Method for controlling a sensorless motor of an air compressor

Through the controller performing speed control and d-axis current forced rotation technology in the air compressor motor, the problems of initial drive and position alignment of the motor under sensorless conditions are solved, and the response performance of the air compressor and vehicle acceleration efficiency are improved. Cooling and sensor heating problems in traditional technology are solved.

CN112628129BActive Publication Date: 2024-11-01HYUNDAI MOTOR CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011003533.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-10-08
Filing Date
2020-09-22
Publication Date
2024-11-01
Estimated Expiration
2040-09-22

AI Technical Summary

Technical Problem

In the prior art, the air compressor motor generates a large amount of heat during high-speed operation, and the cooling performance is limited. The Hall sensor is easily heated, resulting in delayed control response and low efficiency. Especially in the initial driving stage without sensor control, it is difficult to determine the motor position, which affects the impact of Drive performance of air compressor.

Method used

The controller performs speed control, estimates the motor stop state, determines the rotor position, and applies d-axis current to force the rotor to the alignment target position during the stop state, thereby achieving initial drive and position alignment of the motor under sensorless conditions , avoid delays in open loop control.

Benefits of technology

Improve the speed control response of the initial stage of the air compressor drive, enhance the air supply and vehicle acceleration performance of the fuel cell stack, reduce unnecessary torque and cooling requirements for the motor drive, improve driving efficiency and fuel efficiency, and avoid rotor pairing Quasi-failed.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112628129B_ABST
    Figure CN112628129B_ABST
Patent Text Reader

Abstract

A method for controlling a sensorless motor of an air compressor is provided. The method controls the initial drive of the sensorless motor of the air compressor, overcomes related technical problems, and improves control response. At a time point when the motor is determined to be in a stopped state, the rotor position of the motor finally estimated by the sensorless control logic is determined as an alignment target position. An alignment start position is determined from the alignment target position according to a predetermined alignment offset angle. The rotor position of the motor is controlled to change from the determined alignment start position to the alignment target position.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure generally relates to a method for controlling a sensorless motor of an air compressor, and more particularly, to a method for controlling an initial drive of a sensorless motor of an air compressor, which can overcome the problems of related art sensorless control methods and improve control response. Background Art

[0002] A fuel cell system is applied to a hydrogen fuel cell vehicle in an environmentally friendly vehicle. The fuel cell system includes a fuel cell stack, a hydrogen supply unit, an air supply unit, a heat and water management system, and a control unit. The fuel cell stack generates electrical energy through an electrochemical reaction of a reaction gas (i.e., hydrogen is used as a fuel gas and oxygen is used as an oxidizing gas). The hydrogen supply unit supplies hydrogen used as a fuel gas to the fuel cell stack. The air supply unit supplies air including oxygen to the fuel cell stack. The heat and water management system controls the operating speed of the fuel cell stack and performs a water management function. The control unit controls the overall operation of the fuel cell system.

[0003] Figure 1 is a diagram showing the configuration of a fuel cell system known in the related art.

[0004] In the fuel cell system, the oxygen supply unit 10 includes a hydrogen storage (e.g., a hydrogen tank, not shown), a regulator (not shown), a hydrogen supply pipeline 11, a hydrogen pressure control valve (also called a "hydrogen supply valve") 12, a hydrogen pressure sensor 13, a hydrogen recirculator 15, etc.

[0005] In addition, the air supply unit 20 includes an air supply line 21, a filter 22, a blower or air compressor 23, a humidifier 24, etc. In addition to components not shown in the drawings, such as an electric water pump (cooling water pump), a water tank, a radiator, and a three-way valve, the heat and water management system 30 includes a dehydrator 31, a drain valve 32, etc.

[0006] Regarding the hydrogen supply unit 10 that supplies hydrogen (ie, fuel) to the fuel cell stack 1 , high-pressure hydrogen supplied from a hydrogen tank by the hydrogen supply unit 10 is decompressed to a predetermined pressure by a regulator before being supplied to the fuel cell stack 1 .

[0007] When hydrogen decompressed by the regulator is supplied, the amount of hydrogen is controlled by pressure control depending on the operating conditions of the fuel cell stack 1. The opening of a hydrogen pressure control valve or a fuel supply valve (FSV) 12 provided on the hydrogen supply line 11 is controlled according to the required output, thereby supplying a necessary amount of hydrogen to the fuel cell stack 1.

[0008] Here, a fuel cell control unit (FCU, not shown) controls the opening degree of the hydrogen pressure control valve 12 by receiving a sensing value fed back from a hydrogen pressure sensor 13 provided on a hydrogen supply line, so as to control the pressure and amount of hydrogen.

[0009] Furthermore, after reacting in the fuel cell stack 1 , a remaining amount of unreacted hydrogen is discharged through the output end of the anode (or hydrogen electrode or fuel electrode) or is recycled to the stack anode input end through the hydrogen recycler 15 .

[0010] The hydrogen recycler 15 is intended to allow the reuse of unreacted hydrogen. The hydrogen recycler 15 comprises a recirculation line 16 extending from an exhaust line 14 provided on the anode output side of the fuel cell stack 1 to the anode input side.

[0011] In addition, the hydrogen recirculator 15 includes an ejector 17 or a recirculation blower, or an ejector and a recirculation blower, receives unreacted hydrogen through the recirculation line 16 and supplies the unreacted hydrogen to the fuel cell stack 1 .

[0012] In addition, a purge valve 18 for purifying the anode side is provided on the anode output side of the fuel cell stack 1. The purge valve 18 is periodically operated to open and close so that impurities such as nitrogen and water are discharged and removed from the anode of the fuel cell stack 1 together with hydrogen. Thus, the utilization rate of hydrogen is improved.

[0013] Although not shown in the figure, the fuel cell control unit (FCU) not only controls the operating pressure of hydrogen by controlling the opening of the hydrogen pressure control valve 12, but also controls the values ​​18 and 32 in the air compressor 23, the blower and other components of the system.

[0014] Recently, as the performance of fuel cell vehicles has been further improved, the output power required of fuel cells used as a power source for the vehicles has also been increasing.

[0015] Therefore, the air compression ratio and the required flow rate in the air compressor 23 that supplies air to the fuel cell are also increasing. For this reason, it is also necessary to increase the maximum speed of the motor that compresses the air in the air compressor.

[0016] However, increasing the maximum speed of the motor in the air compressor 23 may have the following problems.

[0017] The motor of the air compressor generates a lot of heat when it runs at high speed. Since the motor is easily heated, cooling is basically required to reduce the temperature of the motor.

