 121 for controlling the plurality of electronic expansion valves 121 to perform related operations according to user instructions or system default instructions. Alternatively, the indoor control module 142 may also control the opening degree of each electronic expansion valve 121 according to a user instruction or a system instruction.
It should be understood that, in the embodiment shown in fig. 4, the throttle device 12 is independent of the plurality of indoor units 13, and if the throttle device 12 is located inside the plurality of indoor units 13, the refrigeration cycle principle of the multi-split air conditioning system is still applicable, and will not be described in detail.
Fig. 6 is a block diagram illustrating a hardware configuration of a multi-split air conditioning system according to an exemplary embodiment of the present application. As shown in fig. 6, the multi-split air conditioning system 10 may further include one or more of the following: a first temperature sensor 101, a second temperature sensor 102, a plurality of third temperature sensors 103, a fourth temperature sensor 104, a fifth temperature sensor 105, a plurality of sixth temperature sensors 106, a first pressure sensor 107, a second pressure sensor 108, and a communicator 109.
In some embodiments, the first temperature sensor 101 is connected to the controller 14, and the first temperature sensor 101 may be disposed at the suction port of the compressor 111 for detecting a compressor suction temperature value and transmitting the detected compressor suction temperature value to the controller 14.
In some embodiments, a second temperature sensor 102 is coupled to the controller 14, and the second temperature sensor 102 may be disposed at a discharge port of the compressor 111 for detecting a compressor discharge temperature value and transmitting the detected compressor discharge temperature value to the controller 14.
In some embodiments, a plurality of third temperature sensors 103 are each coupled to the controller 14. A third temperature sensor 103 may be disposed on one of the outdoor units 11, and configured to detect an ambient temperature value of the outdoor unit 11, and send the detected ambient temperature value of the outdoor unit to the controller 14.
In some embodiments, a fourth temperature sensor 104 is coupled to the controller 14, and the fourth temperature sensor 104 may be disposed at the condenser (i.e., the outdoor heat exchanger 112) for detecting a condenser temperature value and transmitting the detected condenser temperature value to the controller 14.
In some embodiments, the fifth temperature sensor 105 is connected to the controller 14, and the fifth temperature sensor 105 may be disposed at the evaporator (i.e., the indoor heat exchanger 131) for detecting an evaporator temperature value and transmitting the detected evaporator temperature value to the controller 14.
In some embodiments, the plurality of sixth temperature sensors 106 are all connected to the controller 14, and one sixth temperature sensor 106 may be disposed in the environment where one indoor unit 13 is located, for detecting an environmental temperature value of the environment where one indoor unit 13 is located, and transmitting the detected environmental temperature value to the controller 14.
In some embodiments, the first pressure sensor 107 is connected to the controller 14, and the first pressure sensor 107 may be disposed at the condenser for detecting a condenser pressure value and transmitting the detected condenser pressure value to the controller 14.
In some embodiments, a second pressure sensor 108 is coupled to the controller 14, and the second pressure sensor 108 may be disposed at the evaporator for detecting an evaporator pressure value and transmitting the detected evaporator pressure value to the controller 14.
In some embodiments, the communicator 109 is coupled to the controller 14 for establishing communication connections with other network entities, such as with terminal devices, and with servers. The communicator 109 may include a Radio Frequency (RF) module, a cellular module, a wireless fidelity (wireless fidelity, WIFI) module, a GPS module, and the like. Taking an RF module as an example, the RF module may be used for receiving and transmitting signals, in particular, transmitting the received information to the controller 14 for processing; in addition, the signal generated by the controller 14 is transmitted. Typically, the RF circuitry may include, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier (low noise amplifier, LNA), a duplexer, and the like.
For example, the multi-split air conditioning system 10 may receive a control instruction sent by the terminal device through the communicator 109, and execute corresponding processing according to the control instruction, so as to implement interaction between the user and the multi-split air conditioning system 10.
In some embodiments, the multi-split air conditioning system 10 may also send its own operation data to the server through the communicator 109, so that the server calculates the operation parameters of the multi-split air conditioning system 10 during the working process according to the operation data of the multi-split air conditioning system 10, and then sends the calculated operation parameters to the multi-split air conditioning system 10. And the controller 14 controls the components of the multi-split air conditioning system 10 to operate according to the operation parameters calculated by the server.
The server may be a single server, or may be a server cluster formed by a plurality of servers. In some embodiments, the server cluster may also be a distributed cluster, and the specific type of the server is not limited in the embodiments of the present application.
It can be understood that the operation capability of the server is higher than that of the multi-split air conditioning system. In some embodiments, in order to reduce the operation pressure of the multi-split air conditioning systems, the multi-split air conditioning systems are all in communication connection with the server, the multi-split air conditioning systems can send their own operation data to the server, and the server builds a fault diagnosis model according to the operation data of the multi-split air conditioning systems. Therefore, the operation pressure of each multi-split air conditioning system can be reduced, and the utilization rate of the computing power resources of the multi-split air conditioning system is improved.
Those skilled in the art will appreciate that the hardware configuration shown in fig. 6 is not limiting of a multi-split air conditioning system, which may include more or fewer components than shown, or may combine certain components, or may have a different arrangement of components.
Fig. 7 is an interaction schematic diagram of the controller 14 and the terminal device 300 of the multi-split air conditioning system according to an exemplary embodiment of the present application.
As shown in fig. 7, the terminal device 300 may establish a communication connection with the controller 14 of the multi-split air conditioning system 10. By way of example, the establishment of the communication connection may be accomplished using any known network communication protocol. The network communication protocol may be various wired or wireless communication protocols such as Ethernet, universal serial bus (universal serial bus, USB), FIREWIRE (FIREWIRE), any cellular network communication protocol (e.g., 3G/4G/5G), bluetooth, wireless Fidelity (wireless fidelity, wi-Fi), NFC, or any other suitable communication protocol. The communication connection may be a bluetooth connection, NFC, zigbee, wireless fidelity (wireless fidelity, wi-Fi), or the like. This is not particularly limited in the embodiments of the present application.
Note that the terminal device 300 shown in fig. 7 is only one example of a terminal device. The terminal device 300 in the present application may be a mobile phone, a tablet computer, a personal computer (personal computer, PC), a personal digital assistant (personal digital assistant, PDA), a smart watch, a netbook, a wearable electronic device, an augmented reality (augmented reality, AR) device, a Virtual Reality (VR) device, a robot, or the like, and the specific form of the terminal device is not particularly limited in the present application.
