Refrigeration cycle device and method for controlling refrigeration cycle device

Abstract

This refrigeration cycle device comprises: a plurality of sensors for detecting a control amount; a plurality of actuators; and a control device that outputs a drive command indicating an operation amount for the plurality of actuators by using sensor group information indicating detection results from the plurality of sensors and drive amount information indicating a drive amount for the plurality of actuators. The control device includes: a feedback control unit that outputs a feedback command for feedback control on the basis of the sensor group information; a characteristic identification unit that sequentially identifies dynamic characteristics indicating the relationship between the operation amount and the control amount for a plurality of combinations of the plurality of actuators and a plurality of control amounts, and outputs identification information indicating the identified dynamic characteristics, on the basis of the sensor group information and the drive amount information; a feedforward control unit that outputs a feedforward command for feedforward control on the basis of the identification information and the sensor group information; and an adder that adds the feedback command and the feedforward command together and outputs the drive command. The characteristic identification unit feeds back the dead time of the dynamic characteristics estimated from the identification information to the process for identifying the next dynamic characteristics.

Description

Refrigeration cycle device and control method for refrigeration cycle device

[0001] The present disclosure relates to a refrigeration cycle device and a control method for a refrigeration cycle device.

[0002] A refrigeration cycle device is composed of a compression section, an expansion section, a condensation section, and an evaporation section. Generally, the compression section has a motor or an internal combustion engine for compressing the refrigerant, and the expansion section may use an electronic expansion valve. The condensation section and the evaporation section may also be equipped with motors for rotating fans for blowing air through them. Thus, a refrigeration cycle device is equipped with multiple actuators. In a refrigeration cycle device, various parameters such as discharge temperature and pressure, evaporation temperature, degree of superheat, and degree of subcooling are controlled by operating these multiple actuators. Therefore, a refrigeration cycle device is a multi-input, multi-output system that handles multiple manipulated variables (inputs) and multiple controlled variables (outputs).

[0003] In a multi-input, multi-output system, each manipulated variable affects all controlled variables. Therefore, if each manipulated variable corresponds to only one controlled variable and is controlled to follow a target value, the manipulated variables will interfere with each other. Therefore, it is desirable to control the manipulated variable of each actuator while taking into account interference from other manipulated variables. For example, Patent Document 1 discloses a technology in which, in a case where a compressor controls the evaporation temperature and an expansion valve controls the superheat degree, fluctuations in the evaporation temperature are considered as disturbances and the opening of the expansion valve is adjusted to suppress the disturbances, thereby suppressing fluctuations in the superheat degree.

[0004] JP 2015-114036 A

[0005] The technology of Patent Document 1 requires that dynamic characteristics related to the opening degree of the expansion valve and the evaporation temperature be identified in advance. However, the dynamic characteristics change depending on the dryness fraction of the refrigerant, the refrigerant flow rate, the temperature and pressure at each part of the refrigeration cycle device, and the piping length. Therefore, if the actuator is controlled using the dynamic characteristics at one or several representative points, fluctuations in the controlled variable cannot be suppressed at operating points where dynamic characteristics different from those at the representative points are exhibited, and the operation of the actuator may become unstable.

[0006] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a refrigeration cycle device that can stabilize the operation of an actuator, and a control method for a refrigeration cycle device.

[0007] A refrigeration cycle apparatus according to the present disclosure includes a control device that outputs drive commands indicating manipulated variables for the multiple actuators using multiple sensors, multiple actuators, and sensor group information indicating detection results of the multiple sensors and drive amount information indicating drive amounts for the multiple actuators. The control device includes a feedback control unit that outputs feedback commands for feedback control based on the sensor group information, a characteristic identification unit that sequentially identifies dynamic characteristics indicating the relationship between the manipulated variables and the controlled variables for multiple combinations of the multiple actuators and the multiple controlled variables based on the sensor group information and the drive amount information, and outputs identification information indicating the identified dynamic characteristics, a feedforward control unit that outputs feedforward commands for feedforward control based on the identification information and the sensor group information, and an adder that adds the feedback command and the feedforward command to output a drive command. The characteristic identification unit feeds back the dead time of the dynamic characteristics estimated from the identification information to the next dynamic characteristic identification process.

[0008] A control method for a refrigeration cycle device includes a control device that outputs drive commands indicating operation amounts of the multiple actuators using multiple sensors for detecting control amounts, multiple actuators, sensor group information indicating the detection results of the multiple sensors, and drive amount information indicating the drive amounts of the multiple actuators. The control device has a feedback control step that outputs feedback commands for feedback control based on the sensor group information, a characteristic identification step that sequentially identifies dynamic characteristics indicating the relationship between the operation amounts and the control amounts for multiple combinations of the multiple actuators and the multiple control amounts based on the sensor group information and the drive amount information, and outputs identification information indicating the identified dynamic characteristics, a feedforward control step that outputs a feedforward command for feedforward control based on the identification information and the sensor group information, and an addition step that adds up the feedback command and the feedforward command to output a drive command. In the characteristic identification step, the dead time of the dynamic characteristics estimated from the identification information is fed back to the next characteristic identification step.

[0009] According to the refrigeration cycle apparatus and the control method for the refrigeration cycle apparatus of the present disclosure, the sequentially identified dynamic characteristics are reflected in the drive command to the actuator. Also, the dead time estimated from the identification information is fed back to the next identification process. Therefore, the operation of the actuator can be stabilized.

[0010] 1 is a schematic diagram showing a configuration example of a refrigeration cycle device according to embodiment 1. FIG. 2 is a hardware configuration diagram showing a refrigeration cycle device according to embodiment 1. FIG. 3 is a block diagram showing a control device according to embodiment 1. FIG. 4 is a block diagram showing a feedback control unit according to embodiment 1. FIG. 5 is a block diagram showing a characteristic identification unit according to embodiment 1. FIG. 6 is a block diagram showing a recursive characteristic identification unit according to embodiment 1. FIG. 7 is a block diagram showing an input / output shift unit according to embodiment 1. FIG. 8 is a diagram showing an estimated value of a step response according to embodiment 1. FIG. 9 is a diagram for explaining the effect of the characteristic identification unit according to embodiment 1. FIG. 10 is a block diagram showing a feedforward control unit according to embodiment 1. FIG. 11 is a flowchart showing the operation of the control device according to embodiment 1. FIG. 12 is a block diagram showing a control device according to embodiment 2. FIG. 13 is a block diagram showing a feedforward control unit according to embodiment 2. FIG. 14 is a block diagram showing a control device according to embodiment 3. FIG. 15 is a block diagram showing a feedforward control unit according to embodiment 3.