[0018] However, the cooling capacity provided by the vehicle is limited, and therefore, the ability to reduce the temperature of the motor of the air compressor is limited.

[0019] In addition, the air compressor also includes a Hall sensor, i.e. a position sensor, to control the position of the motor. However, the Hall sensor is susceptible to heat.

[0020] Therefore, as mentioned above, as the motor speed of the air compressor increases, the motor generates a greater amount of heat. As the motor temperature increases, the use of the Hall sensor may be limited.

[0021] In this regard, it is necessary to omit the Hall sensor and obtain sensorless control technology. If the air compressor can be driven without the Hall sensor, the degree standard for considering cooling performance can be lowered, thereby reducing the cost of the air compressor.

[0022] In the related art sensorless control, a method of estimating the position, angle, and speed of a motor by estimating the back electromotive force of the motor is generally used.

[0023] However, in the related art, when the motor starts to be driven in the early stage, the back electromotive force is not generated, and therefore, the initial position of the motor cannot be determined.

[0024] In this regard, in the early stage of resuming driving after the motor stops, the motor speed may be generated by performing open-loop control for hundreds of milliseconds, and then, sensorless control may be performed based on back electromotive force.

[0025] However, considering the driving characteristics of the air compressor in which response is regarded as important, performing open-loop control for hundreds of milliseconds as described above may be a disadvantage. Summary of the invention

[0026] Therefore, the present disclosure is proposed in view of the above-mentioned problems arising in the related art, and the present disclosure proposes a method for controlling a sensorless motor of an air compressor, and more specifically, proposes a method for controlling the initial drive of a sensorless motor of an air compressor, which can overcome the problems of the sensorless control method of the related art and improve the control response.

[0027] In order to achieve the above-mentioned purpose, according to one aspect of the present disclosure, a method for controlling a sensorless motor of an air compressor is provided. The method may include: performing speed control by a controller to stop the motor of the air compressor; during the execution of speed control, the controller determines that the motor is in a stopped state according to the motor speed estimated by the sensorless control logic; the controller determines the rotor position at the time point when the motor is determined to be in a stopped state, and the rotor position is determined by the rotor position of the motor finally estimated by the sensorless control logic; the controller determines the alignment start position to become the alignment target position from the alignment target position according to a predetermined alignment offset angle; the controller changes the rotor position of the motor from the determined alignment start position to the alignment target position; and in a state where the rotor position of the motor is aligned with the alignment target position, the controller responds to the motor drive request and performs sensorless control by setting the alignment target position at the initial position to drive the motor.

[0028] When determining that the motor is in a stopped state, if the motor speed estimated by the sensorless control logic reaches a predetermined reference speed, the controller may determine that the motor is in a stopped state.

[0029] Additionally, the controller may determine the reference speed to be a speed greater than 0 rpm.

[0030] In addition, when determining the alignment start position, the controller may determine the alignment start position by subtracting the alignment offset value from the predetermined alignment target position in a direction opposite to the rotation direction of the rotor.

[0031] In addition, when controlling the rotor position change of the motor, the controller can control the current applied to the motor to force the rotor to rotate within a predetermined forced rotor rotation time so that the rotor position of the motor changes from the alignment start position to the alignment target position.

[0032] Here, the current may be a phase current, and the controller may control the phase of the phase current applied to the motor so that the rotor position of the motor changes from the alignment start position to the alignment target position within the forced rotor rotation time.

[0033] In addition, the controller may control the phase of the phase current applied to the motor so that the rotor position of the motor changes linearly at a predetermined rate of change.

[0034] In addition, when controlling the change in the rotor position of the motor, the controller can apply a d-axis current to move the rotor position of the motor from an alignment start position to an alignment target position within a predetermined position alignment time starting from a time point when the motor is determined to be in a stopped state; and after a forced rotor rotation time has passed since a time point when the motor is determined to be in a stopped state, stop controlling the phase of the phase current during the application of the d-axis current before the position alignment time has passed.

[0035] Furthermore, the controller may apply a current to the motor so that the rotor position of the motor changes from the alignment start position to the alignment target position at a predetermined rate of change.

[0036] In addition, when controlling the rotor position of the motor to change, the controller may apply a d-axis current to change the rotor position of the motor from the alignment start position to the alignment target position within a predetermined position alignment time.

[0037] Furthermore, the air compressor may be an air compressor for a fuel cell system configured to supply air to a fuel cell stack.

[0038] As described above, the method of controlling a sensorless motor of an air compressor according to the present disclosure can improve the speed control response in the initial stage of driving the air compressor, thereby improving the air supply response of the fuel cell stack, the stack output response, and the vehicle acceleration performance.

[0039] In addition, when the driving of the motor is resumed after stopping, speed control reliability can be obtained in the early stage of driving. Therefore, it is possible to prevent the torque from being applied at an unnecessary position, thereby improving the driving efficiency of the motor and the fuel efficiency of the vehicle.

[0040] Furthermore, the method of controlling a sensorless motor of an air compressor according to the present disclosure can overcome the problem of rotor alignment failure, that is, when the rotor position is aligned with the final position estimated by the sensorless control logic when the motor stops, the rotor may not be aligned to the expected position. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The above and other objects, features and advantages of the present disclosure will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0042] Figure 1 is a diagram showing the configuration of a fuel cell system known in the prior art;

[0043] Figure 2 is a graph showing acceleration and deceleration of a typical air compressor during fuel cell vehicle travel;

[0044] Figure 3 is a graph showing the motor speed in the initial stage during open loop control and driving in the control process of the related art;

[0045] Figure 4A , Figure 4B , Figure 4C and Figure 4D is a graph showing the motor speed when the open loop control is omitted in the initial stage of the motor drive;

[0046] Figure 5is a graph showing a method according to a comparative example, which aligns a rotor position and performs sensorless control at the aligned position when a motor is stopped;

[0047] Figure 6 and Fig. 7A is a diagram showing a rotor position alignment method performed for sensorless motor control when the motor is stopped;

[0048] Figure 7B is a diagram showing a case where no d-axis current is applied;

[0049] Figure 8 shows the rotor position measured during a motor stop state when the rotor is fixed at a 180° position by applying a d-axis current during the motor stop state;

[0050] Fig.9A and Fig. 9B is a diagram of analyzing and comparing examples of motor motions where position alignment has succeeded and failed;

[0051] Fig.10 is a block diagram schematically illustrating a configuration of a system for controlling a sensorless motor according to the present disclosure;