By way of example, taking the terminal device 300 as a mobile phone, a user may download an intelligent home APP on the mobile phone, where the intelligent home APP may be used to manage an intelligent home device, and in this embodiment, the intelligent home device is exemplified by the multi-split air conditioning system 10. Furthermore, the user may select the online device of the multi-split air conditioning system 10, and select a control function to be executed on the multi-split air conditioning system 10 from management options of the multi-split air conditioning system 10. For example, control functions such as start-up, shut-down, switching modes (e.g., cooling mode or heating mode), etc. If the fact that the user clicks a start button of the intelligent home APP on the multi-split air conditioning system 10 is detected, the mobile phone can send a start instruction to the multi-split air conditioning system 10, so that the multi-split air conditioning system 10 starts up to work in response to the start instruction.
The embodiments provided in the present application are specifically described below with reference to the drawings attached to the specification.
As shown in fig. 8, the embodiment of the present application provides a training method of a fault diagnosis model, which may be applied to the controller 14 in the multi-split air conditioning system 10 and may also be applied to a server, which is not limited in this application, and for convenience of description, the following embodiments take the method applied to the controller 14 in the multi-split air conditioning system 10 as an example, and the method includes the following steps:
S101, acquiring normal values and fault values of M characteristic parameters of a first multi-split air conditioning system.
In some embodiments, the first multi-split air conditioning system may be referred to as an experimental multi-split air conditioning system. The memory of the first multi-split air conditioning system stores historical operation data generated when the first multi-split air conditioning system operates under various working conditions.
In some embodiments, before the fault diagnosis model is established, normal values and fault values of M feature parameters of the first multi-split air conditioning system may be obtained, where M is an integer greater than 1.
The characteristic parameters are used for representing parameter information generated in the running process of the multi-split air conditioning system. The characteristic parameters comprise a compressor suction temperature value, a compressor discharge temperature value, a condenser pressure value, an evaporator pressure value, an expansion valve opening value, a compressor current value, an environment temperature value where an outdoor unit is positioned, a compressor suction superheat degree, a compressor discharge superheat degree, an evaporator temperature value and a condenser temperature value.
It can be appreciated that the data generated by the multi-split air conditioning system in the normal operation state is different from the data generated in the case of generating a fault, and one characteristic parameter includes a normal value of the multi-split air conditioning system in the normal operation state and a fault value in the case of generating a fault.
Specifically, as shown in fig. 9, step S101 may be specifically implemented as the following steps:
s1011, acquiring historical operation data of the first multi-split air conditioning system.
The historical operation data comprise normal operation data generated by the first multi-split air conditioning system in a normal operation state and various fault operation data generated when various faults occur in a historical period. It is to be understood that the history period may be a period of time before the current time, and the history period may be set as needed, without limitation.
S1012, analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs.
According to the method, the normal value of one characteristic parameter in the normal operation state of the multi-split air conditioning system is different from the fault value when the fault occurs, so that historical operation data can be analyzed, data generated by the multi-split air conditioning system in the normal operation state and data generated under the condition of generating a target fault are distinguished, and further the normal value and the fault value of each characteristic parameter in M characteristic parameters are obtained according to the data generated by the multi-split air conditioning system in the normal operation state and the data generated under the condition of generating the target fault. The target faults include refrigerant leakage, excessive refrigerant charge, outdoor unit dirt, indoor unit dirt, electronic expansion valve blocking and the like.
In some embodiments, a preset value range corresponding to each characteristic parameter is pre-stored in a database of the first multi-split air conditioning system. When the value of one characteristic parameter is in the corresponding preset value range, the normal value of the characteristic parameter is determined. And when the value of one characteristic parameter is not in the corresponding preset value range, determining the fault value of the characteristic parameter. The preset value range corresponding to each characteristic parameter may be preset by a manager when the first multi-split air conditioning system leaves the factory.
In some embodiments, analyzing the historical operation data, and determining the normal values of the M feature parameters of the first multi-split air conditioning system in the normal operation state of the first multi-split air conditioning system and the fault values when the target fault occurs may include: and analyzing the historical operation data according to a preset value range corresponding to each of the M characteristic parameters, so as to distinguish a normal value and a fault value of each of the M characteristic parameters of the first multi-split air conditioning system.
It can be understood that the fault values of the M feature parameters are different when the multi-split air conditioning system generates different faults, so that the fault values corresponding to the M feature parameters of the first multi-split air conditioning system based on each fault can be resolved while the historical operation data are resolved to determine the fault values and the normal values of the M feature parameters.
S102, determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system.
It can be understood that, under the condition that the first multi-split air conditioning system has a target fault, the fault values of the M feature parameters of the first multi-split air conditioning system generate a certain deviation compared with the normal value, so that the difference value between the fault value of one feature parameter and the normal value can be used as the offset of the feature parameter when the first multi-split air conditioning system has the target fault, and then, according to the difference value between the fault value of each feature parameter in the M feature parameters and the normal value, a feature offset space of the first multi-split air conditioning system when the target fault occurs is constructed, wherein the feature offset space of the first multi-split air conditioning system comprises the offset of each feature parameter in the M feature parameters.
In some embodiments, the space formed by the normal values of the M characteristic parameters of one multi-split air conditioning system may be referred to as the characteristic space of the multi-split air conditioning system.
For example, the feature space of the first multi-split air conditioning system may be represented by the following matrix:
X＝[x 1 x 2 … x n-1 x n ]
wherein x is j And j is a column vector of a certain characteristic parameter, and the value of j is 1-n.
The characteristic offset space of the first multi-split air conditioning system under a certain fault type is as follows:
Δxi＝[Δxi 1 Δxi 2 … Δxi n-1 Δxi n ]
wherein Δxi j And j is a value of 1-n for the offset of a certain characteristic parameter.
S103, correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of the second multi-split air conditioning system.
In some embodiments, the model of the first multi-split air conditioning system is different from the model of the second multi-split air conditioning system. The first multi-split air conditioning system is of a single-cooling type, and the second multi-split air conditioning system is of a heat pump type.
It should be noted that, although there is a certain difference between the feature spaces of the multi-split air conditioning systems of different models, similar feature offset spaces should be provided between the multi-split air conditioning systems of different models because the thermophysical mechanisms that cause the faults of the multi-split air conditioning systems of different models are the same. Therefore, the characteristic offset space of the first multi-split air conditioning system can be corrected according to the model of the second multi-split air conditioning system so as to obtain the characteristic offset space of the second multi-split air conditioning system.