[0011] Embodiment 1. Fig. 1 is a schematic diagram showing an example of the configuration of a refrigeration cycle apparatus 1 according to Embodiment 1. The refrigeration cycle apparatus 1 is, for example, an air conditioner that conditions air indoors. As shown in Fig. 1, the refrigeration cycle apparatus 1 includes an outdoor unit 10, an indoor unit 11, and a control device 12. The outdoor unit 10 includes a compressor 2, a four-way valve 3, an outdoor heat exchanger 4, and an outdoor fan 5. The indoor unit 11 includes an indoor heat exchanger 7, an indoor fan 8, and an expansion valve 9. The outdoor unit 10 and the indoor unit 11 are connected by piping 13 consisting of a liquid pipe and a gas pipe.

[0012] 1 illustrates an example in which the refrigeration cycle apparatus 1 has only one outdoor unit 10 and one indoor unit 11, and only one compressor 2, one outdoor fan 5, one indoor fan 8, and one expansion valve 9. However, the refrigeration cycle apparatus 1 may have multiple indoor units 11 or outdoor units 10, or may have multiple compressors 2, multiple outdoor fans 5, multiple indoor fans 8, and multiple expansion valves 9. The expansion valve 9 of the refrigeration cycle apparatus 1 may be provided in the outdoor unit 10. Furthermore, in addition to the outdoor unit 10 and the indoor unit 11, the refrigeration cycle apparatus 1 may have a relay that relays the flow of refrigerant between the outdoor unit 10 and the indoor unit 11.

[0013] The compressor 2 draws in refrigerant in a low-temperature, low-pressure state, compresses it, and discharges it as refrigerant in a high-temperature, high-pressure state. The four-way valve 3 switches the flow direction of the refrigerant in the refrigerant circuit. The outdoor heat exchanger 4 exchanges heat between the refrigerant and outdoor air. The outdoor heat exchanger 4 functions as a condenser during cooling operation and as an evaporator during heating operation. The outdoor fan 5 is a device that sends outdoor air to the outdoor heat exchanger 4.

[0014] The indoor heat exchanger 7 exchanges heat between the indoor air and the refrigerant. The indoor heat exchanger 7 functions as an evaporator during cooling operation and as a condenser during heating operation. The indoor fan 8 is a device that sends indoor air to the indoor heat exchanger 7. The expansion valve 9 reduces the pressure of the refrigerant to expand it, and is, for example, an electronic expansion valve.

[0015] The compressor 2, the outdoor fan 5, the indoor fan 8, and the expansion valve 9 each have an actuator (motor). The control device 12 inputs operation amounts to these actuators, thereby adjusting the drive amounts of these actuators. Hereinafter, the compressor 2, the outdoor fan 5, the indoor fan 8, and the expansion valve 9 may be collectively referred to as an actuator group 21. The drive amount of the actuators varies depending on the devices in which the actuators are installed, but is represented by, for example, the rotation speed (or operating frequency) of the compressor 2, the rotation speeds of the outdoor fan 5 and the indoor fan 8, and the opening of the expansion valve 9.

[0016] The refrigeration cycle apparatus 1 has a first sensor 14, a second sensor 15, a third sensor 16, a fourth sensor 17, and a fifth sensor 18. The first sensor 14 is a temperature sensor that detects the temperature of the refrigerant discharged from the compressor 2. The detection result of the first sensor 14 corresponds to the discharge temperature, which is one of the multiple controlled variables of the refrigeration cycle apparatus 1. The second sensor 15 is a temperature sensor that detects the pressure of the refrigerant discharged from the compressor 2. The detection result of the second sensor 15 corresponds to the discharge pressure, which is one of the multiple controlled variables of the refrigeration cycle apparatus 1.

[0017] The third sensor 16 is a temperature sensor that detects the temperature of the outdoor heat exchanger 4. The detection result of the third sensor 16 is one of the multiple controlled variables of the refrigeration cycle apparatus 1. The fourth sensor 17 and the fifth sensor 18 are temperature sensors that detect the temperatures of the refrigerant flowing upstream and downstream of the indoor heat exchanger 6. The difference between the detection result of the fourth sensor 17 and the detection result of the fifth sensor 18 corresponds to the temperature difference between the top and bottom of the indoor heat exchanger 6, which is one of the multiple controlled variables of the refrigeration cycle apparatus 1.

[0018] As described above, the first sensor 14, the second sensor 15, the third sensor 16, the fourth sensor 17, and the fifth sensor 18 are sensors for detecting control variables in the refrigeration cycle apparatus 1, and may be collectively referred to as a sensor group 22. The refrigeration cycle apparatus 1 may use sensors other than those described above as sensors for detecting control variables. For example, a sensor for detecting the temperature of air drawn into the indoor unit 11 (room temperature) and a sensor for detecting the temperature of air blown out from the indoor unit 11 may be provided in the housing (not shown) of the indoor unit 11. A sensor for detecting the temperature or pressure of the refrigerant drawn into the compressor 2 may also be used. Furthermore, a sensor for detecting the room temperature may be provided in a remote control for operating the refrigeration cycle apparatus 1.

[0019] As described above, the refrigeration cycle apparatus 1 is a multi-input multi-output system that handles a plurality of manipulated variables (inputs) and a plurality of controlled variables (outputs).

[0020] The control device 12 controls the operation of each actuator provided in the outdoor unit 10 and the indoor unit 11. The control device 12 controls the rotation speed of the compressor 2, the opening degree of the expansion valve 9, and the rotation speeds of the outdoor fan 5 and the indoor fan 8 to control various control variables, such as the temperature and pressure, superheat degree, and subcooling degree, in each device and piping 13 to achieve target values. The configuration and operation of the control device 12 will be described in detail below. While FIG. 1 illustrates a case in which the control device 12 is provided inside the outdoor unit 10, the control device 12 may be provided inside the indoor unit 11 or outside the outdoor unit 10 and the indoor unit 11. The control device 12 may also be divided into an outdoor control device provided in the outdoor unit 10 and an indoor control device provided in the indoor unit 11, and these may be configured to be able to communicate with each other. In this case, it is preferable to assign in advance which indoor control device and which outdoor control device will send control signals to which actuators. A communication repeater may also be provided to relay communication between the indoor control device and the outdoor control device.

