JP6950843B2 - Compressor, air conditioner, refrigerator, compressor control method, compressor control learning data creation method and compressor control trained model creation method - Google Patents

Description

The present invention relates to a compressor driven by a motor, an air conditioner including a compressor, a refrigerator including a compressor, and a compressor control method for controlling the compressor.

Most of the failures of conventional compressors are caused by seizure or abnormal wear in the plain bearings. In the conventional compressor, for example, the refrigerant compressor is provided with a viscosity sensor in order to prevent seizure damage of the refrigerant compressor bearing due to a decrease in the lubricating oil viscosity, and the viscosity of the refrigerant compressor oil reservoir is measured by this sensor. In some cases, the supply of refrigerant to the refrigerant compressor is stopped when the viscosity falls below a predetermined value. This technique temporarily reduces the load on the compressor by stopping the supply of the refrigerant to prevent seizure of the compressor (see, for example, Patent Document 1).

JP-A-2002-206486

In a conventional compressor, the viscosity of the oil sump of the compressor is measured, but it is difficult to completely judge the contact or seizure of the shaft and the bearing from the viscosity of the oil sump. Even if the viscosity is equal to or higher than the predetermined viscosity, the shaft may come into contact with the bearing and seizure or abnormal wear may occur depending on the operating conditions. In addition, since the refrigerant is not supplied, the function as a compressor is stopped, and there is a problem that the efficiency is lowered.

In order to solve the above problems, the purpose is to avoid failure in the bearing portion in a wide range of operating conditions.

The invention according to one claim of the present disclosure is provided on a sliding bearing or a rotating shaft portion into which the sliding bearing is fitted , measures the displacement of the rotating shaft, measures the temperature of the sliding bearing, or the sliding bearing. A sensor that measures the oil film pressure and outputs the measured value as an output signal, and an output signal output from the sensor are calculated and processed to obtain the number of rotations at which the motor should rotate and transmitted as a control signal to the inverter. Obtained as a result of controlling the signal processing unit that controls the rotation speed of the motor, the measured value measured by the sensor, the rotation speed that the motor should rotate with respect to the measured value, and the rotation speed. A storage unit that associates the measured values after control, which are the measured values of the bearings, with each other and stores them as related information, and a machine learning unit that machine-learns the related information stored in the storage unit and outputs a learned model. To prepare.

According to one embodiment of the present disclosure, failure in the bearing portion can be avoided in a wide range of operating conditions. Further, according to another embodiment of the present disclosure, it is possible to avoid contact between the rotating shaft of the compressor and the bearing portion, seizure and abnormal wear due to this contact, without stopping the operation of the compressor. can.

It is a figure which shows the structure of the compressor which shows Embodiment 1 of this disclosure. It is a figure which shows the flow of control of the compressor which shows Embodiment 1 of this disclosure. This is an example of a flow for controlling the rotation speed of the compressor according to the first embodiment of the present disclosure. It is a schematic diagram which shows the eccentric state of the rotating shaft in the slide bearing of the compressor which shows Embodiment 1 of this disclosure. It is a figure which shows the structure which provided the displacement sensor in the rotary shaft part fitted with the slide bearing of the compressor which shows Embodiment 1 of this disclosure. It is a figure which shows the structure of the compressor which shows Embodiment 2 of this disclosure. It is a figure of the configuration example of the hardware of the computer which controls a compressor which shows Embodiment 2 of this disclosure.

Embodiment 1.
FIG. 1 is a vertical cross-sectional view showing a refrigerant compressor according to the first embodiment of the present invention. Here, a refrigerant compressor will be described as an example of the compressor, but the compressor to which this embodiment can be applied is not limited to the refrigerant compressor, and is a compressor having a general sliding bearing such as a compressor or a blower. If it is good. In particular, the compressor has a rotation mechanism that rotates around a rotation axis supported by a slide bearing, and this rotation mechanism can be applied to a compressor that exerts a compression function. Furthermore, it is suitable for a compressor in which an eccentric load is applied to the rotating shaft of the rotating mechanism.

FIG. 1 is a diagram showing a configuration of a compressor showing the present embodiment. In the figure, the left side shows a cross-sectional view of a twin rotary compressor which is one of the refrigerant compressors, and the right side shows a block showing a configuration for controlling this compressor. Although the twin rotary compressor is shown here, it may be a single rotary compressor having one compression mechanism, or a compressor supported by bearings such as a scroll compressor and a screw compressor. The invention can be applied.

In the figure, the refrigerant compressor 6 includes a rotating shaft 1, a plain bearing 2 that supports the rotating shaft 1, an electric motor 3 that rotates the rotating shaft 1, and an inverter 4 that controls the electric motor 3. The electric motor 3 includes a stator 3a and a rotor 3b provided inside the stator 3a and fixed to the rotating shaft 1. The electric motor 3 rotates around the central axis of the rotating shaft 1 by the electromagnetic force generated between the stator 3a and the rotor 3b.

Further, the rotating shaft 1 is provided with a rolling piston having a shape eccentric with respect to the central axis of the rotating shaft 1. The rolling piston is arranged in a cylinder provided so as to cover the rotating shaft 1 from the central axis in the outer peripheral direction. A compression chamber and a suction chamber are configured between the rolling piston and the cylinder. When the motor 3 rotates, the rolling piston fixed to the rotating shaft 1 rotates in the cylinder, and the space of the suction chamber connected to the suction port becomes smaller as the rolling piston rotates, and the internal medium is compressed. It becomes a compression chamber and is connected to the discharge port to discharge the internal medium.

The slide bearing 2 rotatably supports the rotating shaft 1 and can be arranged outside the axial direction of two rollings arranged side by side in the axial direction of the rotating shaft 1. In this case, the slide bearing 2 may form the end face of the compression chamber.

Further, the refrigerant compressor 6 is provided for the sliding bearing 2 or the portion of the rotating shaft 1 to which the sliding bearing 2 is fitted, and is a measured value obtained by measuring a physical quantity indicating the state of the refrigerant compressor 6 using a scientific principle. The signal processing unit 7 is a control unit that calculates and processes the sensor 5 that outputs Has. The signal processing unit may be regarded as a control unit.

The sensor 5 is not contacted or non-contacted, and the physical quantity measured by the sensor 5 indicates the state of the slide bearing 2 or the state of the refrigerant compressor 6 which can be measured by being provided on the portion of the rotating shaft 1 into which the slide bearing 2 is fitted. It may be a physical quantity. The physical quantities of the sensor 5 are, for example, vibration, temperature, and pressure.
The measurement principle of the sensor 5 uses a scientific principle, and an acceleration sensor for vibration, a thermocouple for temperature, and a diaphragm type pressure sensor for pressure can be considered.

More specifically, the sensor 5 can be a displacement sensor. The displacement sensor may be any sensor that can measure the displacement, distance, thickness, etc. of the measurement target. The measurement method of the displacement sensor is not limited to contact or non-contact, and may be any of eddy current type, capacitance type, laser type, ultrasonic type and the like. Further, the displacement sensor may be the displacement of the slide bearing 2, the relative displacement between the slide bearing 2 and the rotating shaft 1, or the change of the relative displacement.

Next, the operation will be described. When the electric motor 3 rotates the rotating shaft 1, the rotor 3b fixed to the rotating shaft 1 rotates. Then, the compressor 6 executes a cycle of sucking, compressing, and discharging the refrigerant between the rotor 3b and the cylinder, and fulfills the functions of compressing and discharging the refrigerant.

Here, the displacement sensor 5 that measures the bearing gap between the rotating shaft 1 of the rotor 3b of the compressor 6 and the inner surface of the slide bearing 2 outputs a signal such as a voltage according to the size of the bearing gap. .. A sensor signal representing the size of the bearing gap between the rotating shaft 1 and the inner surface of the sliding bearing 2 is transmitted to the signal processing unit 7, and the signal processing unit 7 transmits the received sensor signal to the size of the bearing gap. Convert to. The signal processing unit 7 outputs a control signal to the inverter 4 in consideration of the number of rotations of the electric motor 3 to be rotated according to the size of the bearing gap. The signal processing unit 7 may obtain a control signal to the inverter 4 according to the sensor signal and output this control signal to the inverter 4 without converting the above-mentioned sensor signal into the bearing gap.

When the signal processing unit 7 obtains the size of the bearing gap or the control signal according to the sensor signal, the following can be performed. It holds a threshold value that serves as a reference for the soundness of the measured value of the sensor 5 that is set in advance, and when the measured value of the sensor 5 deviates from the sound sound range determined by the threshold value, the measured value of the sensor 5 becomes the sound range. With this as the goal, the rotation speed of the electric motor 3 is feedback-controlled.

The threshold value may be composed of two values, an upper limit threshold value and a lower limit threshold value, as a reference for the soundness of the measured value of the sensor 5. In this example, if the measured value of the sensor 5 is a value between the upper limit threshold value and the lower limit threshold value, it is within a healthy range indicating that the state of the compressor is sound, and a control signal for maintaining the rotation speed up to that point is used. Output to the inverter 4. When the measured value of the sensor 5 is out of the sound range, a signal for feedback-controlling the rotation speed of the motor 3 is output to the inverter 4 with the goal of keeping the measured value of the sensor 5 in the sound range.

In order to obtain a signal for feedback control of the rotation speed of the motor 3 with the goal of keeping the measurement value of the sensor 5 out of the sound range and the measurement value of the sensor 5 in the sound range, the rotation speed of the motor 3 is changed once. Then, the control method may be determined by looking at the measured value of the sensor 5 after that.