Specifically, the memory of the first multi-split air conditioning system stores the corresponding relation between the model and the correction coefficient of the multi-split air conditioning system in advance. After the model of the second multi-split air conditioning system is determined, a target correction coefficient corresponding to the model of the second multi-split air conditioning system can be determined according to the model of the second multi-split air conditioning system and a pre-stored corresponding relation, and then the characteristic offset space of the first multi-split air conditioning system is corrected according to the target correction coefficient to obtain the characteristic offset space of the second multi-split air conditioning system.
In some embodiments, the correction coefficients may be represented in the following matrix:
λi＝[λi 1 λi 2 … λi n-1 λi n ]
and lambdaj is used for representing the correction coefficient corresponding to the model of a multi-split air conditioning system.
For example, the target correction coefficient may be multiplied by the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of the second multi-split air conditioning coefficient.
The corresponding relation between the model of the multi-split air conditioning system and the correction coefficient can be obtained through a large number of actual experiments by workers to which the first multi-split air conditioning system belongs.
In some embodiments, the correction factor ranges between values (0.8,1.2).
S104, determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and the characteristic offset space of the second multi-split air conditioning system.
In some embodiments, operation data of the second multi-split air conditioning system in a normal operation state in unit time can be obtained, and then the operation data is analyzed to determine normal values of M characteristic parameters of the second multi-split air conditioning system.
After the normal values of the M characteristic parameters of the second multi-split air conditioning system are obtained, the fault values of the M characteristic parameters of the second multi-split air conditioning system can be obtained according to the normal values of the M characteristic parameters of the second multi-split air conditioning system and the characteristic offset space of the second multi-split air conditioning system.
For example, normal values of M characteristic parameters of the second multi-split air conditioning system may be added to the characteristic offset space of the second multi-split air conditioning system to obtain fault values of M characteristic parameters of the second multi-split air conditioning system.
S105, training the fault diagnosis model to be trained based on normal values and fault values of M characteristic parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, after obtaining the normal values and the fault values of the M feature parameters of the second multi-split air conditioning system, the normal values and the fault values of the M feature parameters of the second multi-split air conditioning system may be used as a sample set, and the fault diagnosis model to be trained may be trained to obtain the fault diagnosis model after training is completed.
In some embodiments, the fault diagnosis model is a one-dimensional convolutional neural network based fault diagnosis model. Among these, convolutional neural networks (Convolutional Neural Networks) are a special type of multi-layer perceptron. The working principle of convolutional neural networks mainly involves three basic concepts: local receptive fields, pooling, and sharing weights. The one-dimensional convolutional neural network is a convolutional neural network with one-dimensional input data and has classification capability, and can be used for fault detection and diagnosis. Fault detection is a process of detecting whether a system fails, and fault diagnosis is to acquire detailed information of a fault. The detailed information may include fault location, fault type, and fault severity.
The convolution layer is a key structure in a convolution neural network and its function is to convolve its input with a series of filters to form a series of feature maps of the original input and output. When a filter convolves the convolved data, a local receptive field is repeatedly shifted over the input data. The data in the local receptive field is subjected to dot product operation with the weight matrix in the filter, and an output matrix is formed after a fixed offset value is added, wherein the matrix is a characteristic mapping of the original input. The filter has the characteristic of sharing weight for the convolution operation of the input, namely, the weight matrix and bias used by one filter for the data in the local receptive field are the same. The parameters of the convolution layer mainly comprise the size of the local receptive field, the number of filters, the steps, etc. The size of the convolution kernel determines the width of the local receptive field. The number of filters determines how many sets of feature maps of the original inputs the convolution outputs. The stride determines the moving step size of the local receptive field between two adjacent operations.
An activation function may be applied to the output of the convolutional layer for de-linearization. ReLU activation functions are standard functions in modern applications for many different complex tasks. The function of ReLU can be expressed as the following formula (1):
Output a of the ith convolution layer i Can be represented by the following formula (2):
wherein,for convolution operation, σ is the activation function, b i To bias, w i Is a weight matrix.
In some embodiments, as shown in fig. 10, the fault diagnosis model may include: input layer, convolution layer, pooling layer, full connection layer, and output layer.
The output of the convolutional layer may be input into the pooling layer. The pooling layer divides the input data into multiple pooled regions and separately generalizes and forms an output for each pooled region. Typical pooling approaches include maximum pooling, average pooling, etc. The output of the pooling layer is a generalized and compact feature map of its inputs.
In dealing with classification problems, a classification network (softmax) layer is typically employed as the output layer of the convolutional neural network. The softmax layer is a network layer consisting of a softmax function. The softmax function is a numerical function that can convert a D-dimensional raw vector of arbitrary value into a D-dimensional probability vector of values between 0, 1. And determining the classification result of the input data of the softmax layer according to the dimension of the maximum value of the probability vector in the output of the softmax layer.
The softmax function is shown in the following equation (3):
Wherein x is d Is the input of the D-th dimension in the D-dimension original input, P d ∈(0,1)，k∈{1,2,3,…,D}，x k For representing each of the D dimensions.
It can be understood that, because the fault values of the M feature parameters of the second multi-split air conditioning system are obtained according to the feature offset space of the second multi-split air conditioning system, the feature offset space of the second multi-split air conditioning system is related to the feature offset space of the first multi-split air conditioning system, and the feature offset space of the first multi-split air conditioning system is established according to the normal values and the fault values of the M feature parameters of the first multi-split air conditioning system under the condition that the first multi-split air conditioning system has a target fault, the normal values and the fault values of the M feature parameters of the second multi-split air conditioning system are used as a fault diagnosis model completed by sample training, and whether the target fault occurs in the multi-split air conditioning system identical to the model of the second multi-split air conditioning system can be identified and diagnosed.