[0021] (Cooling Operation) Here, the operation of the refrigeration cycle apparatus 1 will be described from the perspective of refrigerant flow. First, cooling operation will be described. The refrigeration cycle apparatus 1 performs cooling operation by switching the four-way valve 3 so that the discharge side of the compressor 2 is connected to the outdoor heat exchanger 4. During cooling operation, the refrigerant drawn into the compressor 2 is compressed by the compressor 2 and discharged in a high-temperature, high-pressure gas state. The high-temperature, high-pressure gas refrigerant discharged from the compressor 2 passes through the four-way valve 3 and flows into the outdoor heat exchanger 4, which functions as a condenser. The refrigerant that flows into the outdoor heat exchanger 4 exchanges heat with outdoor air sent by the outdoor fan 5, condenses, and liquefies. The liquid refrigerant flows into the expansion valve 9, where it is decompressed and expanded to become a low-temperature, low-pressure, two-phase gas-liquid refrigerant. The two-phase gas-liquid refrigerant flows into the indoor heat exchanger 7, which functions as an evaporator. The refrigerant that flows into the indoor heat exchanger 7 exchanges heat with indoor air sent by the indoor fan 8, evaporating and gasifying. At this time, the indoor air is cooled to cool the room. After that, the evaporated refrigerant in a low-temperature, low-pressure gas state passes through the four-way valve 3 and is drawn into the compressor 2.

[0022] (Heating Operation) Next, the heating operation will be described. The refrigeration cycle apparatus 1 performs heating operation by switching the four-way valve 3 so that the discharge side of the compressor 2 is connected to the indoor heat exchanger 7. During heating operation, the refrigerant drawn into the compressor 2 is compressed by the compressor 2 and discharged in a high-temperature, high-pressure gas state. The high-temperature, high-pressure gas refrigerant discharged from the compressor 2 passes through the four-way valve 3 and flows into the indoor heat exchanger 7, which functions as a condenser. The refrigerant that flows into the indoor heat exchanger 7 exchanges heat with indoor air sent by the indoor fan 8, condenses, and liquefies. At this time, the indoor air is heated, thereby heating the room. The liquid refrigerant flows into the expansion valve 9, where it is decompressed and expanded to become a low-temperature, low-pressure, two-phase gas-liquid refrigerant. The two-phase gas-liquid refrigerant flows into the outdoor heat exchanger 4, which functions as an evaporator. The refrigerant that flows into the outdoor heat exchanger 4 exchanges heat with outdoor air sent by the outdoor fan 5, evaporates, and gasifies. The evaporated refrigerant in a gaseous state at low temperature and pressure then passes through the four-way valve 3 and is sucked into the compressor 2 .

[0023] <Control device 12> FIG. 2 is a hardware configuration diagram showing the refrigeration cycle apparatus 1 according to the first embodiment. As shown in FIG. 2, the control device 12 includes a processing device 23 and a storage device 24. The processing device 23 is a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA). The processing device 23 may be various logic circuits or various signal processing circuits other than those described above. The processing device 23 may also be configured by combining multiple circuits of the same or different types among these, and may share and execute each process. If the processing device 23 includes a processor, the functions of the processing device 23 are realized by the processor reading a program stored in a memory.

[0024] The storage device 24 is composed of a RAM (Random Access Memory) that can read and write data from and to the processing device 23, and a ROM (Read Only Memory) that can read data from and to the processing device 23. The storage device 24 may also have non-volatile or volatile semiconductor memory such as a flash memory, an EPROM, or an EEPROM. The storage device 24 may also have a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, a DVD, or the like.

[0025] The control device 12 is connected to the actuator group 21 and the sensor group 22 so as to be able to communicate with them via wired or wireless communication. The control device 12 receives sensor group information SI, which is output from the sensor group 22, and drive amount information AI from the actuator group 21, at each predetermined acquisition period. The sensor group information SI includes the detection results of each sensor. The drive amount information AI includes information indicating the drive amount of each actuator. The processing device 23 of the control device 12 calculates a drive command AR based on this information and outputs it to the actuator group 21. The drive command AR is information used to control the rotation speed of the compressor 2, the opening of the expansion valve 9, and the rotation speeds of the outdoor fan 5 and the indoor fan 8, and indicates these manipulated variables. The manipulated variables refer to the amount of fluctuation in the drive amount for controlling the controlled variables to target values.

[0026] 3 is a block diagram showing the control device 12 according to embodiment 1. As shown in FIG. 3, the control device 12 includes a feedback control unit 31, a characteristic identification unit 32, a feedforward control unit 33a, and an adder .

[0027] The feedback control unit 31 outputs a feedback command FB based on the sensor group information SI. The feedback command FB is information indicating a manipulated variable for stabilizing the control system from the viewpoint of feedback control. The function of the feedback control unit 31 will be described in detail later.

[0028] The characteristic identification unit 32 sequentially identifies the dynamic characteristics based on the sensor group information SI and the drive amount information AI. Here, the dynamic characteristics indicate the relationship between the manipulated variable and the controlled variable. The dynamic characteristics also indicate the degree of fluctuation in the sensor output result (controlled variable) when a certain actuator is operated by the manipulated variable. Sequentially identifying the dynamic characteristics means that the dynamic characteristics identification process is performed each time the sensor group information SI and the drive amount information AI are input to the control device 12 after the acquisition period has elapsed.

[0029] The characteristic identifying unit 32 identifies dynamic characteristics for multiple combinations of actuators and controlled variables. For example, the characteristic identifying unit 32 identifies dynamic characteristics for a combination of the rotation speed of the compressor 2 and the discharge temperature, a combination of the rotation speed of the compressor 2 and the temperature of the outdoor heat exchanger 4, or a combination of the rotation speed of the compressor 2 and the discharge pressure. The characteristic identifying unit 32 also identifies dynamic characteristics for a combination of the opening of the expansion valve 9 and the discharge temperature, or a combination of the opening of the expansion valve 9 and the temperature difference between the upper and lower temperatures of the indoor heat exchanger 6. However, the above combinations are merely examples. Which combination of dynamic characteristics is identified depends on the specifications of the refrigeration cycle apparatus 1 and whether it is intended for home use, commercial use, or hot water supply.