[0052] Fig.11 is a graph showing a method of aligning a rotor position when the motor is stopped during control of a sensorless motor according to the present disclosure;

[0053] Fig.12 is a diagram illustrating a method of aligning a rotor position according to the present disclosure, wherein current application states of a rotor, a stator, and a stator coil, and magnetic flux induced to the stator are illustrated;

[0054] Fig.13 is a graph showing a control method for rotor position alignment according to the present disclosure, which is performed when the motor is stopped;

[0055] Fig.14A and Fig. 14B is a flow chart showing the overall process of controlling a sensorless motor according to the present disclosure;

[0056] Fig.15 is a flowchart showing in more detail a process of setting a position alignment mark and a forced rotor rotation mark in a control unit in a process of controlling a sensorless motor according to the present disclosure;

[0057] Fig.16 is a flow chart showing a process of estimating rotor position during motor stop according to the present disclosure; and

[0058] Fig.17is a flowchart illustrating a process of determining a d-axis current command and a q-axis current command according to a position alignment mark in a control unit according to the present disclosure. DETAILED DESCRIPTION

[0059] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art to which the present disclosure relates can easily put the present disclosure into practice. However, the present disclosure is not limited to the following embodiments and can be implemented in many different forms.

[0060] It will be understood that the terms "comprises," "comprising," "including," "having" and any variations thereof herein are intended to cover a non-exclusive inclusion, unless explicitly described to the contrary.

[0061] The present disclosure relates to a method for controlling a sensorless motor of an air compressor, and more particularly, to a method for controlling an initial drive of a sensorless motor of an air compressor, which can overcome the problems of the sensorless control method of the related art and improve the control response.

[0062] The control method according to the present disclosure can be used to control a sensorless motor of an air compressor. In this context, the air compressor can be Figure 1 The air compressor indicated by reference numeral "23" in the figure is Figure 1 The fuel cell stack indicated by reference numeral “1” in the figure supplies air.

[0063] Hereinafter, the motor is the motor of the air compressor, the speed of the air compressor is the speed of the motor, and the position and angle of the motor are the rotation position and rotation angle of the rotor of the motor.

[0064] Furthermore, hereinafter, the braking device brakes the motor to stop the motor, and the position aligning device aligns and fixes the rotor at a predetermined position.

[0065] Related art sensorless control method

[0066] First, for a better understanding of the present disclosure, a known sensorless control method will be briefly described.

[0067] Figure 2 is a graph showing the acceleration and deceleration of a typical air compressor during fuel cell vehicle driving. Figure 2 As shown, the speed of the air compressor repeatedly accelerates and decelerates rapidly according to the acceleration and deceleration driving pattern of the driver.

[0068] Furthermore, the air supply response of the fuel cell stack is determined based on the speed control response of the air compressor and serves as a key factor in the responsiveness of the stack output.

[0069] Therefore, since the acceleration performance of the vehicle depends on the responsiveness of the stack output, the speed control response of the air compressor is a key factor during fuel cell vehicle driving.

[0070] Reference Figure 2 It is obvious that acceleration and deceleration of the air compressor occurs very frequently during driving, and driving of the motor of the air compressor is repeatedly stopped and resumed.

[0071] That is, if a short period of time during which the motor resumes driving after being stopped is referred to as an initial driving condition of the motor, the initial driving condition of the motor occurs very frequently.

[0072] When based on Figure 2 When the acceleration and deceleration pattern shown controls the motor speed of the air compressor, the stopping and resuming of the motor driving frequently occur. According to the sensorless control of the related art, it is necessary to perform open loop control at the initial stage of each resuming driving.

[0073] When performing known sensorless control processes via back EMF, such as Figure 3 As shown, speed control is performed in the initial stage of motor driving.

[0074] Figure 3 1 is a graph showing the motor speed at the initial stage of driving in the open loop control and known control process. In an air compressor without a Hall sensor (ie, position sensor) and its control system, open loop control of a stopped motor is started at the initial stage of driving.

[0075] When the motor is stopped and the drive is resumed, open loop control is performed in the initial stage of the drive. Since no back electromotive force is generated when the motor is stopped, there is no Hall sensor for detecting the motor position, and it is impossible to estimate the initial position of the motor only by the sensorless control logic of the prior art.

[0076] That is, in the sensorless control logic for estimating the position and speed of the motor by the back electromotive force, the position of the motor can be estimated by the back electromotive force generated at the predetermined motor speed only after the driven motor reaches the predetermined motor speed. Therefore, when the driving of the motor is resumed, in the initial stage of the driving, by generating a magnetic field at a predetermined position in the stopped state, the rotor of the motor rotates using the magnetic field, thereby providing a predetermined or higher speed of the motor.

[0077] exist Figure 3 In FIG. 1 , reference numeral “A” indicates a time period of a drive start portion. Part A is a drive start portion in which an open loop control of the motor and an initial stage of the motor drive are performed. Figure 3 In the example, open-loop control is executed for 0.27 seconds to promote stability in the early stages of motor drive.

[0078] refer to Figure 3 ,It is obvious that after part A, the driving of the motor is successful and the speed of the motor increases rapidly.

[0079] As described above, when the motor of the air compressor is started from a stopped state of the sensorless system, open loop control is performed so that the motor rotates at a predetermined speed or higher, even for a short period of time in the initial stage.

[0080] For example, the open loop control is executed for an average of 0.27 seconds from the stop state. After the motor drive command is instructed, there is a drive delay corresponding to the time period of the open loop control.

[0081] Regarding the response standard of the air compressor, the period until the maximum speed (eg, 100,000 rpm) is reached is about 1 second or less. Therefore, the 0.27 second open-loop control and drive delay is a factor that reduces the response performance by about 30%.

[0082] FIG. 4A to FIG. 4D This is a graph showing the motor speed when the open loop control is omitted in the initial stage of motor driving. FIG. 4A to FIG. 4D The purpose is to illustrate the problems when open-loop control is omitted.

[0083] FIG. 4A to FIG. 4D A case is shown in which the motor speed control is performed by omitting the open-loop control, ie, the response reduction factor, and driving the sensorless control logic.

[0084] like FIG. 4A to FIG. 4D As shown, when the sensorless control logic is directly executed in a state where the position in the early stage of driving is not determined, speed instability oscillation may occur for 100 milliseconds, and the expected speed control may start after a short time of speed oscillation.

[0085] This is because the drive of the rotor is not synchronized with the torque applied by the magnetic field generated from a position independent of the rotor position. Therefore, when the magnetic field rotates once, the rotor is not driven, and the speed estimated by the sensorless control logic has a noisy value.

[0086] In the case where a separate Hall sensor is provided for testing, when speed control is performed by operating the sensorless control logic in a state where the initial position is known by means of the Hall sensor, it can be understood that speed control is directly performed without speed instability, such as Figure 4C and Figure 4D Example shown.

[0087] In this case, torque is applied by generating a magnetic field from a known physical position of the rotor at an early stage. Therefore, the rotor can be synchronized and, therefore, speed control can be performed.

[0088] Therefore, in sensorless control, the initial position of the rotor can be estimated or known.

[0089] Methods for comparing examples

[0090] As described above, in the related art, a predetermined or higher motor speed is obtained by inevitably performing open-loop control, so that the position and speed of the motor can be estimated in the early stage of driving after the motor stops. Here, open-loop control is a response degradation factor. In order to overcome the problems of the related art, a sensorless control method for a driving method can be considered. In the sensorless control method, when the motor stops, the rotor of the motor moves and is fixed at a predetermined position. Afterwards, at the time point when the drive of the motor is resumed, the position of the fixed motor is used as the initial position.

[0091] According to this improved method, the fixed rotor position during the previous motor stop state is used as the initial position when the motor drive is resumed later. Therefore, as an advantage, sensorless control can be started quickly, and the open loop control of the related art is unnecessary.

[0092] In the following, the improved method will be described in more detail.

[0093] Hereinafter, the method of controlling the sensorless motor of the air compressor may be performed by a control unit or a controller, and all control processes (to be described later) may be performed by the control unit.

[0094] Furthermore, the control unit for controlling the sensorless motor may be a typical control unit for controlling a motor of an air compressor in a fuel cell system.

[0095] Furthermore, a sensorless motor is a motor without a Hall sensor for detecting rotor position and speed.

[0096] Since the sensorless motor has no Hall sensor, it is necessary to perform sensorless control of estimating the position and speed of the motor by a predetermined method and then using the estimated values ​​to control the driving of the sensorless motor.

[0097] In a sensorless motor, the initial position cannot be obtained when the motor is stopped because there are no Hall sensors. Figure 5 An improved method to overcome this problem is shown.

[0098] like Figure 5 As shown, an improved method for overcoming the related technical problems provides a process of aligning the rotor position so that the rotor can be at a predetermined angle and position during the process of stopping the motor.

[0099] In addition, when the motor stops, the rotor is fixed in a state of angle and position alignment. Thereafter, in the initial stage of driving to restart the motor speed control, the angle and position of the rotor fixed after alignment are used as the initial angle and position, and sensorless control and speed / current control are performed from this position.

[0100] Reference Figure 5 Describing in more detail, Figure 5 Point T1 in is the time point when the motor stops (hereinafter referred to as the "motor stop point"). Figure 5 Point T2 in FIG. 8 is the time point when the motor drive is resumed and the sensorless control is resumed (hereinafter referred to as the “restart point”).

[0101] like Figure 5 As shown, rotor position alignment is performed, that is, a process of controlling the rotor to be fixed at a predetermined position when the motor is stopped.

[0102] Therefore, after the motor stops, the rotor position is aligned and fixed in a predetermined position.

[0103] Here, the predetermined position may be the rotor position finally estimated by the sensorless control logic at the motor stop determination point, ie, the time point when the control unit determines that the motor is in a stop state satisfying a predetermined condition, which will be described later.

[0104] When the motor is stopped as described above, the rotor is fixed in the final position aligned during the stopping of the motor before the driving of the motor is resumed later.

[0105] Afterwards, when the drive of the motor is resumed, the initial position of the rotor restored to sensorless control is the final position of the rotor fixed after alignment during the motor stop process, and the sensorless control starts from the final position at the motor restart point (i.e. the time point when the motor drive is resumed).

[0106] That is, the motor driving and sensorless control are performed using the final position (ie, the rotor position aligned during the previous motor stop) as the initial position.

[0107] Figure 6 and Fig. 7A is a diagram showing an improved rotor position alignment method according to a comparative example, which is performed for sensorless motor control when the motor is stopped, and Figure 7B is a diagram showing a case where no d-axis current is applied.

[0108] Figure 6 The left part of shows the motor speed during the motor speed control process, in which the motor is accelerated after the motor driving starts, maintained at a constant speed, and then decelerated to stop.

[0109] When the control unit decelerates the motor to a predetermined speed to stop the motor, if the speed of the motor reaches a predetermined reference speed, the motor is determined to be in a stopped state. The rotor position finally estimated by the sensorless control logic at the motor stop determination point is determined to be a fixed position.

[0110] Here, the reference speed may be determined as a positive (+) speed greater than 0 RPM.

[0111] Then, if Figure 6 As shown in the right part of , the d-axis current for controlling the rotor position to a fixed position is applied to the motor so that the rotor of the motor can be fixed at the rotor position finally estimated by the sensorless control logic at the motor stop determination point. In this way, the rotor is fixed at a fixed position.

[0112] Here, the d-axis current for moving and fixing the rotor to a fixed position is applied for a predetermined reference period of time. Due to the application of the d-axis current, the rotor position is moved and fixed at the fixed position.

[0113] After the rotor is fixed, the control unit stops all pulse width modulation (PWM) control processes that generate and output PWM signals and keeps the motor stopped until the motor resumes driving.

[0114] Fig. 7A shows the state of rotor position alignment, Figure 7B The rotor position is misaligned.

[0115] Figure 7B The dotted rectangular portion in the figure represents the state where the rotor rotates due to inertia.

[0116] Since the improved method (comparative example method) is applied to a sensorless motor, the motor is not provided with a Hall sensor (position sensor). Fig. 7A and Figure 7B The rotor angle obtained from the signal (i.e. the angle based on the Hall sensor) is shown in FIG.

[0117] exist Fig. 7A and Figure 7B In this case, “angle based on the Hall sensor” can be understood as indicating the actual position and angle of the rotor.

[0118] Reference Fig. 7A , it can be understood that the state of aligning the rotor position by applying the d-axis current during the motor stop process, and it can be understood that the d-axis current is applied at the stop determination point before the motor speed reaches 0rpm.

[0119] Furthermore, it can be understood that, due to the application of the d-axis current, after the motor stops, the rotor position and angle move and are fixed at a predetermined position, i.e., the position finally estimated by the sensorless control logic at the stop determination point (i.e., a fixed position).

[0120] like Fig. 7A As shown, if the control unit generates and outputs a PWM signal and applies a d-axis current, instead of directly blocking the PWM signal when the motor stops, the rotor position can be aligned to the position finally estimated by the sensorless control logic when the motor stops (i.e., a fixed position), which can be understood from the information about the angle of the Hall sensor signal (i.e., the angle based on the Hall sensor).

[0121] On the contrary, refer to Figure 7B It can be understood that when the d-axis current is not applied during the motor stop process, the rotor rotates (see the rectangular part drawn with dotted lines).

[0122] That is, from the information about the angle (i.e., the angle based on the Hall sensor) obtained from the Hall sensor signal (i.e., the position sensor signal), it is obvious that the rotor position changes as the rotor rotates due to inertia even without applying current after the PWM signal is blocked. After the rotor continues to rotate, the rotor stops at an unknown position.

[0123] As described above, during the motor stop process, the d-axis current is applied so that the rotor position and angle are aligned to the rotor position (i.e., fixed position) and angle finally estimated by the sensorless control logic when the motor stops. Thereafter, in the stopped state, the rotor position is fixed at the aligned position.

[0124] Afterwards, when the driving of the motor starts, the alignment position of the rotor (i.e., the fixed position) is set as the motor restart point (i.e., Figure 5 The initial position at point T2 in the figure is used to directly start sensorless control without open-loop control. This can eliminate the drive delay that occurs in the related art and improve the response of motor control and drive.

[0125] In addition, Figure 8 , the motor speed and rotor position (ie, angle) resulting from repeatedly driving and stopping the motor by performing an improved method of performing initial position alignment, wherein the rotor position is fixed at a position ultimately determined by the sensorless control logic during motor stopping.

[0126] Figure 8 The rotor position measured by applying a d-axis current during a motor stop state when the rotor is fixed at a 180° position (ie, the final estimated position) is shown in order to determine whether the rotor is actually fixed at the expected position (ie, angle).

[0127] exist Figure 8In the 1990s, a separate Hall sensor is used to measure the rotor position (i.e. angle). The Hall sensor used is a sensor with a resolution of 60. When the rotor is positioned at an angle of 180°, the Hall sensor indicates 150° or 210°.

[0128] According to the use Figure 8 From the results of the Hall sensor measurements shown in , it can be understood that the rotor position occasionally deviates from the 180° position to reach the opposite position (alignment failure).

[0129] Therefore, there is a need for an improved method for eliminating such alignment failures.

[0130] Fig.9A and Fig. 9B Examples of motor motions that succeeded and failed position alignment are analyzed and compared. Fig.9A and Fig. 9B middle, Fig.9A A successful example is shown, in which the rotor is positioned to the alignment target position during the motor stop state, while Fig. 9B An example of a failure is shown where the rotor is positioned misaligned with the alignment target position by 180°.

[0131] exist Fig.9A and Fig. 9B In the figure, “Current Speed” indicates the motor speed, and “Estimated Angle” indicates the angle finally estimated by the sensorless control logic when the motor speed reaches the predetermined reference speed when the motor stops.

[0132] In addition, Fig.9A and Fig. 9B In the diagram, "Hall sensor angle" refers to the rotor angle (or position) measured by adding a separate Hall sensor. Although the actual sensorless motor is not equipped with Hall sensors, Hall sensors are added to measure the rotor angle for comparison.

[0133] The Hall sensor angle indicates the actual position of the rotor. A Hall sensor angle close to the estimated angle means the rotor is aligned to the expected position.

[0134] If the rotor position alignment is successful, the current is fixedly applied to the position of the estimated angle obtained by the sensorless control logic, thereby generating an induced magnetic flux. Therefore, the rotor position is fixed at a position consistent with the direction of the magnetic flux generated by the permanent magnet of the rotor. Fig.9A As shown, the actual angle of the rotor (ie, the Hall sensor angle) is close to the estimated angle.

[0135] If the rotor position alignment fails, such as Fig. 9B As shown, the estimated angle has an error of about 180° with the Hall sensor angle.

[0136] This is because, when trying to fix the rotor position at the point in time when the motor drive stops, the rotor rotates further due to inertia instead of stopping at the expected position. In this state, if current is continuously applied to a specific position, the rotor may further rotate about 180° from the expected position before stopping.

[0137] For example, when the rotor is aligned in a direction opposite to the magnetic flux induced by the stator, the push and pull forces can balance, thereby fixing the rotor in an opposite position so as not to rotate in either direction.

[0138] Exemplary Embodiments

[0139] Therefore, the present disclosure aims to provide an improved method that can overcome the problems occurring in the sensorless control method of the related art and the sensorless control method, in which open-loop control is performed in the initial stage of driving when the motor resumes driving after stopping, and in the sensorless control method, when the motor stops, the rotor position of the motor is fixed and aligned according to the final estimated position value finally estimated by the sensorless control logic.

[0140] In particular, according to the present disclosure, a method for controlling a sensorless motor of an air compressor is also provided, which can overcome the problem of rotor alignment failure, that is, when the motor stops, the rotor is not aligned to the expected position through the rotor position alignment performed according to the final estimated position value.

[0141] According to the present disclosure, when the motor stops, the rotor position alignment is performed to control the rotor to be fixed at the alignment target position. This rotor position alignment is similar to the above reference Figure 5 , Figure 6 and Fig. 7A The method described is essentially the same.

[0142] Here, the expected alignment position (hereinafter referred to as the "alignment target position") can be the rotor position finally estimated by the sensorless control logic at the time point (or the motor stop determination point) when the control unit or controller determines that the motor is in a stopped state that satisfies predetermined conditions.

[0143] Furthermore, after the motor stops, before the driving of the motor is resumed later, the rotor is fixed at the position aligned during the motor stopping process (ie, the alignment target position).

[0144] Afterwards, when the drive of the motor is resumed, the initial position of the rotor at the restart point of the sensorless control is the final rotor position (i.e., the alignment target position) that has been fixed after alignment during the motor stop process. The sensorless control starts from the final position at the motor restart point, at which the drive is resumed.

[0145] That is, the subsequent motor driving and sensorless control are performed using the final position, that is, the rotor position aligned during the previous motor stop process, as the initial position.

[0146] Figure 6 The left part of shows the motor speed during the motor speed control process, in which the motor is accelerated after the motor driving starts, maintained at a constant speed, and then decelerated to be in a stopped state.

[0147] When the control unit decelerates the motor to a predetermined speed to stop the motor, if the speed of the motor reaches a predetermined reference speed, the motor is determined to be in a stopped state. The rotor position finally estimated by the sensorless control logic at the motor stop determination point (i.e., the time point when the motor is determined to be in a stopped state) is determined to be aligned with the target position (or fixed position).

[0148] Here, the reference speed may be determined as a positive (+) speed greater than 0 rpm.

[0149] Then, if Figure 6 As shown in the right part of , a d-axis current for controlling the rotor position to the alignment target position is applied to the motor so that the rotor of the motor can be fixed at the alignment target position, that is, the rotor position finally estimated when the motor is determined to be in a stopped state. In this way, the rotor rotates and is fixed at the alignment target position.

[0150] Here, the d-axis current for moving (rotating) and fixing the rotor to the alignment target position is applied for a predetermined reference period of time. Due to the application of the d-axis current, the rotor position is moved and fixed at the alignment target position.

[0151] After the rotor is fixed, the control unit stops all PWM control processes of generating and outputting PWM signals, and maintains the motor stopped state until the driving of the motor is resumed.

[0152] Furthermore, according to the present disclosure, after an alignment start position for rotor position alignment is determined, a d-axis current for position alignment is applied to the motor when the rotor is positioned at the determined alignment start position.

[0153] That is, a new concept of aligning the start position is introduced and used.

[0154] Fig.10 is a block diagram schematically showing a configuration of a system for controlling a sensorless motor according to the present disclosure, Fig.11 is a graph illustrating a method of aligning a rotor position when the motor is stopped in controlling a sensorless motor according to the present disclosure.

[0155] also, Fig.12is a diagram showing a method of aligning a rotor position according to the present disclosure. Fig.12 , the current application states of the rotor, stator and stator coils and the magnetic flux induced to the stator are shown.

[0156] In the drawings, the position and the angle represent the rotation position and the rotation angle of the rotor in the motor 121 .

[0157] According to the present disclosure, the process of aligning the rotor position when the motor 121 stops is described in more detail. If the speed of the motor decelerates to a predetermined reference speed so that the motor stops, the control unit 110 determines that the motor is in a stopped state and determines the final estimated position value (i.e., the rotor position finally estimated by the sensorless control logic at the motor stop determination point) as the alignment target position.

[0158] Subsequently, the control unit 110 determines an alignment start position by subtracting a predetermined alignment offset value from the alignment target position (ie, angle) in a direction opposite to the rotor rotation direction, and starts a control process for rotor position alignment from the alignment start position.

[0159] That is, the control unit 110 controls the inverter 111 to apply a d-axis current to the motor 121 of the air compressor 120 so that the rotor position is aligned with the alignment start position.

[0160] In addition, the control unit 110 controls the rotor position to change linearly within a predetermined period of time until the rotor position reaches the alignment target position from the alignment start position determined as described above.

[0161] Here, the control unit 110 controls the phases (U, V, and W phases) of the phase currents applied when the d-axis current is applied so that the rotor position can be linearly changed from the alignment start position to the alignment target position.

[0162] According to the present disclosure, changing the rotor position means further rotating the rotor, and linearly changing the rotor position means changing the rotor position from the alignment start position to the alignment target position at a predetermined change rate (or gradient).

[0163] In addition, according to the present disclosure, the alignment offset angle (i.e., the angle between the alignment start position and the alignment target position) is an angle that can be tuned and changed to an optimal angle, and the rotor rotation time from the alignment start position to the alignment target position is a time that can be pre-tuned and changed for use.

[0164] As described above, according to the present disclosure, if the alignment target position and the alignment start position are determined by the control unit 110, when the rotor is at the alignment start position, the control unit 110 starts applying the d-axis current to the motor 121. Thereafter, the d-axis current is continuously applied until the rotor reaches the alignment target position.

[0165] Here, the control unit 110 controls the phase of the current so that the rotor position changes linearly at a predetermined change rate (or angle gradient) until the rotor position changes from the alignment start position to the alignment target position.

[0166] During the motor stop as described above, the rotor position is continuously changed by the induced magnetic flux generated by the current applied to the stator coil, thereby performing rotor position alignment.

[0167] In addition, when the rotor position changes continuously, the rotor's rotational inertia decreases continuously. When the induced magnetic flux reaches the aligned target position, the rotor's rotational inertia is smaller than the braking force caused by the induced magnetic flux, thereby fixing the rotor to the desired position.

[0168] Reference above Figure 5 , Figure 6 and Fig. 7A In the described method, if the motor speed estimated by the sensorless control logic reaches the reference speed at a specific time point, a d-axis current for moving (rotating) the rotor to an aligned target position is applied at that time point. Here, starting from the time point when the motor speed reaches the reference speed, a d-axis current for controlling the rotor position to a fixed position (i.e., aligned target position) is applied without current phase control.

[0169] Therefore, in the above method, the moment of inertia of the rotor is sometimes greater than the braking force caused by the induced magnetic flux, so that the rotor further rotates half. Therefore, although it is not intentional, the rotor is positioned in the opposite direction to the induced magnetic flux. Therefore, the rotor is aligned to a position where the thrust and pull forces are balanced, and thus stops at this position, which is problematic.

[0170] In contrast, according to the present disclosure, current control that continuously changes the rotor position from the alignment start position to the alignment target position, that is, current phase control that forcibly rotates the rotor so as to linearly change the position within a predetermined period of time is performed.

[0171] Here, even when the rotor is continuously forced to rotate from the alignment start position to the alignment target position to be in a position opposite to the induced magnetic flux, the position of the induced magnetic flux is continuously moved, so that the rotor can continuously move instead of stopping in the direction where the thrust and pull forces are balanced. Therefore, the rotor can be accurately fixed at the alignment target position.

[0172] Fig.13 is a graph showing a control method for rotor position alignment according to the present disclosure, which is performed when the motor is stopped. Fig.13 As shown, the position alignment mark ① is defined to determine whether to perform the rotor position alignment.

[0173] In addition, if Fig.13 As shown, the forced rotor rotation flag ② is defined to determine whether the rotor is forcibly rotated to perform position alignment.

[0174] The time period for maintaining ① and ② is predetermined by the control unit 110. This time period can be adjusted according to the system.

[0175] Keep the mark ① Fig.13 Time period is the position alignment time, which is set so that the position alignment is performed in the position alignment time.

[0176] In addition, keep the mark ② Fig.13 Time period is the forced rotor rotation time, that is, the time period during which the rotor is forced to rotate. It is arranged so that the rotor position changes linearly and continuously from the alignment start position to the alignment target position.