In some embodiments, in order to enable the trained fault diagnosis model to perform fault diagnosis on the multi-split air conditioning systems of different models, after determining the feature offset space of the first multi-split air conditioning system, the feature offset space of the first multi-split air conditioning system may be corrected according to the models of the multi-split air conditioning systems of different models to obtain the feature offset space of the multi-split air conditioning systems of different models, further obtain normal values of the M feature parameters of the multi-split air conditioning systems of different models, and combine the normal values of the M feature parameters of the multi-split air conditioning systems of different models and the feature offset space of the multi-split air conditioning systems of different models to determine the fault values of the M feature parameters of the multi-split air conditioning systems of different models, and further train the normal values and the fault values of the M feature parameters of the multi-split air conditioning systems of different models as a sample set to obtain the trained fault diagnosis model. Therefore, the trained fault diagnosis model can carry out fault diagnosis on the multi-split air conditioning systems of different models, whether target faults occur in the multi-split air conditioning systems of different models is identified, one fault diagnosis model is not required to be established for each model, and the efficiency of fault diagnosis on the multi-split air conditioning systems is improved.
In some embodiments, in order to enable the trained fault diagnosis model to perform fault diagnosis on the multi-split air conditioning systems with different models and different fault types, when historical data of the first multi-split air conditioning system are analyzed, fault values of M characteristic parameters of the first multi-split air conditioning system when faults with different fault types occur can be analyzed, and because influences of different fault types on the first multi-split air conditioning system are different, a characteristic offset space of the first multi-split air conditioning system under each fault type can be established according to the fault values of M characteristic parameters corresponding to each fault type and normal values of the M characteristic parameters.
For any one of multiple air conditioning systems of different types, correcting the characteristic offset space of the first multiple air conditioning system under each fault type according to the type of the multiple air conditioning system to obtain the characteristic offset space of the multiple air conditioning system under each fault type, further obtaining normal values of M characteristic parameters of the multiple air conditioning system, and further obtaining fault values of M characteristic parameters of the multiple air conditioning system.
And training the fault diagnosis model to be trained according to the fault values and normal values of M characteristic parameters of the multi-split air conditioning systems of different models as a sample set to obtain a trained fault diagnosis model. Therefore, the trained fault diagnosis model can perform fault diagnosis on the multi-split air conditioning systems with different machine types and different fault types, can identify the fault types of the fault multi-split air conditioning systems with different machine types, and further improves the efficiency of fault diagnosis on the multi-split air conditioning systems.
Based on the embodiment shown in fig. 8, the characteristic offset space of the second multi-split air conditioning system is obtained by correcting the characteristic offset space of the first multi-split air conditioning system according to the model of the second multi-split air conditioning system. It can be understood that the characteristic parameters of the multi-split air conditioning systems of different models have certain differences, but the multi-split air conditioning systems of different models should have similar characteristic offset spaces because the thermophysical mechanisms causing the faults of the multi-split air conditioning systems of different models are the same. Therefore, the characteristic offset space of the first multi-split air conditioning system can be corrected according to the model of the second multi-split air conditioning system so as to obtain the characteristic offset space of the second multi-split air conditioning system. And then, the fault values of the M characteristic parameters of the second multi-split air conditioning system can be obtained according to the characteristic offset space of the second multi-split air conditioning system and the normal values of the M characteristic parameters of the second multi-split air conditioning system. And further training the fault diagnosis model to be trained by using the fault values and normal values of the M characteristic parameters of the second multi-split air conditioning system.
Therefore, the fault diagnosis model can be built without collecting fault values of M characteristic parameters when the second multi-split air conditioning system breaks down in an actual working condition, namely, without collecting operation data of the multi-split air conditioning systems of different machine types in a fault operation state, the building period of the fault diagnosis model can be shortened, and further the efficiency of fault diagnosis of the multi-split air conditioning systems is improved.
In some embodiments, as shown in fig. 11, the step S105 may be implemented as the following steps:
s201, determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system.
It can be understood that the number of the M feature parameters of the second multi-split air conditioning system is larger, and some redundant feature parameters irrelevant to the training of the fault diagnosis model are doped, so that in order to increase the training speed of the fault diagnosis model, the target feature parameters with higher fault diagnosis relevance to the subsequent fault diagnosis model need to be extracted from the M feature parameters, and because the trained fault diagnosis model needs to perform fault diagnosis on the multi-split air conditioning systems of different models, the at least one target feature parameter is a feature parameter which is common to the multi-split air conditioning systems of different models and has higher fault diagnosis relevance to the fault diagnosis model.
For example, determining at least one target feature parameter from the M feature parameters of the second multi-split air conditioning system based on the normal values and the fault values of the M feature parameters of the second multi-split air conditioning system may include: and calculating the difference value between the normal value and the fault value of each of the M characteristic parameters of the second multi-split air conditioning system, taking N characteristic parameters corresponding to NG difference values with the difference value larger than a preset threshold value in the M difference values as target characteristic parameters, and further extracting at least one characteristic parameter shared by the multi-split air conditioning systems of various models from the N characteristic parameters as the target characteristic parameter. The preset threshold may be preset by a manager, where N is an integer greater than 1.
It can be understood that the larger the difference between the normal value and the fault value of one characteristic parameter is, the larger the change of the characteristic parameter when the second multi-split air conditioning system is converted from the normal operation state to the target fault state is represented, and the target fault of the second multi-split air conditioning system can be reflected, so that the characteristic parameter can be used as the characteristic parameter with higher fault diagnosis relevance to the subsequent fault diagnosis model.
The target characteristic parameters may include: the compressor suction superheat degree, the compressor discharge superheat degree, the compressor suction temperature value, the compressor discharge temperature value, the evaporator temperature value, the condenser temperature value, the high-pressure value of the second multi-split air conditioning system, the low-pressure value of the second multi-split air conditioning system, the current value, the environmental temperature value of the outdoor unit, the average temperature value of the environments of the plurality of indoor units and the like.
S202, training a fault diagnosis model to be trained according to a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, after determining at least one target characteristic parameter of the second multi-split air conditioning system, training the fault diagnosis model to be trained according to a normal value and a fault value of the at least one target characteristic parameter of the second multi-split air conditioning system as a sample set to obtain a trained fault diagnosis model.