[0030] The characteristic identifying section 32 outputs identification information IDI indicating the identified dynamic characteristic to the feedforward control section 33a. The function of the characteristic identifying section 32 will be described in detail later.

[0031] The feedforward control unit 33a outputs a feedforward command FF based on the sensor group information SI, the drive amount information AI, and the identification information IDI. From the viewpoint of feedforward control, the feedforward command FF is information indicating an operation amount for suppressing the influence of interference that occurs when driving multiple actuators. The function of the feedforward control unit 33a will be described in detail later.

[0032] The adder 34 adds a feedback command FB for feedback control and a feedforward command FF for feedforward control, and outputs the result as a drive command AR to the actuator group 21. Note that adding the feedback command FB and the feedforward command FF means adding the manipulated variable indicated by the feedback command FB and the manipulated variable indicated by the feedforward command FF for each actuator.

[0033] <Feedback control unit> Fig. 4 is a block diagram showing the feedback control unit 31 according to the first embodiment. As shown in Fig. 4, the feedback control unit 31 includes a sensor information processor 41, a target value generator 42, a sensor information selector 43, a feedback controller 44, and a control command selector 45. The function of the feedback control unit 31 is executed independently for each actuator of the refrigeration cycle apparatus 1. The feedback control unit 31 collects outputs of the control command selectors 45 corresponding to all the actuators of the refrigeration cycle apparatus 1 and outputs the collected output as a feedback command FB.

[0034] The sensor information processor 41 processes the sensor group information SI using decimation, a low-pass filter, a high-pass filter, and a band-pass filter. The sensor information processor 41 outputs the processing results for the sensor group information SI as processed sensor group information FSI to the target value generator 42 and the sensor information selector 43.

[0035] The target value generator 42 calculates a target value corresponding to a control variable that is a control objective based on the processed sensor group information FSI, and outputs the calculated target value as control target value information TM. As described above, the processing of the feedback control unit 31 is executed multiple times for each actuator. In each processing of the target value generator 42, a predetermined control variable for the corresponding actuator is selected from the processed sensor group information FSI, and a target value corresponding to the selected control variable is output as control target value information TM. The predetermined control variable for the actuator means the control variable that is controlled by operating the actuator.

[0036] The sensor information selector 43 selects information indicating the detection results for the control variable to be controlled from the processed sensor group information FSI and outputs it as controlled variable measurement value information TD. As described above, the processing of the feedback control unit 31 is executed multiple times corresponding to each actuator. In each processing of the sensor information selector 43, a control variable predetermined for the corresponding actuator at that time is selected from the processed sensor group information FSI as controlled variable measurement value information TD.

[0037] The feedback controller 44 performs feedback control based on the controlled variable measurement value information TD and the control target value information TM, and outputs a pre-selection feedback command FB0. The feedback control method used here may be a known PI control, PID control, proportional-differential-leading PID control, optimal regulator, or two-degree-of-freedom control. Furthermore, when multiple pieces of control target value information TM and controlled variable measurement value information TD are used for one actuator, feedback control is performed independently for each pair of control target value information TM and controlled variable measurement value information TD. The respective execution results are then output as a pre-selection feedback command FB0. The pre-selection feedback command FB0 is information indicating the manipulated variable for the actuator.

[0038] When multiple pieces of control target value information TM and controlled variable measurement value information TD are used for one actuator, the control command selector 45 selects one from multiple pre-selection feedback commands FB0. For example, the control command selector 45 selects the safest one from the multiple pre-selection feedback commands FB0, for example, the smaller one in the case of the rotation speed of the compressor 2. Priorities may be determined in advance for the multiple pieces of control target value information TM, and selection may be performed based on the priorities. In this case, the priority for each controlled variable may be changed depending on the relationship between the control target value information TM and a preset threshold value. The control command selector 45 outputs one selected from the multiple pre-selection feedback commands FB0 as the feedback command FB. When only one pair of control target value information TM and controlled variable measurement value information TD is used for one actuator, the control command selector 45 outputs the pre-selection feedback command FB0 corresponding to this pair as the feedback command FB.

[0039] <Characteristics Identification Unit> Fig. 5 is a block diagram showing the characteristics identification unit 32 according to embodiment 1. As shown in Fig. 5, the characteristics identification unit 32 includes a recursive characteristics identification unit 51 and a dead time estimator 52.

[0040] The sequential characteristic identification unit 51 identifies dynamic characteristics while performing feedback control based on the sensor group information SI, drive amount information AI, and dead time information DEI so that the dead time indicated by the dead time information DEI is eliminated (becomes 0 or 1). The sequential characteristic identification unit 51 outputs identification information IDI indicating the identified dynamic characteristics. The function of the sequential characteristic identification unit 51 will be described in detail later.

[0041] The dead time estimator 52 estimates the dead time based on the identification information IDI and outputs the result as dead time information DEI. The dead time is estimated for each dynamic characteristic specified by a combination of an input (manipulated variable) and an output (controlled variable).

[0042] <Sequential characteristic identification unit> Fig. 6 is a block diagram showing the sequential characteristic identification unit 51 according to embodiment 1. As shown in Fig. 6, the sequential characteristic identification unit 51 includes an input / output shift unit 53, a PID controller 54, and a sequential characteristic identifier 55.

[0043] The input / output shift unit 53 delays the sensor group information SI and the drive amount information AI based on the input / output dead time information IODI, thereby outputting shifted input / output information SSAI in which the influence of dead time of the dynamic characteristics is offset. The function of the input / output shift unit 53 will be described in detail later.

[0044] The PID controller 54 calculates and outputs input / output dead time information IODI indicating the time by which the sensor group information SI and the drive amount information AI are delayed in order to eliminate the dead time indicated by the dead time information DEI (setting it to 0 or 1). The PID controller 54 can be, for example, a known PID feedback controller.

[0045] The recursive characteristic identifier 55 identifies dynamic characteristics based on the shifted input / output information SSAI and outputs the identification result as identification information IDI. The recursive characteristic identifier 55 performs processing every time an acquisition period elapses. As an algorithm for recursive identification, for example, a well-known recursive subspace identification method is used. The recursive subspace identification method is a technique for identifying a state space model by recursively estimating matrix parameters G and H of an input / output matrix equation expressed as Y = G × Z + H × U and performing singular value decomposition on G × Z or G.

[0046] Here, U is a Hankel matrix in which input values ​​of the shifted input / output information SSAI are arranged for each sample. Also, Y is a Hankel matrix in which output values ​​of the shifted input / output information SSAI are arranged for each sample. Also, Z is a Hankel matrix in which input values ​​or input / output values ​​of the shifted input / output information SSAI are arranged for each sample.

[0047] The identification information IDI includes the recursively estimated parameters G and H, the state-space model identified using singular value decomposition, the eigenvalues ​​of the identified state-space model, and the DC gain of the identified state-space model.

[0048] <Input / Output Shift Unit> Fig. 7 is a block diagram showing the input / output shift unit 53 according to embodiment 1. As shown in Fig. 7, the input / output shift unit 53 includes an information processor 56, an input / output shift amount estimator 57, and an input / output shifter 58.

[0049] The information processor 56 performs decimation and at least one filtering process of a low-pass filter, a high-pass filter, and a band-pass filter on the sensor group information SI and the drive amount information AI, and outputs the result as input / output information SAI.

[0050] The input / output shift amount estimator 57 calculates and outputs input / output shift amount information SHI based on the input / output dead time information IODI. Here, the number of pieces of input / output dead time information IODI is equal to the number of combinations of inputs (operation variables) and outputs (control variables), but the number of pieces of input / output shift amount information SHI is equal to the total number of inputs (operation variables) and outputs (control variables). For this reason, the number of pieces of input / output dead time information IODI and the number of pieces of input / output shift amount information SHI may differ. Furthermore, the input / output shift amount information SHI is restricted to be a positive integer in order to be processed by the input / output shifter 58. For this reason, the input / output shift amount estimator 57 approximately calculates the input / output shift amount information SHI from the input / output dead time information IODI.

[0051] As an example, in the case of a two-input, three-output system, 2 × 3 = 6 pieces of input / output dead time information IODI are obtained, but the number of inputs and outputs that can be delayed by the input / output shifter 58 described below is 2 + 3 = 5. In this case, a parameter s that minimizes the evaluation function J expressed by the following equation is estimated using a known numerical optimization method such as the interior point method.

[0052]

[0053] Here, ２ represents the 2-norm, and l^ ａ，ｂ (The "^" is assumed to be attached to the beginning of the immediately preceding character) represents input / output dead time information IODI from the a-th input to the b-th output, and s represents input / output shift amount information SHI.

[0054] The input / output shifter 58 delays the input / output information SAI by the amount of input / output shift amount information SHI, and outputs it as shifted input / output information SSAI.

[0055] As an example, a two-input, one-output second-order ARX (Auto-Regressive eXogeneous) model will be used for explanation. First, a case where input / output shift amount information SHI is not used will be explained. In this case, Y[k] and YU[k] of the following formula are output as post-shift input / output information SSAI. Y[k] = A1 x Y[k-1] + A2 x Y[k-2] + B1 x U1[k-1] + B2 x U1[k-2] + C1 x U2[k-1] + C2 x U2[k-2] = ABC YU[k]

[0056] A1, A2, B1, B2, C1, and C2 are parameters representing dynamic characteristics, and ABC is a vector arranging A1, A2, B1, B2, C1, and C2. Furthermore, Y is processed drive amount information, U1 and U2 are processed sensor information, and YU is a vector arranging Y[k-1], Y[k-2], U1[k-1], U1[k-2], U2[k-1], and U2[k-2]. Furthermore, the operator · represents an inner product, and k, k-1, and k-2 represent data sample numbers.

[0057] On the other hand, when input / output shift amount information SHI is used, the ARX model is expressed by the following equation when the input / output shift amount information SHI of U1 and U2 is S1 and S2. In this case, Y[k] and YU2[k] of the following equation are output as post-shift input / output information SSAI. Y[k]=A1×Y[k-1]+A2×Y[k-2]+B1×U1[k-1-S1]+B2×U1[k-2-S1]+C1×U2[k-1-S2]+C2×U2[k-2-S2]=ABC·YU2[k]

[0058] YU2 represents a vector in which Y[k-1], Y[k-2], U1[k-1-S1], U1[k-2-S1], U2[k-1-S2], and U2[k-2-S2] are arranged.

[0059] Here, a method for estimating the dead time by the dead time estimator 52 will be described. Of the identification information IDI, the parameter H includes an estimated value of the input / output impulse response. FIG. 8 is a diagram showing an estimated value of the step response according to the first embodiment. FIG. 8 shows an estimated value of the step response obtained by integrating the estimated impulse response value. The dead time of each dynamic characteristic specified by a combination of the input (manipulated variable) and the output (controlled variable) may be estimated from a tangent at an inflection point of the step response shown in FIG. 8. Alternatively, the dead time may be estimated by extracting the rising timing of the step response shown in FIG. 8.

[0060] <Effects of the characteristic identification unit> The characteristic identification unit 32 can be considered to operate as a type of control system by regarding the recursive characteristic identifier 55 as the control object, the dead time estimator 52 as a sensor unit, the PID controller 54 as a feedback controller, and the input / output shift unit 53 as an actuator unit. The control target is the dead time information DEI that is the output of the dead time estimator 52, and the shifted input / output information SSAI used for identification is shifted so that the dead time information DEI becomes 0 or 1. By using this method, the dead time can be estimated simultaneously while maintaining the model to be identified at a low order.

[0061] For comparison, consider a case where identification is performed using a recursive characteristic identifier that does not include the dead time estimator 52, the input / output shifter 53, and the PID controller 54. FIG. 9 is a diagram illustrating the effect of the characteristic identifier 32 according to the first embodiment. In FIG. 9, the dashed line indicates the dynamic characteristics identified without considering dead time using a recursive characteristic identifier that does not include the dead time estimator 52, the input / output shifter 53, and the PID controller 54, and the solid line indicates the true dynamic characteristics. As shown in FIG. 9, when identification is performed using a low-order model without considering dead time, dynamic characteristics different from the true dynamic characteristics are identified. If a controller is designed using dynamic characteristics different from the true dynamic characteristics, the tracking ability of the controller may deteriorate, potentially making it impossible to maintain system stability. On the other hand, when identification is performed using a high-order model to consider dead time, the calculation time and required memory amount increase, and high-order models are difficult to use for controller design. Furthermore, because the dead time varies depending on the piping length and the flow rate, even if the dead time is determined to a certain value in advance and identification is performed taking this into consideration, the identification result will differ from the true dynamic characteristics, as shown by the dashed line in Fig. 9. The characteristic identification unit 32 of the first embodiment shifts the post-shift input / output information SSAI so that the dead time information DEI becomes 0 or 1, thereby solving these problems.

[0062] <Feedforward Control Unit> Fig. 10 is a block diagram showing the feedforward control unit 33a according to embodiment 1. As shown in Fig. 10, the feedforward control unit 33a includes a priority determiner 71, an identification information selector 72, a ratio calculator 73, an interference amount calculator 74, and an adjuster 75.

[0063] The priority determiner 71 outputs priority information PSI indicating a priority actuator, which is an actuator for adjusting a target controlled variable, based on the sensor group information SI and the drive amount information AI. The target controlled variable may be selected, for example, from among the multiple controlled variables indicated by the sensor group information SI, one that exceeds a predetermined threshold. Alternatively, a user or an installer of the refrigeration cycle apparatus 1 may pre-select a target controlled variable from among the multiple controlled variables based on the specifications of the refrigeration cycle apparatus 1. The priority actuator is, for example, the actuator that has the greatest influence on the target controlled variable. For example, if the air conditioning capacity (also referred to as cooling capacity or heating capacity) is the target controlled variable, the compressor 2 has a large influence. In this case, the motor of the compressor 2 is designated as the priority actuator. Note that actuators other than the priority actuator are referred to as non-priority actuators.

[0064] The identification information selector 72 selects and outputs, as selected identification information PIDI, identification information IDI indicating the dynamic characteristics of the priority actuator and target controlled variable indicated by the priority information PSI from a plurality of pieces of identification information IDI corresponding to a plurality of dynamic characteristics.

[0065] The ratio calculator 73 calculates ratio information RI from the ratio between a DC gain related to the drive amount of the priority actuator based on the selection identification information PIDI and a DC gain related to the drive amount of the non-priority actuator based on the identification information IDI for the non-priority actuator and the target controlled variable. The ratio calculator 73 determines the drive amount of the non-priority actuator that offsets the fluctuation in the controlled variable when the drive amount of the priority actuator moves by one unit, and the ratio information RI indicates the ratio between the drive amount of the priority actuator and the drive amount of the non-priority actuator at that time. Note that the ratio information RI may also indicate the ratio between the priority actuator and multiple non-priority actuators.

[0066] The interference amount calculator 74 calculates and outputs an interference drive amount DC based on the ratio information RI and the priority information PSI. Specifically, the interference amount calculator 74 calculates, as the interference drive amount DC, a drive amount required for the non-priority actuator to offset the change in the control amount (hereinafter sometimes referred to as the interference amount) from the change in the drive amount of the priority actuator included in the drive amount information AI. The change in the drive amount is the difference between the drive amount after the acquisition period has elapsed (≒ the most recently input drive amount) and the drive amount before the acquisition period has elapsed (≒ the drive amount input immediately before the most recently input drive amount).

[0067] The adjuster 75 determines which actuator will compensate for the interference drive amount DC and outputs the result as a feedforward command FF. The adjuster 75 may cancel out the interference amount using only one non-priority actuator, or may allocate the interference drive amount DC to multiple non-priority actuators to cancel out the interference amount. The non-priority actuators that compensate for the interference drive amount DC to cancel out the interference amount are referred to as cancellation actuators. Furthermore, the adjuster 75 may employ a method of adjusting the control command so that the interference amount is canceled out over time to avoid unexpected behavior. The allocation method may be selected based on whether the difference between the target control amount included in the sensor group information SI and a predetermined limit value is less than a predetermined reference value. Furthermore, only actuators whose drive amounts included in the drive amount information AI do not violate the drive amount limit may be used to cancel out the interference amount.

[0068] The operation of the control device 12 will be described using FIG. 11 . FIG. 11 is a flowchart showing the operation of the control device 12 according to the first embodiment. First, when the sensor group information SI is acquired (step S1), the feedback control unit 31 outputs a feedback command FB based on the sensor group information SI (step S2). After the sensor group information SI and the drive amount information AI are acquired (step S3), the characteristic identification unit 32 identifies dynamic characteristics based on the sensor group information SI and the drive amount information AI and outputs identification information IDI (step S4). Next, the characteristic identification unit 32 estimates dead time based on the identification information IDI and outputs the dead time information DEI (step S5). At this time, the characteristic identification unit 32 feeds back the dead time estimated from the identification information IDI to the next dynamic characteristic identification process. Furthermore, the feedforward control unit 33a outputs a feedforward command FF based on the sensor group information SI, drive amount information AI, and identification information IDI (step S6). The adder 34 then adds the feedback command FB and the feedforward command FF together and outputs the result as a drive command AR to the actuator group 21 (step S7). The control device 12 performs the process of acquiring the sensor group information SI and the drive amount information AI (steps S1 and S3) at each acquisition cycle, and repeatedly performs the processes of steps S2 and S4 to S7 each time the sensor group information SI and the drive amount information AI are acquired.

[0069] As described above, the control device 12 of the refrigeration cycle apparatus 1 is a device that receives the sensor group information SI from the sensor group 22 and the drive amount information AI from the actuator group 21 as inputs. The control device 12 includes a feedback control unit 31 that outputs a feedback command FB, a characteristic identification unit 32 that outputs identification information IDI, a feedforward control unit 33a that outputs a feedforward command FF, and an adder 34 that outputs a drive command AR. The characteristic identification unit 32 sequentially identifies dynamic characteristics for each acquisition period based on the sensor group information SI and the drive amount information AI and calculates the identification information IDI. Furthermore, the calculated identification information IDI is used to estimate dead time in the input / output information SAI, and feedback control is performed to avoid dead time, thereby shifting the input / output information SAI to be used for identifying the next dynamic characteristic.

[0070] This makes it possible to suppress interference between actuators at any operating point without selecting representative operating points in advance and identifying the dynamic characteristics of those operating points. Therefore, the refrigeration cycle apparatus 1 can stabilize the operation of the actuators.

[0071] Furthermore, since interference between multiple actuators can be suppressed, the control gain of the feedback control unit 31 for each actuator can be increased. This allows for faster and more stable control of the refrigeration cycle apparatus 1. Furthermore, because the dynamic characteristics are identified sequentially, faster and more stable refrigeration cycle control can be achieved even when the indoor and outdoor environments or the number of operating indoor units 11 in a multi-model system change. Furthermore, although the piping length may vary depending on the installation location of the refrigeration cycle apparatus 1, interference between the actuators can be suppressed in this case as well by sequentially identifying the dynamic characteristics.

[0072] Embodiment 2. Fig. 12 is a block diagram showing a control device 12 according to embodiment 2. As shown in Fig. 12, embodiment 2 differs from embodiment 1 in that the feedforward command FF input to the adder 34 is output by a feedforward control unit 33b. In embodiment 2, the same parts as those in embodiment 1 are denoted by the same reference numerals and their description will be omitted, and the description will focus on the differences from embodiment 1.

[0073] <Feedforward Control Unit> Fig. 13 is a block diagram showing a feedforward control unit 33b according to embodiment 2. As shown in Fig. 13, the feedforward control unit 33b includes a sensor information processor 81, a target value generator 82, a sensor information selector 83, a subtractor 84, a DC gain selector 85, a divider 86, and a high-response adjuster 87.

[0074] The sensor information processor 81 performs the same processing as the sensor information processor 41 described in the first embodiment, and outputs processed sensor group information FSI2.

[0075] The target value generator 82 calculates a target value corresponding to the target controlled variable based on the processed sensor group information FSI2 and outputs it as control target value information TM2. The target controlled variable may be selected from, for example, a plurality of controlled variables indicated by the sensor group information SI that exceed a predetermined threshold. Alternatively, the user or an installer of the refrigeration cycle apparatus 1 may select the target controlled variable in advance from a plurality of controlled variables based on the specifications of the refrigeration cycle apparatus 1, etc.

[0076] The sensor information selector 83 selects information indicating the detection result for the target controlled variable from the processed sensor group information FSI2 and outputs it as controlled variable measurement value information TD2.

[0077] The subtractor 84 calculates the difference between the controlled variable measurement value information TD2 and the control target value information TM2, that is, the target error, and outputs it as control deviation information REI.

[0078] The DC gain selector 85 extracts a DC gain from the identification information IDI indicating dynamic characteristics of the target controlled variable as an output based on the controlled variable measurement value information TD2 and the identification information IDI. The DC gain selector 85 outputs information indicating the DC gain for the extracted target controlled variable as DC gain information FDCI.

[0079] The divider 86 divides the control deviation information REI by the DC gain information FDCI to calculate the increment of the drive amount required to make the control deviation zero, and outputs the result as control command information FF0.

[0080] The high-response adjuster 87 allocates the increase in drive amount to each actuator and outputs it as a feedforward command FF. The high-response adjuster 87 may allocate to only one actuator to reduce the control deviation to zero, or may allocate to multiple actuators to reduce the control deviation to zero. Furthermore, in order to avoid sudden behavior, the adjuster 75 may adopt a method of adjusting the control command so that the control deviation is reduced to zero over time. Different allocation methods may be used depending on whether the control deviation is less than a predetermined reference value. Furthermore, the increase in drive amount may be allocated only to actuators whose manipulated variable indicated by the actuator drive command AR does not violate a predetermined manipulated variable limit.

[0081] According to the second embodiment, the actuator drive amount required to make the control deviation zero can be added to the control command, thereby realizing highly responsive target value tracking control.

[0082] Embodiment 3. Fig. 14 is a block diagram showing a control device 12 according to embodiment 3. As shown in Fig. 14, embodiment 3 differs from embodiments 1 and 2 in that the feedforward command FF input to the adder 34 is output by a feedforward control unit 33c. In embodiment 3, the same parts as those in embodiments 1 and 2 are denoted by the same reference numerals and description thereof will be omitted, and the description will focus on the differences from embodiments 1 and 2.

[0083] <Feedforward Control Unit> Fig. 15 is a block diagram showing a feedforward control unit 33c according to embodiment 3. As shown in Fig. 15, the feedforward control unit 33c includes a non-interference output unit 91, a high tracking output unit 92, and a control command distributor 93.

[0084] The non-interference output unit 91 is the same as the feedforward control unit 33a described in the first embodiment. The high tracking output unit 92 is the same as the feedforward control unit 33b described in the second embodiment.

[0085] The control command distributor 93 outputs a feedforward command FF based on the control command (non-interference control command) FFa from the non-interference output unit 91, the control command (high-following control command) FFb from the high-following output unit 92, and the sensor group information SI.

[0086] The control command distributor 93 outputs, as a feedforward command FF, one of a non-interference control command, a high tracking control command, or a command indicating a manipulated variable obtained by weighting the manipulated variable indicated by the non-interference control command and the manipulated variable indicated by the high tracking control command, depending on the control state. The control state indicates, for example, the relationship between the target controlled variable and the limit value, and the relationship between the drive amount and the drive amount limit.

[0087] Specifically, when the difference between the target controlled variable and the limit value is equal to or less than a predetermined controlled variable threshold range, or when the difference between the drive variable and the drive variable limit is equal to or less than a predetermined drive variable threshold range, the control command distributor 93 selects the control command group information FFa from the non-interference output unit 91 and outputs it as a feedforward command FF. The target controlled variable may be selected from, for example, a plurality of controlled variables indicated by the sensor group information SI that exceed a predetermined threshold. Alternatively, the target controlled variable may be set in advance based on the specifications of the refrigeration cycle apparatus 1, for example.

[0088] Conversely, if the difference between the target control amount and the limit value exceeds the control amount threshold range, or if the difference between the drive amount and the drive amount limit exceeds the drive amount threshold range, the control command allocator 93 selects the control command group information FFb from the high tracking output unit 92 and outputs it as the feedforward command FF. Furthermore, if the difference between the target control amount and the limit value is within the control amount threshold range and the difference between the drive amount and the drive amount limit is within the drive amount threshold range, the control command allocator 93 can output, as the feedforward command FF, a command indicating a manipulated variable that is a weighted average of the manipulated variable indicated by the non-interference control command and the manipulated variable indicated by the high tracking control command.