[0177] like Fig.13 As shown, when the position alignment mark ① is true, the control unit 110 During this period, the d-axis current command is continuously generated and outputted, so that the inverter 111 can continuously apply a constant d-axis current to the motor 121 of the air compressor 120 .

[0178] In addition, when the forced rotor rotation flag ② is true, the control unit 110 During this period, the rotor is continuously and forcibly rotated from the alignment start position to the alignment target position, thereby linearly increasing the angle indicating the rotor position, such as Fig.13 shown.

[0179] Furthermore, if the forced rotor rotation flag ② is set to false, the rotor position (or angle) is kept constant when the position alignment flag ① is true.

[0180] refer to Fig.13 , it can be understood that the position is aligned with the time Defines the end point of the position alignment. After that, the position alignment mark ① becomes false, thereby completing the rotor position alignment at the end point.

[0181] Fig.14A and Fig. 14B is a flow chart showing the overall process of controlling a sensorless motor according to the present disclosure, Fig.15 1 is a flowchart more specifically illustrating a process of setting a position alignment mark and a forced rotor rotation mark in the control unit 110 in the process of controlling the sensorless motor according to the present disclosure.

[0182] Reference Fig.14A , Fig. 14B and Fig.15 , after driving the motor 121 of the air compressor 120 , the control unit 110 determines whether the motor speed command is set to a motor stop command value, for example, a command value of 0 rpm in S1 .

[0183] Here, if the motor speed command is set to a command value of 0 rpm, the control unit 110 reduces the speed of the motor by performing a typical sensorless deceleration control to stop the motor at S2 .

[0184] Thereafter, in S3, the motor speed estimated by the control unit 110 using the sensorless control logic is compared with a predetermined reference speed rpm. If the speed of the motor reaches and is lower than the reference speed, in S4, the motor is determined to be in a stopped state.

[0185] If it is determined that the motor 121 of the air compressor 120 is in the stopped state as described above, the control unit 110 starts the position alignment, such as Fig.15 As shown, align the position with mark ① (see Fig.13 ) is set to true, and the forced rotor rotation flag ② (see Fig.13 ) is set to true.

[0186] Afterwards, if it is determined that the motor is in the stop position, the control unit 110 sets the position alignment flag and the forced rotor rotation flag to true, and then starts counting time, such as Fig.15 shown.

[0187] After that, the control unit 110 keeps the forced rotor rotation flag ② as true until the time counted from the time point of the start position alignment reaches the forced rotor rotation time (See Fig.13 ).

[0188] If the counted time reaches the forced rotor rotation time, the forced motor rotation flag ② in the control unit 110 is set to false.

[0189] In addition, the control unit 110 maintains the position alignment mark ① (see Fig.13 ) is true until the counted time reaches the position alignment time (See Fig.13 ), even during the forced rotor rotation time Then the forced rotor rotation flag ② is set to false.

[0190] However, if Fig.15 As shown, at the forced rotor rotation mark ② at the forced rotor rotation time After that, if the counted time reaches the position alignment time, the position alignment flag ① in the control unit 110 is set to false.

[0191] Back to Fig.14A If the speed of the motor reaches the reference speed and the motor is determined to be in a stopped state, then in S5, the control unit 110 determines the final estimated position value (i.e., the rotor position finally estimated by the sensorless control logic at the motor stop determination point) as the alignment target position.

[0192] In addition, if the alignment target position is determined, then in S6 the control unit 110 determines the alignment start position by subtracting the alignment offset angle determined in the direction opposite to the rotor rotation direction from the alignment target position.

[0193] Thereafter, if the position alignment mark ① and the fourth rotor rotation mark ② are simultaneously set to true, then in S7, the control unit 110 starts to control the inverter 111 to apply the d-axis current for rotor position alignment to the motor 121 of the air compressor 120 so as to move the rotor position to the alignment target position.

[0194] Afterwards, when the position alignment mark ① is true (i.e., at the predetermined position alignment time During the period of time, the control unit 110 continuously applies the d-axis current (ie, generates and outputs the PWM signal) to the motor 121. When the position alignment time has passed from the motor stop determination point After that, the position alignment mark ① becomes false (i.e., Fig.13 When the position in the d-axis direction is aligned with the end point, the control unit 110 stops applying the d-axis current.

[0195] In addition, if the position alignment mark ① and the fourth rotor rotation mark ② are set to true at the same time and the d-axis current is applied, the control unit 110 performs current phase control to linearly change the rotor position from the alignment start position to the alignment target position while the forced rotor rotation mark ② is true.

[0196] That is, during the forced rotor rotation time determined from the motor stop determination point During the operation, the phases (e.g., U, V, and W phases) of the phase currents applied to the motor 121 via the inverter 111 are controlled so that the rotor position can be changed linearly (i.e., the alignment angle of the rotor is changed between Fig.13 changes continuously and linearly in the

[0197] Therefore, during the forced rotor rotation time During this period, the forced rotor rotation flag ② remains true, and in S7, the rotor is forcibly rotated from the alignment start position to the alignment target position. During the forced rotation period, the rotor position changes at a predetermined rate of change (e.g., Fig.13 The predetermined angle gradient in the curve changes.

[0198] Here, the rate of change of the rotor position from the alignment start position to the alignment target position can be defined by Formula 1:

[0199] Rotor position change rate = (alignment deviation angle) / (forced rotor rotation time) (Formula 1)

[0200] In Formula 1, the alignment offset angle and the forced rotor rotation time are values ​​previously known by the control unit 110. Therefore, the rate of change of the rotor position defined by Formula 1 may be a value that the control unit 110 already knows or a value that can be calculated by the control unit 110.

[0201] In addition, since the rotor position changes at a rate of change from the alignment start position to the alignment target position, the rotor position can be estimated based on the alignment start position and the rate of change of the rotor position within the time period from the forced rotor rotation start point (i.e., the time point when both marks are set to true) serving as the motor stop determination point to the forced rotor rotation end point (i.e., the time point when the forced rotor rotation mark is converted to false).

[0202] That is, the position of the rotor between the alignment start position and the alignment target position can be estimated by Formula 2.

[0203] Estimated position = alignment start position + (change rate × count time) (Formula 2)

[0204] Here, the “estimated position” is an estimated value of the rotor position changing from the alignment start position to the alignment target position, and the “change rate” is the change rate of the rotor position defined by the above formula 1.

[0205] Furthermore, the "counting time" refers to a period of time counted from a point in time when both flags (ie, the position alignment flag ① and the forced rotor rotation flag ②) are set to true because the control unit 110 determines that the motor is in a stopped state.