Based on the embodiment shown in fig. 11, the feature extraction is performed on the M feature parameters of the second multi-split air conditioning system, so as to extract at least one target feature parameter of the M feature parameters of the second multi-split air conditioning system, and further train the fault diagnosis model to be trained according to the normal value and the fault value of the at least one target feature parameter. The training sample set of the fault diagnosis model to be trained is simplified, so that the training speed of the fault diagnosis model to be trained can be improved, and further the efficiency of fault diagnosis of the multi-split air conditioning system is improved.
In some embodiments, as shown in fig. 12, step S202 may be embodied as the following steps:
s301, training a self-encoder model to be trained based on a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained self-encoder model.
It can be understood that, when the fault diagnosis model actually performs fault diagnosis, because the data distribution of the multi-split air conditioning system to be diagnosed is often inconsistent with the data distribution of the second multi-split air conditioning system, the fault diagnosis model has lower accuracy in fault diagnosis according to the original operation data of the multi-split air conditioning system to be diagnosed. The original operation data of the multi-split air conditioning system to be diagnosed needs to be subjected to data reconstruction, so that the distribution of the original operation data of the multi-split air conditioning system to be diagnosed after the data reconstruction can be consistent with the data distribution of the second multi-split air conditioning system.
While the inputs and outputs of the fault diagnosis model in the training process should be consistent with those of the actual fault diagnosis. Based on the above, the self-encoder model to be trained can be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system, and the self-encoder model after training is obtained.
In some embodiments, original operation data of the multi-split air conditioning system to be diagnosed is processed through an encoder compression and decoder of the self-encoder model and is converted into a data set similar to data distribution of the second multi-split air conditioning system by the self-encoder model, so that the fault diagnosis model can better perform fault diagnosis on the multi-split air conditioning system to be diagnosed.
As shown in fig. 13, the self-encoder includes two parts, an encoder and a decoder. The raw characteristic data (i.e., normal values and fault values of at least one target characteristic parameter of the second multi-split air conditioning system) may be mapped to encoder blocks of the hidden representation and the hidden representation mapped to decoder blocks of the reconstructed input. The encoder compresses the original characteristic data, and the compressed original characteristic data is restored by the decoder. The input features are extracted by minimizing the difference between the raw feature data and the reconstructed raw feature data, referred to as reconstruction errors. The method specifically comprises the following steps:
step 1, converting normal values and fault values of at least one target characteristic parameter into a characteristic vector matrix X= { X 1 ,X 2 ,…,X n (wherein X is 1 ,X 2 ,…,X n Representing normal values and fault values of at least one target characteristic parameter.
Step 2, the encoder compresses the original characteristic data to the hidden layer, which can be specifically shown in the following formula (4):
H＝σ e (W e X+B e ) Formula (4)
Wherein sigma e () As Sigmoid function, X is original characteristic data set, W e For encoder weight parameters, B e For encoder bias, H is the mapping of the original feature data at the hidden layer.
Step 3, the decoder reconstructs the original feature data and outputs the reconstructed original feature data, which can be specifically shown in the following formula (5):
wherein sigma d () As Sigmoid function, W d Is the weight parameter of the decoder, B d Is the deviation of the decoder and,is the reconstructed raw feature data.
Step 4, training the self-encoder model to minimize the reconstruction error and obtain the corresponding W e 、W d And
the reconstruction error calculation formula is shown in the following formula (6):
wherein,to reconstruct the error for softmax, X i For the original characteristic data>And reconstructing the characteristic data.
Further, the self-encoder model training is completed.
S302, inputting a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system into the trained self-encoder model to obtain the normal value and the fault value of the at least one target characteristic parameter after data reconstruction.
As is clear from S301, the self-encoder model has a data reconstruction function for data. In some embodiments, the normal value and the fault value of the at least one target characteristic parameter of the second multi-split air conditioning system are input into the self-encoder model after training is completed, the normal value and the fault value of the at least one target characteristic parameter after data reconstruction can be obtained, and then the fault diagnosis model to be trained can be trained according to the normal value and the fault value of the at least one target characteristic parameter after data reconstruction.
S303, training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after data reconstruction to obtain a trained fault diagnosis model.
In some embodiments, the normal value and the fault value of at least one target characteristic parameter after data reconstruction are used as a sample set to train the fault diagnosis model to be trained, so as to obtain a fault diagnosis model after training.
The foregoing embodiments focus on a training process for a fault diagnosis model provided in connection with embodiments of the present application, and in some embodiments, as shown in fig. 14, after the training of the fault diagnosis model is completed, the method further includes the following steps:
S401, acquiring first operation data of a third multi-split air conditioning system and second operation data of a fourth multi-split air conditioning system.
The third multi-split air conditioning system and the fourth multi-split air conditioning system are all multi-split air conditioning systems with target faults, and the machine type of the third multi-split air conditioning system is different from the machine type of the fourth multi-split air conditioning system.
S402, analyzing the first operation data and the second operation data, and determining fault values of M characteristic parameters of the third multi-split air conditioning system and fault values of M characteristic parameters of the fourth multi-split air conditioning system.
The manner in which the first operation data and the second operation data are parsed may refer to the description of S1012 above, and will not be described herein.
S403, extracting the fault value of at least one target characteristic parameter of the third multi-split air conditioning system and the fault value of at least one target characteristic parameter of the fourth multi-split air conditioning system from the fault values of M characteristic parameters of the third multi-split air conditioning system and the fault values of M characteristic parameters of the fourth multi-split air conditioning system.
For the description of S403, reference may be made to the description of S104, which is not repeated herein.
S404, inputting the fault value of at least one target characteristic parameter of the third multi-split air conditioning system and the fault value of at least one target characteristic parameter of the fourth multi-split air conditioning system into the trained self-encoder model to obtain the fault value of at least one target characteristic parameter of the third multi-split air conditioning system after data reconstruction and the fault value of at least one target characteristic parameter of the fourth multi-split air conditioning system after data reconstruction.
S405, inputting the fault value of at least one target characteristic parameter of the third multi-split air conditioning system after data reconstruction and the fault value of at least one target characteristic parameter of the fourth multi-split air conditioning system after data reconstruction into a trained fault diagnosis model to obtain a fault diagnosis result of the third multi-split air conditioning system and a fault diagnosis result of the fourth multi-split air conditioning system.