Specifically, the signal processing unit 7 transmits a control signal for increasing the rotation speed by a certain number (for example, a predetermined rotation speed per minute) to the inverter 4, and the sensor 5 after a predetermined time (specifically, 1 minute). If the measured value of is changing in the direction within the sound range, a control signal for continuously increasing the rotation speed of the motor 3 is output to the inverter 4. On the contrary, if the measured value of the sensor 5 after a predetermined time changes further out of the sound range, a control signal for lowering the rotation speed of the motor 3 is output to the inverter 4. This is repeated, and when the measured value of the sensor 5 after control is within the sound range, a control signal for returning the rotation speed of the motor 3 to the original value is output to the inverter 4 and maintained.

FIG. 2 shows a flowchart relating to the control method of the present embodiment. With reference to the figure, a method of controlling the rotation speed of the electric motor 3 by the inverter 4 using 6 steps is shown.

Step S0: In the first control step S0, a device including the refrigerant compressor 6, for example, a sensor air conditioner or a refrigerating device is activated. The electric motor 3 of the refrigerant compressor 6 rotates, the rotating shaft 1 and the rotor 3b fixed to the rotating shaft 1 rotate, and the compressor 6 is sucked, compressed, and discharged.

Step S1: In the first control step S1, the sensor 5 starts measuring the state of the slide bearing 2 while the compressor 6 is sucking, compressing, and discharging. Specifically, the displacement sensor 5 measures the gap between the slide bearing 2 and the rotating shaft 1. The value measured by the sensor 5 depends on the physical quantity measured by the sensor. The physical quantity to be measured depends on the type of the sensor 5.

Step S2: In the second control step S2, the measured value measured by the sensor 5 is compared with the threshold value as a reference for the soundness of the preset measured value, and it is determined whether or not the measurement value is within the soundness range. For example, a case where the threshold value of the upper limit of the bearing gap judged to be sound is set in advance will be described. If the measured value measured by the sensor 5 or the value obtained by converting the measured value into the size of the bearing gap is less than the threshold value, the process proceeds to step S3.

Step S3: Third control Step S3 confirms the presence or absence of a stop command for the refrigerant compressor 6, the air conditioner, or the refrigerating equipment. If there is a stop command, the refrigerant compressor 6, air conditioner or refrigeration equipment is stopped. If there is no stop command, the process proceeds to step S4.

Step S4: In the fourth control step S4, the signal processing unit 7 outputs a control signal of the rotation speed of the motor 3 preset according to the operation settings of the refrigerant compressor 6, the air conditioner or the refrigerator to the inverter 4. Then, after shifting to step S6 described later, the process returns to step S1. The control of the above steps S1 to S6 to S1 is a basic operation when the state of the refrigerant compressor 6 is sound.

Next, in step S2, a case where the measured value of the sensor 5 or the size of the bearing gap is equal to or larger than the threshold value will be described. If the measured value of the sensor 5 or the size of the bearing gap is equal to or larger than the threshold value, the process proceeds to step S5.

Step S5: In the fifth control step S5, the signal processing unit 7 generates a control signal (command) for changing the rotation speed of the electric motor 3 and outputs the control signal (command) to the inverter 4. The inverter that receives this control signal controls the motor 3 to change the rotation speed of the motor 3 and proceeds to step S6.

Step S6: In the sixth control step S6, the signal processing unit 7 outputs a control signal (command) for maintaining the rotation speed of the electric motor 3 to the inverter 4. Upon receiving this control signal, the inverter 4 controls the motor 3 to maintain the rotation speed of the motor 3 and returns to step S1.

The process returns from step S6 to step S1 and proceeds to step S2. At this time, the second control step S2 compares the measured value measured by the sensor 5 with the threshold value that serves as a reference for the soundness of the measured value set in advance, and determines whether or not it is within the soundness range. As a result of the determination, if it is within the sound range, that is, if the sensor measurement value measured by the sensor 5 or the value obtained by converting the measurement value into the size of the bearing gap is less than the threshold value, the process proceeds to steps S3 and S6. As a result, the control for maintaining the rotation speed of the electric motor 3 is performed. On the other hand, as a result of the judgment, if it is out of the sound range, that is, if the sensor measurement value measured by the sensor 5 or the value obtained by converting the measured value into the size of the bearing gap remains equal to or more than the threshold value, the bearing gap is within the sound range. That is, the steps are executed in the order of steps S1, S2, S5, S6, S1 until the value becomes less than the threshold value.

The above is the case where the preset threshold value in step S2 is the upper limit threshold value of the bearing gap, but when it is the lower limit threshold value of the bearing gap, it is as follows. In step S2, if the bearing gap exceeds the threshold value, the process proceeds to step S3, and if it is equal to or less than the threshold value, the process proceeds to step S5. The above steps are continued until a stop command is given for the refrigerant compressor 6, air conditioner or refrigeration equipment.

In the above, the case where the threshold value is the upper limit threshold value and the lower limit threshold value of the bearing gap has been described separately, but the threshold value is the upper limit threshold value and the lower limit threshold value, and the sound range is the upper limit threshold value and the lower limit threshold value. It may be between. In this case, when the signal processing unit 7 determines in step S2 whether or not it is within the sound range, the measured value measured by the sensor 5 or the size of the bearing gap exceeds the lower threshold value and is less than the upper limit threshold value. If is, it is judged to be within the healthy range, and other than this is excluded from the healthy range.

Further, when it is out of the sound range, the control signal for changing the rotation speed of the motor 3 generated by the signal processing unit 7 in step S5 is the case where the measured value or the size of the bearing gap is equal to or less than the lower limit threshold value. In the case of the upper limit threshold value or more, the control signal is usually different, and whether the rotation speed of the motor 3 is increased or decreased is reversed.

FIG. 3 shows a flowchart relating to another control method according to the present embodiment. In the figure, A to E at the end are names for distinguishing the scenes of the control step, and the steps having the same step S ○ before the end have the same internal processing.

When it is determined that the state of the refrigerant compressor 6, the air conditioner or the refrigerating equipment is within the sound range, steps S1A, S2A, S3 and S4A having the same contents as the above-mentioned control steps S1, S2, S3, S4 and S1. , S1A are executed sequentially.

Next, a case where the control shifts to a process of changing the rotation speed of the motor 3, that is, a case where the measured value of the sensor 5 or the size of the bearing gap is determined to be out of the sound range in step S2 will be described. .. First, a case where the threshold value is set as the upper limit of the bearing gap will be described.

Step S1A: Corresponds to the above step S1. Similar to the above, the displacement sensor 5 measures the gap between the slide bearing 2 and the rotating shaft 1 while the compressor 6 is sucking, compressing, and discharging.

Step S2A: Corresponding to the above step S2, when the measured value measured by the sensor 5 or the size of the bearing gap is equal to or larger than the threshold value, the signal processing unit 7 first “increases” the motor rotation speed as out of the sound range. After shifting to step S5-1 to generate a control signal to be generated and output to the inverter 4, the process proceeds to step S1B. If the measured value measured by the sensor 5 or the size of the bearing gap is less than the threshold value, the process proceeds to step S3 as being within the sound range.

Step S5-1: The signal processing unit 7 generates a control signal that "increases" the rotation speed of the electric motor 3 and outputs it to the inverter 4 to "increase" the rotation speed of the electric motor 3.

Step S1B: The process of step S1 is performed again. In particular, the sensor 5 measures the state of the sliding bearing and proceeds to step S7A.

Step S7A: In step S7A, the latest measured sensor measurement value or bearing clearance size is compared with the previously measured sensor measurement value or bearing clearance size. Here, a case where the signal processing unit 7 determines whether the latest bearing gap is increased with respect to the previous bearing gap will be described.
1) If the latest bearing clearance is smaller than the previous bearing clearance, it means that the direction is better as a result of the change. Therefore, go to step S6B for maintaining the rotation speed of the motor 3. After migrating, the process is transferred to step S2A.
2) If the latest bearing clearance increases or does not change from the previous bearing clearance, the result of the change will be worse or unchanged. Therefore, on the contrary, the rotation speed of the motor 3 is "decreased". The process is transferred to step S5-2.

Step S6B: Similar to step S6, the signal processing unit 7 generates a control signal for maintaining the rotation speed of the motor 3 and outputs the control signal to the inverter 4 to maintain the rotation speed of the motor 3.

Step S5-2: The signal processing unit 7 generates a control signal that “decreases” the rotation speed of the electric motor 3 and outputs it to the inverter 4 to “decrease” the rotation speed of the electric motor 3. Next, the process proceeds to step S1C.

Step S1C: The process of step S1 is performed again. In particular, the sensor 5 measures the state of the slide bearing and proceeds to step S7B.

Step S7B: In step S7B, the latest measured sensor measurement value or the size of the bearing gap is compared with the previously measured sensor measurement value or the size of the bearing gap.
1) If the latest bearing gap is smaller than the previous bearing gap, it means that the direction is better as a result of the change, so the process proceeds to step S6C.
2) If the latest bearing clearance does not increase or change from the previous bearing clearance, it will not deteriorate or change as a result of the change. In this step, even if the rotation speed of the latest motor 3 is "decreased" or "increased", it does not deteriorate or change. In this case, the process proceeds to step S6E.