[0206] In the present disclosure, the rotor position estimated by Equation 2 may be used by the control unit 110 to control the rotor position.

[0207] Fig.16 is a flow chart showing a process of estimating the rotor position during motor stop according to the present disclosure. Fig.16 As shown, before the driving of the motor 121 stops, neither the position alignment mark ① nor the forced rotor rotation mark ② is true. Here, the estimated position of the rotor is an estimated position value obtained by the sensorless control logic.

[0208] In addition, if the motor 121 stops and the position alignment mark ① and the forced rotor rotation mark ② are simultaneously set to true, the control unit 110 controls the phase of the current applied to the motor 121 of the air compressor 120 via the inverter 111 so that the rotor is forced to rotate from the alignment start position to the alignment target position. Here, the phase of the current is controlled so that the rotor position changes linearly at a predetermined rate of change.

[0209] Therefore, if both flags are true, such as Fig.16 As shown, the rotor position is estimated as the value of "alignment start position + (change rate × counting time)", as shown in Formula 2.

[0210] However, if the time counted from the motor stop determination point reaches the forced rotor rotation time and the forced rotor rotation flag ② is converted to false, when the position alignment flag ① is true, the estimated position of the rotor is aligned with the target position.

[0211] After that, return to FIG. 14. After starting the current phase control for forcing the rotor to rotate, in S8, the control unit 110 determines whether the time counted from the motor stop determination point has reached the predetermined forced rotor rotation time. If the counted time has reached the forced rotor rotation time, then in S9, the control unit 110 stops the current phase control for linearly changing the rotor position, but continues to apply the d-axis current for fixing and maintaining the rotor position in the alignment target position (see Fig. 14B ).

[0212] Here, the forced rotor rotation flag ② in the control unit 110 is set to false, and the position alignment flag ① is maintained to true.

[0213] Afterwards, in S10, the control unit 110 determines whether the counted time after starting the motor stop determination and applying the d-axis current has reached the position alignment time. If the counted time has reached the position alignment time, the control unit 110 stops applying the d-axis current and PWM control, thereby completing the rotor position alignment in S11.

[0214] Thereafter, if a speed command for driving the motor is generated in S12 and the driving of the motor is resumed, the control unit 110 starts sensorless control of the motor 121 by setting the alignment target position as an initial position in S13 .

[0215] Fig.17 1 is a flowchart showing a process of determining a d-axis current command and a q-axis current command according to a position alignment mark in a control unit of the present disclosure. If the position alignment mark ① in the control unit 110 is true, a d-axis current command is generated. When the control unit 110 performs PWM output according to the current command, the d-axis current can be applied to the motor 121.

[0216] As described above, the method of controlling a sensorless motor of an air compressor according to the present disclosure can improve the speed control response in the initial stage of driving the air compressor, thereby improving the air supply response of the fuel cell stack, the stack output response, and the vehicle acceleration performance.

[0217] In addition, when the driving of the motor is resumed after stopping, speed control reliability can be obtained in the early stage of driving. Therefore, it is possible to prevent the torque from being applied at an unnecessary position, thereby improving the driving efficiency of the motor and the fuel efficiency of the vehicle.

[0218] Furthermore, the method of controlling a sensorless motor of an air compressor according to the present disclosure may overcome the problem of rotor alignment failure, wherein the rotor may not be aligned to the expected position when the rotor position is aligned with the final position estimated by the sensorless control logic when the motor is stopped.

[0219] That is, the present disclosure can prevent position alignment failures that may occur occasionally, such as Figure 8 shown.

[0220] Although exemplary embodiments of the present disclosure have been described for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the disclosure as disclosed in the accompanying claims.

Claims

1. A method for controlling a sensorless motor of an air compressor, the method The following steps are involved: Speed ​​control is performed by the controller to stop the motor of the air compressor; During the speed control, the controller determines that the motor is in a stopped state according to the motor speed estimated by the sensorless control logic; The controller determines a rotor alignment target position at a time point when the motor is determined to be in a stopped state, wherein the rotor alignment target position is determined by a rotor position of the motor finally estimated by the sensorless control logic; The controller determines an alignment start position from the alignment target position according to a predetermined alignment offset angle; The controller changes the rotor position of the motor from the determined alignment start position to the alignment target position; as well as In a state where the rotor position of the motor is aligned with the alignment target position, the controller performs sensorless control by setting the alignment target position at an initial position in response to a motor drive request to drive the motor, The controller determines the alignment start position by subtracting the alignment offset angle from a predetermined alignment target position in a direction opposite to a rotation direction of the rotor.

2. The method according to claim 1, in, When the motor speed estimated by the sensorless control logic reaches a predetermined reference speed, the controller determines that the motor is in a stopped state.

3. The method according to claim 2, in, The controller determines the reference speed as a speed greater than 0 rpm.

4. The method according to claim 1, in, When changing the rotor position, the controller controls the current applied to the motor to forcibly rotate the rotor within a predetermined forced rotor rotation time so that the rotor position of the motor changes from the alignment start position to the alignment target position.

5. The method according to claim 4, in, The current includes a phase current, and the controller controls the phase of the phase current applied to the motor so that a rotor position of the motor changes from the alignment start position to the alignment target position within the forced rotor rotation time.

6. The method according to claim 5, in, The controller controls the phase of a phase current applied to the motor so that a rotor position of the motor changes linearly at a predetermined rate of change.

7. The method according to claim 5, in, When changing the rotor position, the controller: Applying a d-axis current for moving the rotor position of the motor from the alignment start position to the alignment target position within a predetermined position alignment time starting from the time point when the motor is determined to be in a stopped state; and After the forced rotor rotation time has elapsed from a time point when the motor is determined to be in a stopped state, the phase of the phase current is stopped from being controlled during the application of the d-axis current before the position alignment time has elapsed.

8. The method according to claim 4, in, The controller applies a current to the motor so that a rotor position of the motor changes from the alignment start position to the alignment target position at a predetermined change rate.

9. The method according to claim 4, in, When changing the rotor position, the controller applies a d-axis current to change the rotor position of the motor from the alignment start position to the alignment target position within a predetermined position alignment time.

10. The method according to claim 1, in, When changing the rotor position, the controller applies a d-axis current to change the rotor position of the motor from the alignment start position to the alignment target position within a predetermined position alignment time.

11. The method according to claim 1, in, The air compressor includes an air compressor for a fuel cell system to supply air to a fuel cell stack.