The fault diagnosis result of the third multi-split air conditioning system indicates the fault type of the third multi-split air conditioning system, and the fault diagnosis result of the fourth multi-split air conditioning system indicates the fault type of the fourth multi-split air conditioning system.
As shown in fig. 15, an embodiment of the present application provides a training apparatus for performing the training method of the fault diagnosis model shown in fig. 8, where the training apparatus 2000 includes: an acquisition unit 2001 and a processing unit 2002. In some embodiments, the training device 2000 may further include a storage unit 2003.
In some embodiments, the obtaining unit 2001 is configured to obtain normal values and fault values of M feature parameters of the first multi-split air conditioning system, where M is an integer greater than 1.
A processing unit 2002 for: determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein the characteristic offset space comprises a difference value between the fault value and the normal value of each characteristic parameter; correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of a second multi-split air conditioning system, wherein the model of the second multi-split air conditioning system is different from that of the first multi-split air conditioning system; determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and a characteristic offset space of the second multi-split air conditioning system; and training the fault diagnosis model to be trained based on normal values and fault values of M characteristic parameters of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, the processing unit 2002 is specifically configured to: determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system; and training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
In some embodiments, the processing unit 2002 is specifically configured to: training the self-encoder model to be trained based on the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a self-encoder model after training; inputting the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system into the trained self-encoder model to obtain the normal value and the fault value of at least one target characteristic parameter after data reconstruction; and training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after data reconstruction to obtain a trained fault diagnosis model.
In some embodiments, the acquiring unit 2001 is specifically configured to: acquiring historical operation data of a first multi-split air conditioning system; analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs.
In some embodiments, the storage unit 2003 is configured to store historical operating data of the first multi-split air conditioning system.
In some embodiments, storage unit 2003 is used to store the trained self-encoder model.
In some embodiments, the storage unit 2003 is used to store the trained fault diagnosis model.
The units in fig. 15 may also be referred to as modules, for example, the processing units may be referred to as processing modules.
The individual units in fig. 15 may be stored in a computer-readable storage medium if implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solution of the embodiments of the present application may be essentially or a part contributing to the prior art or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the methods described in the embodiments of the present application. The storage medium storing the computer software product includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The embodiment of the present application further provides a hardware structure schematic of a training device, as shown in fig. 16, where the training device 3000 includes a processor 3001, and optionally, a memory 3002 and a communication interface 3003 connected to the processor 3001. The processor 3001, the memory 3002, and the communication interface 3003 are connected by a bus 3004.
The processor 3001 may be a central processing unit (central processing unit, CPU), a general purpose processor network processor (network processor, NP), a digital signal processor (digital signal processing, DSP), a microprocessor, a microcontroller, a programmable logic device (programmable logic device, PLD), or any combination thereof. The processor 3001 may also be any other apparatus having processing functionality, such as a circuit, a device, or a software module. The processor 3001 may also include a plurality of CPUs, and the processor 3001 may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. A processor herein may refer to one or more devices, circuits, or processing cores for processing data (e.g., computer program instructions).
The memory 3002 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (random access memory, RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (electrically erasable programmable read-only memory, EEPROM), a compact disc read-only memory (compact disc read-only memory) or other optical disk storage, optical disk storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, as embodiments of the present application are not limited in this regard. The memory 3002 may be separate or integrated with the processor 3001. Wherein the memory 3002 may contain computer program code. The processor 3001 is configured to execute computer program code stored in the memory 3002, thereby implementing the methods provided by the embodiments of the present application.
The communication interface 3003 may be used to communicate with other devices or communication networks (e.g., ethernet, radio access network (radio access network, RAN), wireless local area network (wireless local area networks, WLAN), etc.). The communication interface 3003 may be a module, a circuit, a transceiver, or any device capable of enabling communications.
Bus 3004 may be a peripheral component interconnect standard (peripheral component interconnect, PCI) bus or an extended industry standard architecture (extended industry standard architecture, EISA) bus, among others. The bus 3004 may be classified into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in fig. 16, but not only one bus or one type of bus.
Embodiments of the present application also provide a computer-readable storage medium comprising computer-executable instructions that, when run on a computer, cause the computer to perform any of the methods provided by the above embodiments.
The present embodiments also provide a computer program product comprising computer-executable instructions which, when run on a computer, cause the computer to perform any of the methods provided by the above embodiments.
The embodiment of the application also provides a chip, which comprises: a processor and an interface through which the processor is coupled to the memory, which when executed by the processor executes a computer program or computer-executable instructions in the memory, cause any of the methods provided by the embodiments described above to be performed.
In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using a software program, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer-executable instructions. When the computer-executable instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer-executable instructions may be stored in or transmitted from one computer-readable storage medium to another, for example, from one website, computer, server, or data center by wired (e.g., coaxial cable, fiber optic, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). Computer readable storage media can be any available media that can be accessed by a computer or data storage devices including one or more servers, data centers, etc. that can be integrated with the media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.
Although the present application has been described herein in connection with various embodiments, other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed application, from a review of the figures, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.
Although the present application has been described in connection with specific features and embodiments thereof, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and drawings are merely exemplary illustrations of the present application as defined in the appended claims and are considered to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the present application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.
The foregoing is merely a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any changes or substitutions within the technical scope of the present disclosure should be covered in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (6)