[0089] According to the third embodiment, it is possible to select whether to perform the more stable tracking control described in the first embodiment or the more responsive tracking control described in the second embodiment, depending on whether the target controlled variable or the drive variable has a margin relative to the limit value.

[0090] REFRIGERATION CYCLE DEVICE, 2 COMPRESSOR, 3 FOUR-WAY VALVE, 4 OUTDOOR HEAT EXCHANGER, 5 OUTDOOR FAN, 6 INDOOR HEAT EXCHANGER, 7 INDOOR FAN, 8 EXPANSION VALVE, 10 OUTDOOR UNIT, 11 INDOOR UNIT, 12 CONTROL DEVICE, 13 PIPING, 14 FIRST SENSOR, 15 SECOND SENSOR, 16 THIRD SENSOR, 17 FOURTH SENSOR, 18 FIFTH SENSOR, 21 ACTUATOR GROUP, 22 SENSOR GROUP, 23 PROCESSING DEVICE, 24 MEMORY DEVICE, 31 FEEDBACK CONTROL SECTION, 32 CHARACTERISTICS IDENTIFICATION SECTION, 33a, 33b, 33c FEEDFORWARD CONTROL SECTION, 34 ADDER, 41 SENSOR INFORMATION PROCESSOR, 42 DESIGN VALUE GENERATOR, 43 SENSOR INFORMATION SELECTOR, 44 FEEDBACK CONTROLLER, 45 CONTROL COMMAND SELECTOR, 51 SEQUENTIAL CHARACTERISTICS IDENTIFICATION SECTION, 52 DEAD TIME ESTIMATOR, 53 INPUT / OUTPUT SHIFT SECTION, 54 PID CONTROLLER, 55 SEQUENTIAL CHARACTERISTICS IDENTIFICATION SECTION, 56 Information processor, 57 Input / output shift amount estimator, 58 Input / output shifter, 71 Priority determiner, 72 Identification information selector, 73 Ratio calculator, 74 Interference amount calculator, 75 Adjuster, 81 Sensor information processor, 82 Target value generator, 83 Sensor information selector, 84 Subtractor, 85 DC gain selector, 86 Divider, 87 High response adjuster, 91 Non-interference output unit, 92 High tracking output unit, 93 Control command distributor.

Claims

1. A refrigeration cycle apparatus comprising: a plurality of sensors for detecting controlled variables; a plurality of actuators; and a control device that outputs drive commands indicating manipulated variables of the plurality of actuators using sensor group information indicating detection results of the plurality of sensors and drive amount information indicating drive amounts of the plurality of actuators, wherein the control device has: a feedback control unit that outputs feedback commands for feedback control based on the sensor group information; a characteristic identification unit that sequentially identifies dynamic characteristics indicating relationships between the manipulated variables and the controlled variables for a plurality of combinations of the plurality of actuators and the plurality of controlled variables based on the sensor group information and the drive amount information, and outputs identification information indicating the identified dynamic characteristics; a feedforward control unit that outputs feedforward commands for feedforward control based on the identification information and the sensor group information; and an adder that adds the feedback command and the feedforward command and outputs the drive command, wherein the characteristic identification unit feeds back dead time of the dynamic characteristics estimated from the identification information to a next identification process of the dynamic characteristics.

2. The refrigeration cycle device of claim 1, wherein the characteristic identification unit comprises: a PID controller that outputs input / output dead time information for eliminating the dead time; an input / output shift unit that outputs shifted input / output information obtained by delaying the sensor group information and the drive amount information based on the input / output dead time information; and a sequential characteristic identifier that sequentially identifies the dynamic characteristics based on the shifted input / output information.

3. The refrigeration cycle device according to claim 2, wherein the input / output shift unit comprises: an information processor that applies at least one of decimation, low-pass filtering, high-pass filtering, and band-pass filtering to the sensor group information and the drive amount information and outputs the result as input / output information; an input / output shift amount estimator that outputs input / output shift amount information for delaying the input / output information based on the input / output dead time information; and an input / output shifter that delays the input / output information based on the input / output shift amount information.

4. A refrigeration cycle device according to any one of claims 1 to 3, wherein the feedforward control unit outputs the feedforward command indicating the operation amount of a priority actuator that has a greater influence on the selected controlled variable than other actuators, and the operation amount of a cancellation actuator that cancels out fluctuations in the selected controlled variable caused by operation of the priority actuator.

5. A refrigeration cycle device according to any one of claims 1 to 3, wherein the feedforward control unit outputs the feedforward command indicating the operation amounts of the plurality of actuators that makes the difference between the selected controlled variable and the target value zero.

6. The refrigeration cycle apparatus according to any one of claims 1 to 3, wherein the feedforward control section comprises: a non-interference output section that outputs a non-interference control command indicating an operation amount of a priority actuator that has a larger influence on the selected controlled variable than other actuators, and an operation amount of a cancellation actuator that cancels out a fluctuation in the selected controlled variable caused by the operation of the priority actuator; a high-following output section that outputs high-following control commands indicating operation amounts of the plurality of actuators that make the difference between the selected controlled variable and a target value zero; and a control command distributor that outputs, as the feedforward command, either the non-interference control command, the high-following control command, or a command indicating an operation amount that is a weighted average of the operation amount indicated by the non-interference control command and the operation amount indicated by the high-following control command, depending on the control state.

7. A control method for a refrigeration cycle device comprising: a plurality of sensors for detecting controlled variables; a plurality of actuators; and a control device that outputs drive commands indicating operation variables of the plurality of actuators using sensor group information indicating detection results of the plurality of sensors and drive amount information indicating drive amounts of the plurality of actuators, wherein the control device has: a feedback control step of outputting feedback commands for feedback control based on the sensor group information; a characteristic identification step of sequentially identifying dynamic characteristics indicating relationships between the operation variables and the controlled variables for a plurality of combinations of the plurality of actuators and the plurality of controlled variables based on the sensor group information and the drive amount information, and outputting identification information indicating the identified dynamic characteristics; a feedforward control step of outputting feedforward commands for feedforward control based on the identification information and the sensor group information; and an addition step of adding the feedback command and the feedforward command to output the drive command, wherein in the characteristic identification step, dead time of the dynamic characteristics estimated from the identification information is fed back to the next characteristic identification step.