Step S6C: Similar to step S6, the signal processing unit 7 generates a control signal for maintaining the rotation speed of the motor 3, outputs the control signal to the inverter 4, maintains the rotation speed of the motor 3, and proceeds to step 2B. do.

Step S2B: In step S7B, as a result of "decreasing" the rotation speed of the motor 3, the direction is good. Therefore, when the measured value measured by the sensor 5 or the size of the bearing gap is equal to or larger than the threshold value. The signal processing unit 7 shifts to step S5-2 to generate a control signal for "decreasing" the motor rotation speed and output it to the inverter 4. If the measured value or the size of the bearing gap is less than the threshold value, it is within the sound range, and the rotation speed of the motor 3 is shifted to the continuation step S6D.

Step S6D: Similar to step S6, the signal processing unit 7 generates a control signal for maintaining the rotation speed of the motor 3, outputs the control signal to the inverter 4, maintains the rotation speed of the motor 3, and then proceeds to step S1A. Moving. This step is a process when it is within the sound range.

Step S6E: Similar to step S6, the signal processing unit 7 generates a control signal for maintaining the rotation speed of the motor 3, outputs the control signal to the inverter 4, maintains the rotation speed of the motor 3, and then proceeds to step S1A. Migrate and implement a series of flows. By this processing flow, the rotation speed of the motor 3 whose bearing gap is less than the threshold value is searched.

S3 shown in FIG. 3 indicates that when the measured value measured by the sensor 5 or the size of the bearing gap is less than the threshold value in step S2A, the process moves to step S3 as being within the sound range. In step S3, as in step S3 described above, the presence or absence of a stop command for the refrigerant compressor 6, the air conditioner or the refrigerating equipment is confirmed, and if there is a stop command, the refrigerant compressor 6, the air conditioner or the refrigerating equipment is confirmed. To stop. If there is no stop command, the process is transferred to steps S4 / S6A. S4 / S6A described in FIG. 3 means that the same processing as in step S4 and step S6 described above is performed. That is, in step S4, the signal processing unit 7 outputs a control signal of the rotation speed of the motor 3 preset according to the operation settings of the refrigerant compressor 6, the air conditioner or the refrigerator to the inverter 4, and in step S6A, The signal processing unit 7 outputs a control signal (command) for maintaining the rotation speed of the electric motor 3 to the inverter 4.

In the above control flow, the motor rotation speed increase command step S5-1 that gives a command to increase the rotation speed of the motor 3 and the motor rotation speed decrease command step S5-2 that commands to decrease the rotation speed of the motor 3 It doesn't matter which order is executed first. That is, in the above description, although the motor rotation speed increase command step S5-1 was performed first after step S2A, the motor rotation speed decrease command step S5-2 may be executed first. Further, when the execution order is reversed in this way, the steps to be executed in succession to both steps are shifted in the same manner as described above.

Further, it is conceivable to set the threshold value for the bearing gap as both the upper limit and the lower limit. In this case, in steps S2A and S2B, the rotation speed of the motor 3 may be controlled so that the measured bearing clearance is within the sound range within the upper and lower limits of the threshold value.

Further, in steps S7A and S7B, it may be determined whether or not the latest bearing clearance is reduced with respect to the previous physical quantity (previous bearing clearance). When determining whether the latest bearing clearance has decreased with respect to the previous physical quantity, the step of shifting may be changed depending on the determination result, depending on whether the threshold value is the upper limit or the lower limit of the bearing clearance. For example, when the threshold value is set to the lower limit, it is determined in step S7A whether the physical quantity is decreasing, and if it is decreasing, the process proceeds to step 5-2, and the signal processing unit 7 "decreases" the rotation speed of the electric motor 3. A control signal is generated and output to the inverter 4 to "decrease" the rotation speed of the electric motor 3. On the other hand, when the physical quantity has not decreased, the process is transferred to step S6B, and the signal processing unit 7 generates a control signal for maintaining the rotation speed of the motor 3 and outputs the control signal to the inverter 4 to rotate the motor 3. Keep the number. Further, in step S7B, it is determined whether the physical quantity is decreasing, and if the physical quantity is decreasing or not changing, the process proceeds to step S6E, and if the physical quantity is increasing, the process proceeds to step S6C.

That is, when the threshold value is the upper limit threshold value of the bearing gap and the measured bearing gap is equal to or more than the threshold value, the latest bearing gap is reduced in steps S7A and 7B so as to be less than the threshold value. The rotation speed of the electric motor 3 may be controlled. On the contrary, when the threshold value is the lower limit of the bearing clearance and the bearing clearance is equal to or less than the threshold value, in steps S7A and S7B, the latest bearing clearance is set so as not to decrease further. The rotation speed of the electric motor 3 may be controlled so as to exceed the above.

The signal processing unit 7 specifies a target rotation speed of the electric motor 3, and the inverter 4 controls the target rotation speed. The control of the rotation speed of the electric motor 3 performed by the inverter 4 is performed by a classical control theory such as achieving a desired behavior as an input / output system expressed by a transmission function, and fuzzy control controlled by a control model using a fuzzy set. In addition to the theory, a control method based on any control theory may be used.

Next, in the above-mentioned processing, the signal processing unit 7 changes the rotation speed of the electric motor 3 when it is determined that the measured value or the size of the bearing gap is out of the sound range. Specifically, in step S5, step S5-1, and step S5-2, a method of determining the amount of change in the rotation speed of how much the rotation speed of the electric motor 3 is changed when it is determined to be out of the sound range will be described. do.

Here, the amount of change in the motor rotation speed will be described as being defined as ΔN. The unit of the number of revolutions such as ΔN is the number of revolutions per unit time, for example, the number of revolutions per minute [RPM]. The simplest way to give the amount of change ΔN of the rotation speed of the motor 3 is to add or subtract a preset value from the current rotation speed of the motor 3. That is, it is as shown in the following equation 1.

N ＝ N0 ＋ a ・ ΔN ・ ・ ・ ・ (1)
However,
N: Motor speed after change, N0: Motor speed before change,
a: Coefficient in the range of −1 ≤ a ≤ 1, ΔN: Constant value that can be set arbitrarily

Here, in the above-mentioned steps (steps S5, S5-1, S5-2), the rotation speed N of the electric motor 3 can be changed discretely by sequentially changing the coefficient a in the equation 1.

Further, ΔN may be given as in the following equation 2 according to the motor rotation speed N0 before the change.

ΔN = a ・ N ₀ ... (2)

As in the above equation 2, if ΔN is obtained by multiplying the rotation speed N0 of the motor 3 before the change by the coefficient a, the rotation speed of the motor can be changed as in the following equation 3 without setting ΔN in advance. Can be done.

N = N ₀ + ΔN = (1 + a) ・ N ₀ ... (3)

Here, the coefficient a is first given an appropriate initial value (0.1, etc.), and in steps S5-1 and S5-2, the value of the coefficient a is changed according to the determination results in steps S7A and S7B. For example, if the number of rotations is increased and the bearing gap is directed toward the healthy range, the coefficient a can be increased from 0.1 to 0.2.

The method of changing the rotation speed of the motor 3 shown above is an example, and the present embodiment can be applied regardless of the method of changing the rotation speed of the motor. Further, with respect to the rotation speed of the electric motor 3 designated by the signal processing unit 7, the control method of the electric motor rotation speed by the inverter 4 is pulse width modulation, that is, PWM (Pulse Width Modulation) or pulse amplitude modulation, that is, PAM. In addition to (Pulse Amplifier Modulation), any control method can be applied to the present invention.

Next, the threshold value of the bearing gap (hereinafter, _{also referred to as Hlim} ) will be described. By attaching the displacement sensor 5 to the slide bearing 2, the bearing gap (distance between the rotating shaft 1 and the inner surface of the slide bearing 2) of the slide bearing 2 can be measured. By measuring the bearing clearance with the displacement sensor 5, it is possible to quantitatively evaluate the contact between the rotating shaft 1 and the bearing (sliding bearing 2), the margin of contact even if not contacted, and the like.

FIG. 4 is a schematic view showing a cross section of the slide bearing 2 perpendicular to the rotation center axis of the rotation axis 1. The figure shows that the rotating shaft 1 is in an eccentric state. As shown in the figure, when a load (arrow in the figure) acts on the rotating shaft 1, the rotating shaft 1 is eccentric in the load direction L1 (arrow in the figure) with respect to the bearing center C2 of the slide bearing 2. At this time, for example, when the displacement sensor 5 is attached to the first quadrant and the second quadrant (the portion of the slide bearing 2 on the opposite side of the load direction L1) with respect to the bearing center line, the measured bearing clearance is , Increases according to the degree of eccentricity of the rotating shaft 1. In this case, the motor rotation speed may be controlled in the direction in which the bearing gap is reduced.

On the other hand, when the displacement sensor 5 is attached to the third quadrant and the fourth quadrant (the portion of the slide bearing 2 on the direction side of the load direction L1) with respect to the bearing center line, the bearing gap is reduced, so that the bearing gap is reduced. The motor rotation speed may be controlled so as to increase.

Further, depending on the lubrication state of the bearing, it is conceivable that the load direction L1 and the eccentric direction are opposite to each other, unlike the above. _{In this case, the threshold value Hlim} for the bearing gap according to the first embodiment is set for both the upper limit threshold value and the lower limit threshold value, and the motor rotation speed is set so that the bearing gap is within the range between the upper limit threshold value and the lower limit threshold value. Should be controlled. The range between the upper limit threshold and the lower limit threshold of the bearing gap is within the above-mentioned sound range.

FIG. 5 is a cross-sectional view of a refrigerant compressor in which the displacement sensor 5 is attached to a rotating shaft portion fitted to the slide bearing 2. This is done by embedding a displacement sensor in a portion of the rotating shaft 1 inside the slide bearing 2 toward the outside of the diameter method of the rotating shaft 1. As shown in the figure, by attaching the displacement sensor 5 to the rotating shaft portion fitted to the sliding bearing 2, the bearing clearance of the sliding bearing 2 can be measured from the rotating shaft 1 side over the entire circumference. That is, the lower limit threshold value and the upper limit threshold value are set for the minimum value and the maximum value of the bearing gap measured during one rotation of the rotation shaft 1, respectively, and the minimum value is below the lower limit threshold value or the maximum value is the upper limit threshold value. The rotation speed of the electric motor 3 may be controlled so as not to exceed the value. Also in this case, the range in which the measured minimum and maximum values of the bearing gap are within the range between the upper limit threshold value and the lower limit threshold value is within the above-mentioned sound range.

A method of setting the value of the threshold value _{Hlim of the bearing gap will be described.} For example, an arbitrary value or a ratio to the difference between the outer diameter of the rotating shaft and the inner diameter of the bearing can be set in advance. _{The value of the threshold value Hlim} of the bearing clearance may be set according to the size of the diameter clearance of the bearing of the slide bearing 2. For example, the following setting method of Equation 4 can be considered.

_Hlim = b ・ C ・ ・ ・ ・ (4)
However, b: a coefficient in the range of 0 <b <1, C: bearing diameter clearance = bearing inner diameter-shaft outer diameter

Here, the bearing diameter gap C may be a bearing radius gap. In this case, the range of the coefficient b is 0 <b <2. Further, instead of the bearing diameter clearance, the surface roughness of either the outer diameter of the rotating shaft 1 or the inner surface of the slide bearing 2 may be used. Alternatively, it may be the sum of the surface roughness of the outer diameter of the rotating shaft 1 and the surface roughness of the inner surface of the slide bearing 2, or the square root of the sum of squares. In either case, the value obtained by multiplying the value instead of the bearing diameter clearance by the coefficient b can be set as the value _{of the threshold value Hlim of the bearing clearance.} However, when surface roughness is used, the coefficient b may be a value of 0 or more, and there is no limitation of 1 or less.

Here, it will be described that the state of the slide bearing 2 such as the bearing gap is changed by changing the rotation speed of the motor 3. When the electric motor 3 rotates, the rotating shaft 1 and the rolling piston are integrally rotated around the rotation center axis of the rotating shaft 1, and the compression chamber between the cylinder and the rolling piston is sequentially deformed to form the inside of the compression chamber. The process of sucking, compressing, and discharging the medium is repeated. At this time, the plain bearings 2 that support the rotating shafts 1 on both sides of the rolling piston in the axial direction receive an eccentric load from the rotating shaft 1 depending on each step, that is, the position around the rotating shaft of the rolling piston. Here, when the rotation speed of the motor 3 changes, the magnitude of the oil film pressure generated in the bearing gap changes, so that the eccentric position of the rotating shaft changes, and inevitably the distribution of the bearing gap changes, and the plain bearing The state of 2 will change.

According to the present embodiment, the bearing clearance indicating the state of the refrigerant compressor 6 is measured by the displacement sensor 5, and the bearing clearance threshold having a margin against contact and seizure between the rotating shaft 1 and the slide bearing 2. By controlling the rotation speed of the motor 3 with the inverter 4 according to the measured bearing clearance and the set threshold value, the contact between the rotating shaft 1 and the slide bearing 2 can be made from the region where the load of the compressor is large to the region where the load is small. Burning can be avoided.

Further, the sensor 5 provided on the rotating shaft 1 to which the sliding bearing 2 or the sliding bearing 2 is fitted and outputs the measured measured value as an output signal and the output signal output from the sensor 5 are calculated and processed to rotate the motor. The control unit that obtains the desired rotation speed and transmits it as a control signal to the inverter 4 to control the rotation speed of the motor 3 avoids the failure of the bearing 2 in a wide range of operating conditions. Further, it is possible to avoid contact between the rotating shaft 1 of the compressor 6 and the bearing 2 and seizure and abnormal wear due to this contact without stopping the operation of the compressor 6.

(When using a vibration sensor)
Next, an example in which a vibration sensor is used as the sensor 5 provided for the portion of the slide bearing 2 or the rotating shaft 1 into which the slide bearing 2 is fitted will be described. The vibration sensor, which is the sensor 5, measures the vibration of the slide bearing 2 or the slide bearing 2. The vibration sensor here is not limited to the acceleration sensor, and can be applied regardless of the measurement principle, contact, or non-contact. When a vibration sensor is used as the sensor 5, the vibration sensor is not only provided on the portion of the rotating shaft 1 to which the slide bearing 2 or the slide bearing 2 is fitted, but also the slide bearing 2 or the slide bearing 2 is fitted. It may be provided in a compressor 6 that is not vibrationally insulated from the portion of the rotating shaft 1 or a housing that houses the compressor 6.

When the rotating shaft 1 and the sliding bearing 2 are separated by an oil film and operated in a non-contact state, the vibration generated by the rotation of the rotating shaft 1 is dampened by the viscosity of the lubricating oil. Therefore, the level of vibration propagating to the bearing 2 is smaller than that when the rotating shaft 1 and the bearing 2 are in contact with each other. Therefore, the sensor 5 of the vibration sensor can detect the contact between the rotating shaft 1 and the bearing 2 by measuring the vibration level (magnitude of vibration amplitude) or the acceleration of the rotating shaft 1 or the bearing 2. Further, when contact is detected, the rotation speed of the electric motor 3 can be changed to avoid seizure.

Next, the operation will be described. The basic control method is the same as in the case of the displacement sensor. However, regarding the vibration level or the threshold value of acceleration, it is preferable to set an upper limit value because the vibration of the shaft or the bearing becomes large when the contact occurs as described above. For the vibration level or the threshold value of acceleration, an arbitrary upper limit value a _lim may be set. In this case, the above-mentioned healthy range is a range in which the vibration level or acceleration is equal to or less than the _{upper limit value a lim.}

The signal processing unit 7 obtains a signal for feedback control of the rotation speed of the electric motor 3 with the target of keeping the measured value of the sensor 5 which is a vibration sensor within the sound range with the _{upper limit value a lim or less as the sound range.}

Specifically, when the measured value of the vibration sensor _{exceeds the upper limit value a lim} , the rotation speed of the motor 3 is changed, and the measured value of the vibration sensor, which is the sensor 5 after the rotation speed is changed, is sound. The control is changed depending on whether or not it is within the range. Here, changing the rotation speed means increasing or decreasing the rotation speed.

For example, when the measured value of the vibration sensor changes in a direction outside the sound range, that is, when the measured value of the vibration sensor becomes larger, the direction is opposite to the direction of change in the rotation speed of the motor 3 performed immediately before. Change it and see the measured value of the vibration sensor after that. When the measured value of the vibration sensor becomes small, it is heading in a good direction, so that the rotation speed of the motor 3 is maintained. _{Alternatively} , when the measured value of the vibration sensor is larger than the upper limit value a lim by a predetermined value or more, the rotation speed of the motor 3 is changed immediately before with the aim of further improving, that is, reducing the vibration. The vibration of the bearing 2 can be quickly focused within the sound range by further changing the direction.

If the measured value of the vibration sensor does not decrease even if the rotation speed of the motor 3 is changed in the direction opposite to that immediately before, the amount of change in the rotation speed is reduced and the rotation speed of the motor 3 is changed again. The same thing is repeated to search for the rotation speed at which the measured value of the vibration sensor becomes smaller after changing the rotation speed. When the rotation speed at which the measured value of the vibration sensor becomes small can be searched, the rotation speed is maintained or the rotation speed at which the measured value becomes smaller is searched, and the motor 3 is controlled to the rotation speed at which the state becomes good. By controlling in this way, the state in which abnormal vibration is generated is removed, and seizure of the slide bearing 2 is avoided without stopping the operation.

Here, it will be described that the vibration of the bearing changes when the rotation speed of the motor 3 is changed when the vibration of the bearing becomes large. When the rotating shaft and the bearing come into direct contact with each other, acceleration due to friction is generated as compared with the case of non-contact. As described above, when the rotation speed of the motor 3 is changed to bring it into a non-contact state, the vibration acceleration decreases. Furthermore, if the bearing clearance is very small even in the non-contact state, the magnitude of the oil film pressure in the bearing clearance is proportional to the reciprocal of the cube of the bearing clearance, so a very high pressure is generated and the eccentricity of the rotating shaft is generated. Since the position fluctuates, vibration acceleration is generated according to the amount of fluctuation. Therefore, it is possible to change the vibration of the bearing by changing the rotation speed of the motor 3, and changing the rotation speed of the motor 3 to keep the vibration of the bearing within the sound range causes the above bearing gap to be within the sound range. Equivalent to

According to the present embodiment, the vibration level or acceleration of the rotating shaft 1 or the bearing 2 generated when the rotating shaft 1 and the bearing 2 come into contact with each other or the distance between the rotating shaft 1 and the bearing 2 becomes small is vibrated. By controlling the rotation speed of the electric motor 3 with the inverter 4 so that the vibration level or acceleration measured by the sensor is lower than the vibration level or acceleration threshold measured by the sensor, seizure of the rotating shaft 1 and the slide bearing 2 is avoided. can do.

(When using a temperature sensor)
Next, an example in which a temperature sensor for measuring the bearing temperature is used as the sensor 5 provided for the portion of the slide bearing 2 or the rotating shaft 1 into which the slide bearing 2 is fitted will be described. The temperature sensor, which is the sensor 5, measures the temperature of the slide bearing 2 or the slide bearing 2. Here, the temperature sensor is not limited to a specific sensor such as a thermocouple, and can be applied regardless of the measurement principle, contact, or non-contact.

When seizure occurs between the rotary shaft 1 and the slide bearing 2, contact between the rotary shaft 1 and the slide bearing 2 occurs before seizure occurs. When contact occurs, friction occurs at the contact portion, most of the energy of the friction is consumed as heat energy, and the temperature of the slide bearing 2 and the peripheral portion connected to the slide bearing 2 rises. Therefore, by measuring the temperature inside the refrigerant compressor 6, for example, the bearing temperature, the contact between the rotating shaft 1 and the sliding bearing 2 is detected, and by appropriately controlling this detection as a trigger, the rotating shaft 1 and the sliding bearing 2 slide. It is possible to avoid seizure with the bearing 2.

Next, the operation will be described. The basic control method is the same as when the sensor 5 is a displacement sensor. However, it is preferable to set an upper limit value for the threshold value for the measured value of the sensor 5, which is a temperature sensor, because the temperature rises due to heat generation due to friction between the rotating shaft 1 and the slide bearing 2 as described above. The temperature threshold value of the slide bearing 2 may be set to an arbitrary upper limit value, Slim. Then, the healthy range becomes less than or equal to the upper limit value Slim.

The signal processing unit 7 feedback-controls the rotation speed of the electric motor 3 with the target of keeping the measured value of the sensor 5, which is a temperature sensor, within the sound range, with the upper limit value of Slim or less as the sound range.

Specifically, when the measured value of the temperature sensor is out of the sound range, that is, exceeds the upper limit value Slim, the rotation speed of the motor 3 is changed to change the rotation speed of the sensor 5 (temperature sensor). The control is changed depending on whether the measured value approaches the healthy range.

The control of the motor 3 when the sensor 5 is used as the temperature sensor and the sound range is within the upper limit value Slim is basically the same as when the above-mentioned vibration sensor is used as the sensor 5. That is, in the control of the vibration sensor, the temperature sensor _{can be controlled by reading the vibration sensor as the temperature sensor and reading the upper limit value a lim} as the upper limit value Slim.

Here, it will be described that when the temperature of the bearing rises, the temperature of the bearing changes when the rotation speed of the motor 3 is changed. When the rotating shaft and the bearing come into direct contact with each other, the temperature rises due to frictional shear between the materials, and the temperature of the bearing and the rotating shaft rises due to heat conduction in the material. As described above, when the rotation speed of the motor 3 is changed to bring it into a non-contact state, the temperature rise is suppressed by heat diffusion by the oil film. Further, if the bearing gap is very small even in the non-contact state, the fluid shear generated in the bearing oil film becomes large, the oil film temperature rises, and the temperatures of the bearing and the rotating shaft rise. Therefore, the temperature of the bearing can be changed by changing the rotation speed of the electric motor 3, and changing the rotation speed of the electric motor 3 to keep the temperature of the bearing and the rotating shaft within a sound range can reduce the above-mentioned bearing gap. It is equivalent to keeping it within the healthy range.

According to the present embodiment, the frictional heat generated by the contact between the rotating shaft 1 and the slide bearing 2 is measured by the temperature sensor (sensor 5), and the temperature threshold (upper limit value Slim) is measured by the temperature sensor. By controlling the rotation speed of the electric motor 3 with the inverter 4 so that the temperature drops, seizure of the rotating shaft 1 and the slide bearing 2 can be avoided.

(When using a pressure sensor)
Next, as the sensor 5, a refrigerant compressor having a pressure sensor for measuring the pressure generated in the oil film of the slide bearing 2 of the compressor 6 will be described. The pressure sensor, which is the sensor 5, is not limited to a specific sensor such as a diaphragm type pressure sensor, and may be any one that can measure the pressure generated by the oil film of the slide bearing 2 regardless of the measurement principle.

When seizure occurs between the rotary shaft 1 and the slide bearing 2, contact between the rotary shaft 1 and the slide bearing 2 occurs before seizure occurs. Generally, the magnitude of the pressure generated in the oil film of a plain bearing is proportional to the reciprocal of the cube of the bearing clearance. Therefore, immediately before the rotating shaft 1 and the inner surface of the slide bearing 2 come into contact with each other, the oil film becomes a thin film, and the inside of the oil film is in a very high pressure state. Therefore, the contact between the rotating shaft 1 and the slide bearing 2 can be detected by measuring the pressure generated in the oil film of the bearing of the compressor 6 using the pressure sensor. By appropriately controlling the rotation speed of the electric motor 3 using this detection as a trigger, seizure between the rotating shaft 1 and the slide bearing 2 can be avoided.

Next, the operation will be described. The basic control method is the same as in Forms 1 and 2. However, regarding the pressure threshold value, it is preferable to set an upper limit value because the oil film pressure becomes very high in the vicinity of the contact state as described above. For the pressure threshold measured by the pressure sensor, an arbitrary upper limit value Plim may be set. Then, in the sound range in this case, the measured value of the sensor 5 becomes the upper limit value Plim.

The signal processing unit 7 feedback-controls the rotation speed of the electric motor 3 with the target of keeping the measured value of the sensor 5, which is a pressure sensor, within the sound range, with the upper limit value Plim or less as the sound range.

Specifically, when the measured value of the pressure sensor is out of the sound range, that is, exceeds the upper limit value Plim, the rotation speed of the motor 3 is changed to change the rotation speed, and then the sensor 5 (pressure sensor). The control is changed depending on whether or not the measured value of is approaching the healthy range.

The control of the motor 3 when the sensor 5 is a pressure sensor and the sound range is within the upper limit value Plim is basically the same as when the above-mentioned vibration sensor is the sensor 5. That is, in the control of the vibration sensor, the pressure sensor _{can be controlled by reading the vibration sensor as a pressure sensor and reading the upper limit value a lim} as the upper limit value P lim.

Here, it will be described that the pressure of the bearing changes when the rotation speed of the motor 3 is changed when the pressure of the bearing rises. Since the magnitude of the pressure generated in the bearing oil film is proportional to the reciprocal of the cube of the bearing clearance, a very high pressure is generated, and the size of the bearing clearance changes as the rotation speed of the motor 3 changes, resulting in a pressure sensor. The measured value of is also changed. Therefore, the rotation speed of the motor 3 can be changed to change the pressure of the bearing, and changing the rotation speed of the motor 3 to keep the measured value of the oil film pressure within the sound range makes the above bearing gap sound. Corresponds to being within the range.

According to the present embodiment, the pressure generated in the oil film between the rotating shaft 1 and the slide bearing 2 is measured by the pressure sensor, and the pressure measured by the pressure sensor is lower than the pressure threshold (upper limit value Plim). By controlling the rotation speed of the electric motor 3 with the inverter 4 as described above, seizure of the rotating shaft 1 and the slide bearing 2 can be avoided.

As described above, the pressure of the oil film measured by the pressure sensor is proportional to the reciprocal of the cube of the bearing clearance, so that the pressure rises sharply as the bearing clearance becomes smaller. Therefore, it is easy to detect unhealthy, but it is resistant to noise and can be controlled efficiently.

In the above, the rotation speed of the motor 3 is controlled based on the measured value of one of the sensors of each type, but the rotation speed of the motor 3 is controlled based on the measured values of two or more sensors. Is also good. This is because the inside of the compressor may have high temperature and high pressure depending on the operating conditions, and acceleration may occur in other sliding elements. Therefore, by attaching a plurality of sensors to the bearing and controlling the rotation speed of the motor 3 based on the measured values, the state of the bearing and the rotating shaft can be comprehensively judged, and the contact state of the bearing and the rotating shaft can be detected. Increased accuracy. In addition, the sensor may have temperature characteristics, and by combining the temperature sensor and other sensors, it is possible to correct the temperature characteristics of the sensor to be used, and it is possible to measure with higher accuracy. ..

Next, the operation when the rotation speed of the electric motor 3 is controlled based on the measured values of the plurality of sensors will be described. The signal processing unit 7 compares the measured value of each sensor with the sound range set for each sensor, and changes the rotation speed of the motor 3 when even one sensor measured value is out of the sound range. Enter control. First, the rotation speed of the motor 3 is controlled with the goal of keeping the measured value of the sensor exceeding the sound range within the sound range. Further, if there is a sensor outside the sound range other than the measured value of the sensor that is out of the sound range first, the rotation speed of the electric motor 3 is controlled so that the sensor outside the sound range is within the sound range. do. This is repeated to control the rotation speed of the motor 3 until the measured values of all the sensors are within the sound range.

With the above configuration, a sensor having a short time constant, for example, a displacement sensor, can detect that it is out of the sound range and start control to keep it within the sound range. After that, by performing the above control until the measured value of a sensor having a long time constant, for example, a temperature sensor, is within the sound range, the state can be sufficiently sound and the state of the slide bearing 2 is stabilized. be able to.

Further, when a displacement sensor (vibration sensor, pressure sensor) having temperature characteristics is used, the temperature can be corrected by combining with the temperature sensor, and the accuracy of detecting the contact state of the bearing and the rotating shaft is improved. In this case, if the measured value of the temperature is fed back to the signal processing unit 7 and the temperature is corrected by the displacement sensor or the like, the accuracy is further improved.

Embodiment 2.
In the above embodiment, when the sensor measurement value measured by the sensor for the state of the bearing deviates from the preset sound range, the rotation speed of the motor is controlled with the goal of keeping the sensor measurement value within the sound range. However, the above control process may be machine-learned to obtain a trained model, and the trained model may be applied to the measured current sensor measurement values to estimate the number of rotations at which the motor should rotate. .. In this case, the trained model correlates the sensor measurement value, the rotation speed of the electric motor obtained by the signal processing unit with respect to the measurement value, and the sensor measurement value measured after controlling to this rotation speed. It can be stored as association information, and the stored association information can be obtained by machine learning. With this configuration, the number of revolutions that the motor should rotate to keep the sensor measurement value within the sound range from the learned model obtained from the past control history without exploratory control of the number of revolutions of the motor. A certain estimated rotation speed can be obtained and controlled.

That is, in the present embodiment, in the above embodiment, while controlling the rotation of the compressor based on the sensor information, the learning data is collected with the sensor information as the state and the control as the action, and the collected learning data is used as a machine. It learns, creates a trained model, and stores it in the learning model storage unit. When the control is tentatively obtained using the trained model and the probability of improvement of the state exceeds a predetermined value, the above control means transmits an operation signal to be controlled using the trained model to the device. You may switch.

In the following, an example in which a displacement sensor is used as the sensor will be mainly described, but other sensors that measure vibration, temperature, and pressure may be used as in the above embodiment.

FIG. 6 is a diagram showing a configuration of a compressor showing the present embodiment. In the figure, the left side is a cross-sectional view of the refrigerant compressor, and the right side is a block diagram showing a configuration for controlling the compressor. In the following, the same reference numerals as those in the above-described embodiment represent the same or corresponding ones. Further, as in the above embodiment, a single rotary compressor having a single compression mechanism may be used, or a compressor supported by bearings such as a scroll compressor and a screw compressor may be used.

In the figure, the refrigerant compressor 6 includes a rotating shaft 1, a slide bearing 2 that supports the rotating shaft 1, an electric motor 3, and an inverter 4. The electric motor 3 includes a stator 3a and a rotor 3b. The mechanical structure of the refrigerant compressor 6 is the same as that of the above embodiment.

Further, the refrigerant compressor 6 includes a sensor 5 provided for the sliding bearing 2 or the portion of the rotating shaft 1 to which the sliding bearing 2 is fitted, and calculates and processes the output signal output from the sensor 5, and the electric motor 3 operates. It has a signal processing unit 7 that sends a signal to send a control signal to the inverter 4 in order to obtain the number of rotations to be rotated.

The sensor 5 does not distinguish between contact and non-contact, and measures the displacement of the sliding bearing 2 or the relative displacement between the sliding bearing 2 and the rotating shaft 1 or the change in the relative displacement. It may be a displacement sensor to be measured, a temperature sensor, or a pressure sensor to measure the pressure of the oil film in the sliding bearing 2.

The refrigerant compressor 6 sends the information of the physical amount converted by the signal processing unit 7, the voltage signal of the sensor, the control signal to the inverter which is the result of obtaining the control amount based on the information, or the air conditioner or the refrigerating equipment. A storage unit 8 that associates the operation signals of Using the learning model storage unit 10 and the learned model to store the information of the above-mentioned physical quantity newly measured, the control signal from the voltage signal of the sensor to the inverter 4, or the operation signal to the air conditioner or the refrigerating device. A control amount calculation unit 11 for output is provided. Here, an example is shown in which the storage unit 8 stores the voltage signal of the sensor and the operation signal to the air conditioner or the refrigerating device in association with each other. The control amount calculation unit 11 may be regarded as an estimation unit.

Further, the learning data stored in the storage unit 8 is, in addition to the above, the result of obtaining the control amount based on the information, the result of transmitting the control signal to the inverter 4, the operation signal to the device, or the sensor having a change. Voltage signal may be associated.

The storage unit 8 may be a memory, a storage disk, or a semiconductor memory regardless of whether it is built in the refrigerant compressor 6 or mounted externally. Further, the storage medium and storage method of the storage unit 8 are not specified.

Further, the refrigerant compressor 6 of the present embodiment is provided with a storage unit 8, a machine learning unit 9, and a learning model storage unit 10, and a network or the like of the learned model stored in the learning model storage unit 10 is provided. It may be transmitted to the outside via.

Further, the refrigerant compressor 6 of the present embodiment may be configured to read the learned model from the outside into the learning model storage unit 10, and the control amount calculation unit 11 may be operated using the learned model. .. In this case, the refrigerant compressor 6 may not have the storage unit 8 and the machine learning unit 9. Further, the refrigerant compressor 6 may transmit learning data to the outside via a network or the like, and perform machine learning by a machine learning unit 9 provided outside to construct a learned model. In this case, even if the refrigerant compressor 6 reads the learned model into the learning model storage unit 10 from the outside via the network and operates the control amount calculation unit 11 using the learned model learned from the outside. good.

The storage unit 8 stores the slide bearing 2 of the compressor or the information of the sensor provided in the rotating portion where the slide bearing 2 is fitted as the input information indicating the state of the refrigerant compressor 6. The sensor 5 is a displacement sensor that measures the displacement of the rotating shaft 1, and may be information corresponding to the displacement of the rotating shaft 1 in the slide bearing 2. Further, the information stored in the storage unit 8 may be the input information indicating the state of the compressor, which is obtained by measuring the bearing gap of the slide bearing 2 from the above displacement. Alternatively, the sensor 5 may be a vibration sensor, a temperature sensor, or a pressure sensor that measures the oil film pressure of the slide bearing.

Further, the storage unit 8 uses the information of the physical quantity converted by the signal processing unit 7 as the input information indicating the action, the control signal to the inverter 4 which is the result of obtaining the control amount based on this information, or the air conditioner. Alternatively, the operation signal to the refrigeration equipment is stored. Further, at this time, the input information indicating the state and the input information indicating the behavior are associated and stored as learning data.

Further, the storage unit 8 stores as input information indicating a state in which the state has changed as a result of executing the action. These may be stored as time series information.

Next, the machine learning unit 9 learns a learning model to be an output by inputting a data set stored in the storage unit 8 based on the input information indicating the state and the input information indicating the action. The machine learning of the machine learning unit 9 may be reinforcement learning using a value function. In this case, the machine learning unit 9 is valuable when the measured value of the sensor 5 obtained by controlling the electric motor 3 with the rotation speed output by the signal processing unit 7 changes from outside the healthy range to within the healthy range. You may reward the function to build a trained model. Further, any learning algorithm may be used as the learning algorithm used by the machine learning unit 9. As an example, the case where reinforcement learning is applied will be described below.

Reinforcement learning is that an agent (behavior) in a certain environment observes the current state and decides the action to be taken. Agents are rewarded by the environment by choosing an action and learn how to get the most reward through a series of actions.

Q-learning and TD-learning are known as typical methods of reinforcement learning executed by the machine learning unit 9. For example, in the case of Q-learning, the general update equation (behavioral value table) of the behavioral value function Q (s, a) is expressed by the following equation.

Q (st, at) ← Q (st, at) + α (rt + 1 + γmaxQ (st + 1, a) --Q (st, at)) ・ ・ ・ (5)

In Equation 5, st represents the environment at time t, and at represents the action at time t. Depending on the action at, the environment changes to st + 1. rt + 1 represents the reward given by the change in the environment. Further, γ represents a discount rate, and α represents a learning coefficient. In addition, γ is in the range of 0 <γ ≦ 1, and α is in the range of 0 <α ≦ 1. When Q-learning is applied, the learning content that is the output becomes the action at.

Here, in the machine learning unit 9, st is input information indicating a state, at is input information indicating an action, and st + 1 is a state changed by the action at. rt + 1 is a reward given by changing st to st + 1.

More specifically, st is information on the slide bearing of the compressor or the sensor provided in the rotating portion where the slide bearing is fitted. Further, the action at is the information of the physical quantity converted by the signal processing unit, the control signal to the inverter 4 which is the result of obtaining the control amount based on the information, the operation signal to the air conditioner or the refrigerating equipment, or the compression. It may be information indicating the number of rotations of the machine.

In the update formula represented by the equation 5, if the action value of the best action a at time t + 1 is larger than the action value Q of the action a executed at time t, the action value Q is increased, and vice versa. , Reduce the action value Q. In other words, the action value function Q (s, a) is updated so that the action value Q of the action a at time t approaches the action value of discretion at time t + 1. By doing so, the best behavioral value in one environment is propagated to the behavioral value in the previous environment.

The machine learning unit 9 further includes a reward calculation unit and a function update unit.

The reward calculation unit calculates the reward based on the state variable. The reward calculation unit calculates the reward r based on the reward standard. For example, in the case of the reward increase standard, the reward r is increased (for example, the reward of "1" is given). On the other hand, in the case of the reward reduction standard, the reward r is reduced (for example, "-1" is given).

For example, assuming that the state is a bearing clearance between the bearing of the bearing 2 and the rotating shaft 1, the reward is large when the bearing clearance decreases, and the reward is small when the bearing clearance increases. Further, if the bearing clearance becomes smaller than the above threshold value, the reward increases.

Further, assuming that the sensor 5 is a temperature sensor and the state is the temperature of the bearing 2, the reward is large when the bearing temperature decreases, and the reward is small when the bearing temperature increases. Furthermore, if the bearing temperature falls below the threshold value, the reward will increase.

Further, when the sensor 5 is a vibration sensor and the state is an output signal of a vibration sensor provided on the bearing, the reward is large when the vibration level of the vibration sensor decreases, and the reward is small when the vibration level of the vibration sensor increases. Also, if the vibration level becomes smaller than the threshold value, the reward increases.

Further, assuming that the sensor 5 is a pressure sensor and the state is an output signal of a pressure sensor that measures the oil film pressure of a sliding bearing provided on the bearing, a reward is obtained in the case of a change approaching an appropriate output signal range of the predetermined pressure sensor. Increases, and in the case of a change that deviates from the output signal range, the reward is small. Further, if the signal of the pressure sensor is in the output signal range, the reward will be increased.

The function update unit updates the function for determining the output (learning content) according to the reward calculated by the reward calculation unit. For example, in the case of Q-learning, the action value function Q (st, at) represented by Equation 5 is used as a function for calculating the output (learning content). As a result of updating the action value function Q based on the relational information stored in the storage unit 8 in this way, this action value function Q becomes a learned model. The learning model storage unit 10 stores the learned model obtained above.

In the present embodiment, the case where the learning algorithm used by the machine learning unit 9 is reinforcement learning has been described, but the present invention is not limited to this. As for the learning algorithm, in addition to reinforcement learning, supervised learning, unsupervised learning, semi-supervised learning, and the like can also be applied.

Further, as the learning algorithm described above, deep learning that learns the extraction of the feature amount itself can also be used, and the machine follows other known methods such as neural networks, genetic programming, functional logic programming, and support vector machines. Learning may be performed.

Next, the control quantity calculation unit 11 reads the learned model output by the machine learning unit 9 from the learning model storage unit 10, and uses the learned model to actually measure the state information, that is, the sensor 5. From the newly measured information on the physical quantity or the voltage signal of the sensor 5, information for outputting a control signal to the inverter 4 or an operation signal to the air conditioner or the refrigeration equipment is obtained.

For example, in the case of the above Q-learning, the control amount calculation unit 11 substitutes the newly measured output information of the sensor as s into the learned action value function Q (s, a), and substitutes the action value function Q (s). , a) Find the action a that maximizes the value. The action a obtained in this way is output by the control amount calculation unit 11 as a control signal to the inverter 4 or an operation signal to the air conditioner or the refrigerating device.

In the trained model obtained by the machine learning unit 9, when arbitrary state information is input, the compressor rotates so that the state information becomes an appropriate value, that is, if the state information is not within the above sound range, the state information becomes an appropriate value. Output a signal that makes the number appropriate. Therefore, the control amount calculation unit 11 that calculates the control amount using the trained model can output an appropriate signal.

Next, the hardware that executes the storage unit 8, the machine learning unit 9, the learning model storage unit 10, and the control amount calculation unit 11 of the present embodiment will be described with reference to FIG. 7.

In the figure, the calculation device 100 that inputs / outputs to and from the sensor 5 and the inverter 4 stores a central processing unit CPU (Central Processing Unit) 101 and a program executed by the CPU 101, and temporarily stores information in the middle of calculation. Then, the storage device 102 that stores the calculation result, the signal from the sensor 5 that is the input information is read, temporarily stored in the storage device 102, and the result calculated by the CPU 101 is taken out from the storage device 102 and output to the inverter 4. It has F 103.

The storage unit 8 is executed by the storage device 102. The storage unit 8 reads the signal from the sensor 5 from the I / F 103, reads the output signal output to the inverter 4, associates it, and stores it in the storage device 102 as learning data. The machine learning unit 9 is executed by the CPU 101. The learning data stored in the storage device 102, that is, the storage device 102 is read, the learning program stored in the storage device 102 is executed by the CPU 101, and the learned model is output to the storage device 102. The learning model storage unit 10 is executed by the storage device 102. The control amount calculation unit 11 is executed by the CPU 101 based on the program stored in the storage device 102. The control amount calculation unit 11 acquires the signal from the sensor 104 via the I / F 103, temporarily stores it in the storage device 102, and uses the learned model stored in the storage device 102 to cause the CPU 101 to use the learned model. , The control amount is calculated from the signal of the sensor 104, and this control amount is output to the inverter 4 via the I / F 103.

Further, the signal processing unit 7 may be realized by the calculation device 100. If the signal processing unit 7, the storage unit 8, the machine learning unit 9, the learning model storage unit 10, and the control amount calculation unit 11 are realized by the same calculation device 100, the processing can be executed by one calculation device 100, which is efficient.

Next, the processing of the present embodiment will be described. In the learning process, first, learning data is collected and a learning data set is generated. The processing process for changing the rotation speed of the motor 3 described in the above embodiment controls the rotation speed of the compressor according to the state, and the measurement information during this period and the rotation speed at which the motor 3 transmitted to the device should rotate. Alternatively, the operation signal is stored in the storage unit 8 as learning data. The storage unit 8 sends the information of the physical quantity converted by the signal processing unit 7, the voltage signal of the sensor, the control signal to the inverter 4, which is the result of obtaining the control amount based on the information, or the air conditioner or the refrigerating equipment. The operation signals of are associated with each other and stored in the storage unit 8 as learning data.

The machine learning unit 9 has information on the physical quantity converted by the signal processing unit 7 stored in the storage unit 8, the voltage signal of the sensor, and the control signal to the inverter, which is the result of obtaining the control amount based on the information. A trained model is created by using the above training algorithm from the relation of the operation signal to the air conditioner or the refrigeration equipment. In this process, if the amount of training data to be collected increases, the learning algorithm advances the learning of the trained model, and the trained model becomes a trained model that can output actions that are appropriate outputs in response to various states.

When the trained model can be created, the process of utilizing the trained model is started. The control amount calculation unit 11 applies the learned model to the measured output signal of the current sensor to obtain the rotation speed at which the compressor (motor) should rotate or the output signal to the inverter 4, and obtains the output signal to the inverter 4. Or output to the inverter 4. The trained model is a model in which an appropriate output is required for various states by the machine learning unit 9. Therefore, the control amount calculation unit 11 can apply the trained model to the measured output signal of the sensor 5 to output an appropriate output.

In the learning process, the machine learning unit 9 machine-learns the learning data stored in the storage unit 8 to create a learned model. At the initial stage of learning, there is little learning data, and even if a trained model is created, the result calculated by the control amount calculation unit 11 based on the trained model may not always be a good result.

Therefore, in the initial stage of learning, control is performed by the rotation speed change processing process described in the above embodiment, and after a predetermined period, the control amount calculation unit 11 outputs using the learned model learned by the machine learning unit 9. The compressor is controlled by the control amount.

Here, after the predetermined period, a period of one year from the start of learning may be used. This is because it is thought that one year will learn about all the conditions. Further, instead of the above-mentioned rotation speed change processing process during learning, the compressor 6 is temporarily controlled by the control amount calculated by the control amount calculation unit 11 using the learned model learned up to that point, and for a predetermined period. It may be determined whether to switch to the trained model in earnest depending on whether or not the sensor measurement value is within the threshold value.

Specifically, when the signal processing unit 7 controls the motor 3 by the number of rotations calculated and output from the output signal measured by the sensor 5, the following processing is performed. Of the number of times the signal processing unit 7 obtained the estimated rotation speed from the control amount calculation unit 11 within the latest predetermined period, the estimated rotation speed obtained by the control amount calculation unit 11 from the output signal measured by the sensor 5. , The same output signal is calculated, and the number of times that the rotation speed output by the signal processing unit 7 matches is obtained. Next, the matching rate, which is the rate obtained by dividing the number of matching times obtained by the signal processing unit 7 by the total number of times the estimated number of rotations obtained by the control amount calculation unit 11 within a predetermined period, becomes a predetermined threshold value. When the coincidence rate exceeds the threshold rate, the motor 3 is controlled at the estimated rotation speed obtained by the control amount calculation unit 11. On the contrary, when the coincidence rate becomes equal to or less than the threshold value, the signal processing unit 7 may switch from the output signal measured by the sensor 5 to control the motor 3 according to the calculated and output rotation speed.

In the past, even if a physical quantity indicating the operating state of a compressor was measured by a sensor, how to control the operation of the compressor was different for each individual due to variations in the manufacturing of the compressor. When controlled by an exploratory method, it took time to converge, and during this time, failures occurred and efficiency deteriorated.

In the present embodiment, the sensor output value and the control value are associated with each other in the rotation speed change processing process by the exploratory method of the above embodiment, collected as a learning data set, stored in the storage unit 8, and stored in the learning data set. The machine learning unit 9 obtains a trained model from, and the control amount calculation unit 11 uses the obtained learned model to obtain an appropriate control value for the output value of the newly measured current sensor. do. As a result, once the trained model is generated, it can be controlled without using an exploratory method, so that the sensor measurement value can be quickly brought within the healthy range. That is, there is an effect that the time for the measured value of the sensor 5 to be within the threshold value within the sound range is shortened, failure and deterioration are prevented, and efficiency is improved.

In addition, if learning is insufficient, it may not converge or it may take time to converge. Therefore, the control amount calculation unit 11 using the trained model may be temporarily tried, and if the control amount calculation unit 11 converges within a predetermined time, the rotation speed change processing process may be switched to the control amount calculation unit 11. If it does not converge within a predetermined time, the learning data is further collected while returning to the rotation speed change processing process and controlling it. After that, the machine learning unit 9 tries to control the trained model separately, and repeats the process until it converges within a predetermined time. By doing so, there is an effect that the trained model can be safely used by replacing it with the control amount calculation unit 11 after sufficient learning.

1 Rotating shaft 2 Bearing, plain bearing 3 Motor 4 Inverter 5 Sensor, displacement sensor 6 Compressor, refrigerant compressor 7 Signal processing unit 8 Storage unit 9 Machine learning unit 10 Learning model storage unit 11 Control amount calculation unit

Claims (19)

In a compressor having a rotating shaft, a plain bearing that supports the rotating shaft, an electric motor that rotates the rotating shaft, and an inverter that controls the electric motor.
Provided on the sliding bearing or the rotating shaft portion where the sliding bearing is fitted, the displacement of the rotating shaft is measured, the temperature of the sliding bearing is measured, or the oil film pressure of the sliding bearing is measured. A sensor that outputs the measured value as an output signal,
A signal processing unit that calculates and processes the output signal output from the sensor, obtains the rotation speed at which the motor should rotate, and transmits the control signal to the inverter to control the rotation speed of the motor .
The measured value measured by the sensor, the rotation speed at which the motor should rotate with respect to the measured value, and the measured value of the sensor obtained as a result of controlling the rotation speed. A storage unit that associates certain post-control measurement values with each other and stores them as related information,
A compressor including a machine learning unit that machine-learns the related information stored in the storage unit and outputs a trained model. The compression according to claim 1, further comprising a control amount calculation unit that estimates an estimated rotation speed, which is the rotation speed at which the motor should rotate by applying the learned model to the output signal of the current measured sensor. Machine. The compressor according to claim 1 or 2, wherein the trained model output from the machine learning unit is transmitted to the outside via a network. The output signal of the sensor, the control signal obtained by the signal processing unit, and learning data associated with the output signal of the sensor obtained as a result of transmitting the control signal are transmitted to the outside via a network. The compressor according to any one of claims 1 to 3. The compressor according to claim 2, wherein the control amount calculation unit estimates the estimated rotation speed using the learned model read from the outside. In a compressor having a rotating shaft, a plain bearing that supports the rotating shaft, an electric motor that rotates the rotating shaft, and an inverter that controls the electric motor.
Provided on the sliding bearing or the rotating shaft portion where the sliding bearing is fitted, the displacement of the rotating shaft is measured, the temperature of the sliding bearing is measured, or the oil film pressure of the sliding bearing is measured. A sensor that outputs the measured value as an output signal,
A signal processing unit that calculates and processes the output signal output from the sensor, obtains the rotation speed at which the motor should rotate, and transmits the control signal to the inverter to control the rotation speed of the motor.
The measured value measured by the sensor, the number of rotations at which the motor should rotate with respect to the measured value, and the measured value of the sensor obtained as a result of controlling the number of rotations. The training data, which is the related information associated with the measured values after control, is output to the outside, and the training data is machine-learned to output the trained model. A control amount calculation unit that reads the trained model from the outside, applies the trained model to the output signal of the current sensor measured, and estimates the estimated rotation speed, which is the rotation speed at which the motor should rotate. Compressor with and. In a compressor having a rotating shaft, a plain bearing that supports the rotating shaft, an electric motor that rotates the rotating shaft, and an inverter that controls the electric motor.
Provided on the sliding bearing or the rotating shaft portion where the sliding bearing is fitted, the displacement of the rotating shaft is measured, the temperature of the sliding bearing is measured, or the oil film pressure of the sliding bearing is measured. A sensor that outputs the measured value as an output signal,
A signal processing unit that calculates and processes the output signal output from the sensor, obtains the rotation speed at which the motor should rotate, and transmits the control signal to the inverter to control the rotation speed of the motor.
The measured value measured by the sensor, the rotation speed at which the motor should rotate with respect to the measured value, and the measured value of the sensor obtained as a result of controlling the rotation speed. The trained model output from the externally provided machine learning unit is read from the outside by machine learning the training data, which is the related information that is related to the measured values after control. A compressor including a control amount calculation unit that estimates an estimated rotation speed, which is the rotation speed at which the motor should rotate, by applying the learned model to the output signal of the current measured sensor. The signal processing unit increases or decreases the rotation speed of the motor when the measured value deviates from a sound range determined by a threshold value that is a reference for the soundness of the measured value of the sensor, which is preset by the measured value. The compressor according to any one of claims 3 to 7, wherein the rotation speed of the motor is controlled to be increased, decreased, or maintained according to the measured value measured after the change in the direction. The signal processing unit does not change the measured value after changing the rotation speed of the motor when the measured value deviates from the sound range determined by the threshold value for the soundness of the measured value of the sensor. The compressor according to any one of claims 3 to 7, wherein when the deterioration occurs, the rotation speed of the motor is changed in the opposite direction to the increase / decrease in which the rotation speed of the motor is changed. The sensor is a displacement sensor provided on the rotating shaft portion fitted to the sliding bearing, and any one of claims 1 to 9 for measuring the bearing clearance of the sliding bearing over the entire circumference from the rotating shaft. The compressor according to item 1. The sensor is a displacement sensor that measures the bearing clearance of the slide bearing.
The signal processing unit increases or decreases the rotation speed of the motor when the measured value deviates from a sound range determined by a threshold value that is a reference for the soundness of the measured value of the sensor, which is preset by the measured value. The rotation speed of the motor is controlled to be increased, decreased, or maintained by the measured value measured after the change.
The threshold value of the sound range is the surface roughness of either the outer diameter of the rotating shaft or the inner surface of the plain bearing, or the sum or square of the surface roughness of the outer diameter of the rotating shaft and the surface roughness of the inner surface of the plain bearing. The compressor according to any one of claims 1 to 7 and 9, which is a value obtained by multiplying the sum square root by a coefficient. The sensor is a pressure sensor that measures the oil film pressure of the slide bearing.
The compressor according to claim 8 or 9, wherein the sound range is equal to or less than a preset upper limit value. The sensor includes a plurality of types of sensors: a displacement sensor that measures the displacement of the rotating shaft, a temperature sensor that measures the temperature of the slide bearing, and a pressure sensor that measures the oil film pressure of the slide bearing.
The signal processing unit compares the measured value of each of the sensors with the sound range set for each sensor, and when even one of the measured values of the sensor deviates from the sound range, the rotation of the electric motor. The compressor according to any one of claims 1 to 9, wherein the number is changed and the measured values of all the sensors are repeated until they are within the respective sound ranges. An air conditioner including the compressor according to any one of claims 1 to 13. A refrigerator comprising the compressor according to any one of claims 1 to 13. Compressor control learned to create a trained model for controlling a compressor having a rotating shaft, a plain bearing that supports the rotating shaft, an electric motor that rotates the rotating shaft, and an inverter that controls the electric motor. It ’s a model creation method.
The slide bearing is provided on a rotation shaft portion to which the slide bearing or the slide bearing is fitted , and a sensor for measuring the displacement of the rotation shaft, the temperature of the slide bearing, or the oil film pressure of the slide bearing is used to measure the slide bearing. A measurement step that outputs the measured value that measures the state of
A control step that calculates and processes the output signal output from the sensor, obtains the rotation speed at which the motor should rotate, and transmits it as a control signal to the inverter to control the rotation speed of the motor .
A storage unit that stores the output signal of the sensor, the rotation speed at which the motor should rotate with respect to the output signal, and the output signal of the sensor obtained as a result of controlling the rotation speed as association information. Steps and
A compressor-controlled trained model creation method including a machine learning step of machine learning the association information stored in the storage unit and outputting the trained model. The compressor control trained model creation method according to claim 16 is included.
A compressor control method comprising an estimation step of applying the trained model to the measured current output signal of the sensor to estimate the number of revolutions to be rotated by the motor. Compression that creates learning data for performing machine learning to control a compressor having a rotating shaft, a sliding bearing that supports the rotating shaft, an electric motor that rotates the rotating shaft, and an inverter that controls the electric motor. It is a data creation method for machine control learning.
The slide bearing is provided on a rotation shaft portion to which the slide bearing or the slide bearing is fitted, and a sensor for measuring the displacement of the rotation shaft, the temperature of the slide bearing, or the oil film pressure of the slide bearing is used to measure the slide bearing. A measurement step that outputs the measured value that measures the state of
A control step that calculates and processes the output signal output from the sensor, obtains the rotation speed at which the motor should rotate, and transmits it as a control signal to the inverter to control the rotation speed of the motor.
A storage unit that stores the output signal of the sensor, the number of rotations at which the motor should rotate with respect to the output signal, and the output signal of the sensor obtained as a result of controlling the number of rotations as association information. A data creation method for compressor control learning with steps. In the control step, the measured value does not change or deteriorates after changing the rotation speed of the motor when the measured value deviates from the sound range determined by the threshold value for the soundness of the measured value of the sensor. The method for creating data for compressor control learning according to claim 18, wherein the rotation speed of the motor is changed in the opposite direction to the increase / decrease in which the rotation speed of the motor is changed.

Images