1. A method of training a fault diagnosis model, comprising:
acquiring historical operation data of a first multi-split air conditioning system;
analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs; wherein M is an integer greater than 1, and the target fault includes one or more of refrigerant leakage, refrigerant charge excess, outdoor unit fouling, indoor unit fouling, and electronic expansion valve blockage;
determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein the characteristic offset space comprises a difference value between the fault value and the normal value of each characteristic parameter;
Correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of a second multi-split air conditioning system, wherein the model of the second multi-split air conditioning system is different from that of the first multi-split air conditioning system;
determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and a characteristic offset space of the second multi-split air conditioning system;
determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system; the M characteristic parameters comprise a compressor suction temperature value, a condenser pressure value, an evaporator pressure value, an environment temperature value where an outdoor unit is positioned, a compressor suction superheat degree and a compressor discharge superheat degree; the target characteristic parameters are characteristic parameters shared by the first multi-split air conditioning system and the second multi-split air conditioning system;
and training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
2. The method according to claim 1, wherein training the fault diagnosis model to be trained according to the normal value and the fault value of the at least one target feature parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model comprises:
training the self-encoder model to be trained based on a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained self-encoder model;
inputting the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system into a self-encoder model after training is completed, and obtaining the normal value and the fault value of at least one target characteristic parameter after data reconstruction;
and training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after the data reconstruction, so as to obtain the fault diagnosis model after training.
3. A training device, comprising:
an acquisition unit configured to: acquiring historical operation data of a first multi-split air conditioning system; analyzing the historical operation data, and determining normal values of M characteristic parameters of the first multi-split air conditioning system in a normal operation state of the first multi-split air conditioning system and fault values when a target fault occurs; wherein M is an integer greater than 1, and the target fault includes one or more of refrigerant leakage, refrigerant charge excess, outdoor unit fouling, indoor unit fouling, and electronic expansion valve blockage;
A processing unit for: determining a characteristic offset space of the first multi-split air conditioning system based on normal values and fault values of M characteristic parameters of the first multi-split air conditioning system, wherein the characteristic offset space comprises a difference value between the fault value and the normal value of each characteristic parameter; correcting the characteristic offset space of the first multi-split air conditioning system to obtain the characteristic offset space of a second multi-split air conditioning system, wherein the model of the second multi-split air conditioning system is different from that of the first multi-split air conditioning system; determining fault values of M characteristic parameters of the second multi-split air conditioning system based on normal values of the M characteristic parameters of the second multi-split air conditioning system and a characteristic offset space of the second multi-split air conditioning system; determining at least one target characteristic parameter from M characteristic parameters of the second multi-split air conditioning system based on normal values and fault values of the M characteristic parameters of the second multi-split air conditioning system; the M characteristic parameters comprise a compressor suction temperature value, a condenser pressure value, an evaporator pressure value, an environment temperature value where an outdoor unit is positioned, a compressor suction superheat degree and a compressor discharge superheat degree; the target characteristic parameters are characteristic parameters shared by the first multi-split air conditioning system and the second multi-split air conditioning system; and training the fault diagnosis model to be trained according to the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained fault diagnosis model.
4. The training device of claim 3, wherein the device comprises a plurality of sensors,
the processing unit is specifically configured to: training the self-encoder model to be trained based on a normal value and a fault value of at least one target characteristic parameter of the second multi-split air conditioning system to obtain a trained self-encoder model;
inputting the normal value and the fault value of at least one target characteristic parameter of the second multi-split air conditioning system into a self-encoder model after training is completed, and obtaining the normal value and the fault value of at least one target characteristic parameter after data reconstruction;
and training the fault diagnosis model to be trained based on the normal value and the fault value of at least one target characteristic parameter after the data reconstruction, so as to obtain the fault diagnosis model after training.
5. A training device, comprising: a processor and a memory;
the memory stores instructions executable by the processor;
the processor is configured to, when executing the instructions, cause the training device to implement the method of any of claims 1 to 2.
6. A computer readable storage medium comprising computer instructions which, when run on a computer, cause the computer to perform the method of any one of claims 1 to 2.
CN202210682549.8A 2022-04-11 2022-06-16 Training method and device for fault diagnosis model Active CN115095953B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202210682549.8A CN115095953B (en) 2022-06-16 2022-06-16 Training method and device for fault diagnosis model
PCT/CN2023/071431 WO2023197711A1 (en) 2022-04-11 2023-01-09 Multi-split air-conditioning system, fault positioning method, and fault diagnosis model training method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210682549.8A CN115095953B (en) 2022-06-16 2022-06-16 Training method and device for fault diagnosis model

Publications (2)

Publication Number Publication Date
CN115095953A CN115095953A (en) 2022-09-23
CN115095953B true CN115095953B (en) 2024-04-09

Family

ID=83290839

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210682549.8A Active CN115095953B (en) 2022-04-11 2022-06-16 Training method and device for fault diagnosis model

Country Status (1)

Country Link
CN (1) CN115095953B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023197711A1 (en) * 2022-04-11 2023-10-19 青岛海信日立空调系统有限公司 Multi-split air-conditioning system, fault positioning method, and fault diagnosis model training method

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107015486A (en) * 2017-04-20 2017-08-04 重庆大学 A kind of air-conditioner water system regulating valve intelligent fault diagnosis method
KR20210055998A (en) * 2019-11-08 2021-05-18 엘지전자 주식회사 Air conditioner controlling method and air conditioner controlling apparatus
CN113177594A (en) * 2021-04-29 2021-07-27 浙江大学 Air conditioner fault diagnosis method based on Bayesian optimization PCA-extreme random tree
CN113203589A (en) * 2021-04-29 2021-08-03 华中科技大学 Distributed fault diagnosis method and system for multi-split air conditioning system
JPWO2021260764A1 (en) * 2020-06-22 2021-12-30
CN114065613A (en) * 2021-10-27 2022-02-18 中国华能集团清洁能源技术研究院有限公司 Multi-working-condition process industrial fault detection and diagnosis method based on deep migration learning
CN114492534A (en) * 2022-01-30 2022-05-13 华中科技大学 Construction method and application of cross-size motor bearing fault diagnosis model
CN114484735A (en) * 2022-03-11 2022-05-13 青岛海信日立空调系统有限公司 Multi-split system fault diagnosis and energy-saving potential identification method and multi-split system
WO2022112991A1 (en) * 2020-11-30 2022-06-02 Jio Platforms Limited System and method of predicting failures

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109753998A (en) * 2018-12-20 2019-05-14 山东科技大学 The fault detection method and system, computer program of network are generated based on confrontation type
KR102279351B1 (en) * 2020-11-27 2021-07-20 성한 주식회사 A fault diagnosis method of based ai on optimization of air conditioner

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107015486A (en) * 2017-04-20 2017-08-04 重庆大学 A kind of air-conditioner water system regulating valve intelligent fault diagnosis method
KR20210055998A (en) * 2019-11-08 2021-05-18 엘지전자 주식회사 Air conditioner controlling method and air conditioner controlling apparatus
JPWO2021260764A1 (en) * 2020-06-22 2021-12-30
WO2022112991A1 (en) * 2020-11-30 2022-06-02 Jio Platforms Limited System and method of predicting failures
CN113177594A (en) * 2021-04-29 2021-07-27 浙江大学 Air conditioner fault diagnosis method based on Bayesian optimization PCA-extreme random tree
CN113203589A (en) * 2021-04-29 2021-08-03 华中科技大学 Distributed fault diagnosis method and system for multi-split air conditioning system
CN114065613A (en) * 2021-10-27 2022-02-18 中国华能集团清洁能源技术研究院有限公司 Multi-working-condition process industrial fault detection and diagnosis method based on deep migration learning
CN114492534A (en) * 2022-01-30 2022-05-13 华中科技大学 Construction method and application of cross-size motor bearing fault diagnosis model
CN114484735A (en) * 2022-03-11 2022-05-13 青岛海信日立空调系统有限公司 Multi-split system fault diagnosis and energy-saving potential identification method and multi-split system

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
New fault diagnostic strategies for refrigerant charge fault in a VRF system using hybrid machine learning method;Ding, XL (Ding, Xinlei) 等;JOURNAL OF BUILDING ENGINEERING;20210131;全文 *
基于小波神经网络方法的空调系统传感器漂移故障诊断;杨云雨;杜志敏;晋欣桥;;能源技术;20080220(01);全文 *
多联机系统故障类型识别及故障细化诊断模型研究;刘倩 等;制冷学报;全文 *

Also Published As

Publication number Publication date
CN115095953A (en) 2022-09-23

Similar Documents

Publication Publication Date Title
CN111964236B (en) Method and device for calculating outdoor environment temperature and air conditioner
CN115095953B (en) Training method and device for fault diagnosis model
CN112728712B (en) Multi-split air conditioner running capacity detection method, multi-split air conditioner, storage medium and device
EP3196563A1 (en) Control device for air conditioning system, air conditioning system, and method for determining anomaly of air conditioning system
CN114754413B (en) Multi-split air conditioning system and fault positioning method
CN115264973B (en) Water chilling unit and ideal energy efficiency ratio determining method thereof
CN114877550B (en) Control method of throttle valve, water chiller, and computer-readable storage medium
CN111397104B (en) Control method and device of air conditioning system, control equipment, medium and air conditioning system
CN111503924B (en) Four-way valve fault detection method and device
CN113465131A (en) Air conditioner and method for determining refrigerant quantity
CN111306695B (en) Compressor load data optimization method and device, computer equipment and storage medium
CN114110999A (en) Method and device for controlling opening degree of internal machine of multi-connected unit, electronic equipment and storage medium
CN114992798B (en) Air conditioning system and indoor machine capability calculating method thereof
CN116336608B (en) Abnormal data transmission method and device for energy-saving air conditioning system
CN115111825B (en) Method and device for determining operation frequency of compressor
WO2024114675A1 (en) Multi-split air conditioning system and control method thereof
CN116608150B (en) Method, apparatus and computer readable storage medium for determining rotational speed
CN115978747A (en) Heat pump unit and control method thereof
CN114608848B (en) Method, device, system and storage medium for detecting operation capability of multi-split system
CN114110994B (en) Air conditioning system throttling control method and device based on pipe length
CN114593497B (en) Bypass loop refrigerant flow detection method, device, system and storage medium
CN114576825B (en) Air conditioner control method, air conditioner and readable storage medium
CN114811743A (en) Air conditioning system and indoor unit capacity calculation method thereof
CN116428711A (en) Multi-split air conditioning system and control method thereof
CN115950045A (en) Multi-split air conditioning system and control method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant