Control device for synchronous motor

Abstract

A control device (1) for a synchronous motor (200) controls a synchronous motor (200) using the phase (rotor position estimation value θest) of a dc-axis, where the rotor position of the synchronous motor (200) is defined as a d-axis, an axis advanced by 90 degrees from the d-axis is defined as a q-axis, an axis having a phase difference of θe from the d-axis is defined as the dc-axis, and an axis advanced by 90 degrees from the dc-axis is defined as a qc-axis. A processor of the control device (1) for the synchronous motor (200) calculates θe by using the voltages (voltage commands vdc, vqc) of the dc axis and the qc axis, the angular velocity (electric angular velocity estimation value ω^) of the dc axis, a characteristic value (resistance value R, static q-axis inductance Lq) of the synchronous motor (200), and the magnetic fluxes (Φdc, Φqc) of the dc axis and the qc axis, without using an approximate value for the angular velocity (electric angular velocity true value ω) of the d-axis. Thus, even if the synchronous motor (200) suddenly accelerates or decelerates, the accuracy of the phase difference θe is ensured.

Description

Synchronous motor control device

[0001] The present invention relates to a control device for a synchronous motor.

[0002] Electrically powered vehicles, such as hybrid vehicles and electric vehicles, often use permanent magnet synchronous motors (hereinafter referred to as synchronous motors) with permanent magnets in their rotors as their drive sources. To drive and control such synchronous motors, an inverter, which is a power converter, is generally used. Current vector control is commonly known as a method for controlling the torque of a synchronous motor using an inverter. To perform current vector control, it is necessary to accurately grasp the magnetic pole position of the rotor of the synchronous motor. As one means for grasping this magnetic pole position, a resolver, which is a position sensor, is often provided in the synchronous motor.

[0003] However, providing a resolver in a synchronous motor increases the number of parts in the system, increases the system size, and increases costs.

[0004] Furthermore, the resolver has an output winding wound thereon for outputting a position signal, and there is a risk of failure due to breakage or short circuit of the winding, which also leads to a decrease in the reliability of the system.

[0005] Therefore, position sensorless vector control is known as a technique for driving a synchronous motor without using a sensor such as a resolver.

[0006] Position sensorless vector control performs current vector control by obtaining rotor position information based on information such as the voltage applied to a synchronous motor and the current that flows due to the voltage. This eliminates the need for position sensors such as resolvers, preventing system size, increased costs, and reduced reliability. Methods for estimating rotor position are broadly divided into methods based on the position dependency of the inductance of a synchronous motor and methods based on the induced voltage of a synchronous motor. It is generally known that the former is used in the low-speed range where the induced voltage of a synchronous motor is zero or small, and that position sensorless control is performed by switching to the latter method in the medium- to high-speed range where the induced voltage becomes high.

[0007] Patent Document 1 discloses a method based on extended electromotive force (EEMF) as a position sensorless control method for a synchronous motor in the medium-to-high speed range. Patent Document 2 also discloses a method based on magnetic flux that enables highly accurate position estimation regardless of the speed range of the synchronous motor.

[0008] JP 2001-251889 A JP 2009-284558 A

[0009] In Patent Document 1, the rotor position of a synchronous motor is defined as the d-axis, the axis perpendicular to the d-axis is defined as the q-axis, a coordinate system in which there is a phase difference between these dq-axes and θe is defined as the dc-qc-axes, and information on θe is obtained by estimating the extended induced voltage, thereby implementing position sensorless control. However, there was a problem in that the extended induced voltage could not be detected accurately unless the speed of the synchronous motor was in the medium to high speed range.

[0010] In contrast to this, in Patent Document 2, an observer is configured to estimate magnetic flux using a voltage command value, current detection value, and current estimation value of a synchronous motor based on the voltage equation of the synchronous motor, and the magnetic flux estimation value obtained by the observer is used to estimate position. In a synchronous motor with a permanent magnet in the rotor, flux linkage due to the permanent magnet is always generated regardless of speed, and by utilizing this, a position sensorless system can be achieved over a wide speed range.

[0011] However, when current flows through the windings of a synchronous motor, the current generates magnetic flux, which causes the magnetic flux linking the stator of the synchronous motor to become too dense and reduce inductance (magnetic saturation). As a result, the pole placement of the observer changes, which can degrade the observer's estimation performance (response and accuracy). While it is possible to take magnetic saturation into account by reflecting inductance changes due to magnetic saturation in the observer's estimation formula according to the current detection value, there is a problem with the computational complexity involved in performing pole placement calculations in real time.

[0012] In addition to inductance changes due to magnetic saturation, inductance can also be classified as dynamic inductance or static inductance. The former is expressed by the gradient of magnetic flux near the current operating point, while the latter represents the gradient of magnetic flux between the current operating point and the origin. The extended electromotive force method typified by Patent Document 1 requires the use of quantities different from the d-axis magnetic flux and the q-axis magnetic flux, such as the product of the q-axis static inductance and the d-axis current, or the product of the d-axis dynamic inductance and the q-axis current, making it difficult to consider static and dynamic inductance. Even in a configuration using an observer such as Patent Document 2, separating static and dynamic inductances poses the problem of complexity.

[0013] Furthermore, in both Patent Document 1 and Patent Document 2, it is necessary to assume that the speed of the dc axis (=estimated speed value) is equal to the speed of the d axis (true speed value), and in areas where this assumption does not hold, such as during sudden acceleration or deceleration of the synchronous motor, position estimation errors may occur, causing the control system to become unstable.

[0014] An object of the present invention is to provide a control device for a synchronous motor that ensures the accuracy of the phase difference θe (axis error) even when the synchronous motor is suddenly accelerated or decelerated.

[0015] In order to achieve the above object, the present invention provides a synchronous motor control device that defines the rotor position of a synchronous motor as the d-axis, the axis 90 degrees ahead of the d-axis as the q-axis, the axis having a phase difference θe with the d-axis as the dc-axis, and the axis 90 degrees ahead of the dc-axis as the qc-axis, and controls the synchronous motor using the phase of the dc-axis, and has a processor that calculates the phase difference θe using the voltages on the dc-axis and the qc-axis, the angular velocity of the dc-axis, characteristic values ​​of the synchronous motor, and the magnetic flux of the dc-axis and the qc-axis, without using an approximation of the angular velocity of the d-axis.

[0016] According to the present invention, the accuracy of the phase difference θe (axis error) is ensured even when the synchronous motor is suddenly accelerated or decelerated. Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiments.

[0017] FIG. 10 is a diagram showing an overall image of a drive device (control device) for a synchronous motor according to a first embodiment of the present invention. FIG. 11 is a block diagram showing main functions of a position estimator according to the first embodiment of the present invention. FIG. 12 is a diagram showing main functions of a dc-qc axis magnetic flux calculator according to the first embodiment of the present invention. FIG. 13 is a block diagram showing main functions of a modified position estimator. FIG. 14 is a diagram showing main functions of a speed calculator according to the first embodiment of the present invention. FIG. 15 is a diagram showing a modified dc-qc axis magnetic flux calculator. FIG. 16 is a diagram showing a modified axis error calculator. FIG. 17 is a configuration diagram of a drive device (control device) for a synchronous motor according to a second embodiment of the present invention. FIG. 18 is a configuration diagram of a flux control unit shown in FIG. 19. FIG. 19 is a configuration diagram of a position estimator shown in FIG.

[0018] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The following description and drawings are examples for explaining the present invention, and some omissions and simplifications have been made as appropriate for clarity of explanation. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.

[0019] 1 is a configuration diagram of a drive device for a synchronous motor according to this embodiment. The drive device for a synchronous motor according to this embodiment can take into account changes in inductance due to magnetic saturation, as well as static and dynamic inductance, and further has a position estimation technique that does not require the assumption that the estimated speed and the true speed value are equal.

[0020] The drive device for a synchronous motor includes a position estimation unit 100A, a synchronous motor 200, a power converter 300, a current command calculation unit 400, a current detection unit 500, a current control unit 600, a coordinate inverse conversion unit 700A, and a coordinate conversion unit 700B, and drives the synchronous motor 200. In this embodiment, the position estimation unit 100A, the current command calculation unit 400, the current control unit 600, the coordinate inverse conversion unit 700A, and the coordinate conversion unit 700B are realized by a control device 1. The control device 1 is, for example, a microcomputer, and is composed of a memory (storage device), a CPU (processor), an input / output circuit (communication device), etc. The CPU functions as each unit by executing a program stored in the memory.

[0021] The synchronous motor 200 is a permanent magnet type synchronous motor, and is configured by using a permanent magnet (ferromagnetic material) in the rotor and an armature winding in the stator. Note that in this embodiment, no position sensor is provided to detect the magnetic pole position of the synchronous motor 200, and instead the magnetic pole position is estimated by a position estimator 100A.

[0022] The power converter 300 is, for example, an inverter, and performs pulse width modulation (PWM) on the voltage commands Vu*, Vv*, and Vw* from the coordinate inverse transformation unit to control the on / off of the semiconductor switch elements of the power converter when driving the synchronous motor 200. In this way, the voltages Vu, Vv, and Vw are applied to the synchronous motor 200, thereby driving the synchronous motor 200.

[0023] The current command calculation unit 400 receives a torque command T* as a drive command for the synchronous motor 200 from a higher-level control device (not shown), calculates current commands idc* and iqc*, which are target values ​​of the current to be flowed to the synchronous motor 200, and transmits them to the current control unit 600.

[0024] The current detection unit 500 is configured by current sensors 500u, 500v, and 500w that detect three-phase currents flowing through the synchronous motor 200. The current sensors 500u, 500v, and 500w are arranged in the respective phases of the synchronous motor 200.

[0025] The current detection unit 500 detects three-phase currents Iu, Iv, and Iw and transmits signals to the coordinate conversion unit 700B. While the current sensors 500u, 500v, and 500w are shown as being disposed in each of the three phases of the synchronous motor 200, they may be disposed in only two phases (e.g., the u-phase and the v-phase) by taking advantage of the fact that the sum of the three-phase AC currents is zero. Alternatively, the three-phase current of the synchronous motor 200 may be obtained from the current flowing through a DC bus (not shown) on the input side of the power converter 300. Such a configuration allows the number of current sensors to be reduced, thereby achieving cost reduction.

[0026] The current control unit 600 receives current commands idc* and iqc* from the current command calculation unit and current values ​​idc and iqc from the coordinate conversion unit 700B, and adjusts and controls voltage commands vdc and vqc so that idc* and idc, and iqc* and iqc, respectively. The current control unit 600 can be configured in various ways, such as by combining commonly known proportional-integral control, control that compensates for the armature reaction of a synchronous motor, and control that compensates for the dead time of the power converter 300.

[0027] The coordinate inverse transformation unit 700A receives voltage commands vdc and vqc, which are signals from the current control unit, and performs conversion processing to convert them into three-phase AC voltage commands Vu*, Vv*, and Vw* based on the estimation result θest of the position estimation unit 100A.

[0028] The coordinate conversion unit 700B converts the three-phase AC currents Iu, Iv, and Iw obtained from the current detection unit 500 into currents idc and iqc in the dc-qc coordinate system based on θest, which is the estimation result of the position estimation unit 100A.

[0029] The position estimation unit 100A is a functional block that brings about the main effect of this embodiment, and receives the output signals vdc and vqc of the current control unit 600 and the output signals idc and iqc of the coordinate conversion unit 700B, and based on these, calculates the rotor position of the synchronous motor 200 as its estimated value θest.

[0030] Next, the basic principle of the present invention will be explained. Equations (1) and (2) are voltage equations on the dc-qc coordinate system using the extended induced voltage, which is the conventional technique.

[0031]

[0032]

[0033] where R is the winding resistance of the synchronous motor, Ld is the static d-axis inductance of the synchronous motor, Lq is the static q-axis inductance of the synchronous motor, Ldh is the dynamic d-axis inductance of the synchronous motor, Lqh is the dynamic q-axis inductance of the synchronous motor, ω is the angular velocity of the d-axis (the true value of the electrical angular velocity of the synchronous motor), ω^ is the angular velocity of the dc-axis (the estimated value of the electrical angular velocity of the synchronous motor), Eex is the extended induced voltage, p is the differential operator, and p() is the time derivative in parentheses.

[0034] As shown in equations (1) and (2), the extended induced voltage components of the dc and qc axes are expressed by trigonometric functions using the axis error θe. Therefore, by calculating the extended induced voltage components of the dc and qc axes, it is possible to estimate the axis error θe.

[0035] However, equations (1) and (2) require ω, which is the true value of the electrical angular velocity. In position sensorless control, since the true value ω is unknown, the extended induced voltage components of the dc axis and qc axis must be estimated using the approximation that ω is equal to ω^, which is the estimated value of ω. This approximation holds when the synchronous motor is in a steady state with a constant rotation speed, but it does not hold during transient operation such as a sudden change in rotation speed, which can result in errors in position estimation.

[0036] Furthermore, dynamic inductance and static inductance are mixed in equations (1) and (2), and if these are taken into consideration in the methods described in Patent Documents 1 and 2 in order to improve the accuracy of position estimation, the calculations become complicated.

[0037] Furthermore, while the product of inductance and current generally results in magnetic flux, the second term of equation (2) contains the product of the dynamic d-axis inductance and the qc-axis current, and the third term contains the product of the static q-axis inductance and the dc-axis current, and since these are products of mutually orthogonal axes, they are physical quantities separate from the magnetic flux of the dc-axis or qc-axis. This creates a problem that makes it difficult to configure a magnetic flux observer such as that disclosed in Patent Document 2.

[0038] Therefore, in this embodiment, these problems are solved by the following means.

[0039] In the dq coordinate system, which is the coordinate system of a synchronous motor, if magnetic fluxes on the d and q axes are used instead of d and q axis currents, the following equations (3) and (4) are obtained.

[0040]

[0041]

[0042] Here, Φd is the d-axis magnetic flux (magnetic flux generated by the d-axis current and the permanent magnet), Φq is the q-axis magnetic flux (magnetic flux generated by the q-axis current), and Φm is the flux linkage due to the permanent magnet.

[0043] Equation (3) can be transformed into the following equation (5).

[0044]

[0045] Equation (5) is obtained by modifying equation (3) so that the coefficient of Φd in the first term on the right side becomes (R / Lq).

[0046] When the d-axis and q-axis voltage equations expressed based on the magnetic fluxes of equations (5) and (4) are converted from the dq coordinate system to dc-qc coordinates with a phase difference of θe, the following equations (6) and (7) are obtained.

[0047]

[0048]

[0049] As shown in equations (6) and (7), the term for the axis error θe appears on each of the dc axis and the qc axis, and the axis error can be estimated in the same way as the conventional extended electromotive force.

[0050] Furthermore, since the term ω, which is the true value of the electrical angular velocity, is not included and is expressed only by the estimated value ω^, it is possible to estimate the position with high accuracy even when transient velocity fluctuations occur.

[0051] Furthermore, since the right-hand sides of equations (6) and (7) are expressed using only magnetic flux containing information on inductance (only static inductance), there is no need to distinguish between dynamic and static inductance, which makes it possible to improve the accuracy of position estimation and simplify calculations.

[0052] From the relationship between equations (6) and (7), the axis error θe can be expressed using the following equation (8).

[0053]

[0054] Note that arctan is an arctangent function. The required characteristics of the synchronous motor 200 are two types: the resistance value R and the static q-axis inductance Lq. Since the d-axis inductance is not required, the synchronous motor 200 is robust with respect to the electrical constants.

[0055] Next, an embodiment using the basic principle of the present invention described above will be described with reference to Fig. 2. Fig. 2 is a configuration diagram of a position estimation unit 100A according to the first embodiment.

[0056] The position estimation unit 100 A includes a dc-qc axis magnetic flux calculation unit 110 , an axis error calculation unit 120 , a speed calculation unit 130 , and a position calculation unit 140 .

[0057] The dc-qc axis magnetic flux calculation unit 110 receives the dc-qc axis currents idc and iqc as input values ​​and calculates the dc-qc axis magnetic fluxes Φdc and Φqc.

[0058] The axis error calculation unit 120 calculates the axis error θe using as input values ​​vdc, vqc which are dc-qc axis voltages, Φdc, Φqc which are dc-qc axis magnetic fluxes, and ω^ which is an estimated electrical angular velocity value.

[0059] The speed calculation unit 130 receives the axis error θe as an input value and calculates ω^, which is an estimated value of the electrical angular speed.

[0060] The position calculation unit 140 receives the electrical angle estimate ω^ as an input value and calculates the rotor position estimate θest of the synchronous motor 200 .

[0061] Hereinafter, each functional block constituting the position estimation unit 100A will be described.

[0062] 3 shows the dc-qc-axis magnetic flux calculation section 110 in the position estimation section 100 A. The dc-qc-axis magnetic flux calculation section 110 includes a dc-axis magnetic flux calculation section 111 and a qc-axis magnetic flux calculation section 112 .

[0063] The dc-axis magnetic flux calculation unit 111 has the characteristics of the dc-axis magnetic flux Φdc with respect to the dc-axis and qc-axis currents idc and iqc stored in a table, etc. The dc-axis magnetic flux calculation unit 111 references and outputs the value of the dc-axis magnetic flux Φdc according to the values ​​of the dc-axis and qc-axis currents idc and iqc.

[0064] The qc-axis magnetic flux calculation unit 112 has the characteristics of the qc-axis magnetic flux Φqc with respect to the dc-axis and qc-axis currents idc and iqc stored in a table, etc. The qc-axis magnetic flux calculation unit 112 references and outputs the value of the qc-axis magnetic flux Φqc according to the values ​​of the dc-axis and qc-axis currents idc and iqc.

[0065] The dc-axis magnetic flux calculation unit 111 and the qc-axis magnetic flux calculation unit 112 may create tables using characteristic analysis values ​​of the synchronous motor 200 and experimental data obtained in advance.

[0066] As described above, the dc-axis magnetic flux calculation unit 111 and the qc-axis magnetic flux calculation unit 112 include inductance information because they are characteristics that indicate the magnetic flux relative to the current. Therefore, by using these magnetic fluxes, it is not necessary to distinguish between dynamic and static inductance, which simplifies calculations and improves the accuracy of position estimation.

[0067] The axis error calculation unit 120 calculates the axis error θe, which is the phase difference between the d-axis and dc-axis, based on the above-mentioned equation (8). In this embodiment, the inputs to the axis error calculation unit 120 are the dc-axis and qc-axis voltages vdc, vqc, the dc-axis and qc-axis magnetic fluxes Φdc, Φqc, and the estimated speed ω^. This is because the time derivative terms p(Φdc) and p(Φqc) are omitted to more simply express equation (8). This makes it possible to simplify the calculation. In other words, the axis error θe can be calculated using the following equation (9) instead of equation (8):

[0068]

[0069] However, because equation (9) does not take into account the time derivative of the magnetic flux, dynamic inductance is not taken into account. This may result in a transient position estimation error, so calculation may be performed as in equation (8). When calculation is performed using the configuration of equation (8), as shown in FIG. 4, a differential calculation unit 150 is provided that differentiates the dc-axis and qc-axis magnetic fluxes Φdc and Φqc, and the results of the differential calculation are sent to the axis error calculation unit 120 as pΦdc and pΦqc. Note that the difference from the position estimation unit 100A in FIG. 2 is the presence or absence of the differential calculation unit 150, so blocks having equivalent functions to those in FIG. 2 are denoted by the same reference numerals in FIG. 4.

[0070] The calculation in the differential calculation unit 150 may be a pure time differential, or a filter may be provided for the differential value to prevent noise amplification. In that case, phase lead compensation may be applied to compensate for the delay caused by the filter.

[0071] Furthermore, the dc-axis magnetic flux Φdc and the qc-axis magnetic flux Φqc can be calculated based on the voltage equations, in addition to the above-mentioned characteristic table for current. The voltage equations on the dc-qc axes can also be expressed by the following equations (10) and (11).

[0072]

[0073]

[0074] Therefore, from the relationship between equations (10) and (11), the dc-axis magnetic flux Φdc and the qc-axis magnetic flux Φqc can be obtained from the following equations (12) and (13).

[0075]

[0076]

[0077] Since the calculations of Equations (12) and (13) are composed of simple arithmetic operations, the calculation load on the processor can be reduced compared to when the magnetic flux characteristic table for current described above is provided. Furthermore, as in the case of providing a table, the magnetic flux can be derived while taking into account static and dynamic inductance. In this case, the dc-qc-axis magnetic flux calculation unit 110 is configured to calculate the dc-axis magnetic flux Φdc and the qc-axis magnetic flux Φqc using the dc-axis voltage vdc and the qc-axis voltage vqc, and the dc-axis current idc and the qc-axis current iqc as inputs, as shown in FIG. 6 .

[0078] The axis error calculation section 120 may also have the configuration shown in Fig. 7. By substituting Φqc in equation (13) for Φqc in equation (6), the term for the axis error θe in equation (6) can be expressed by the following equation (14).

[0079]

[0080] Similarly, by substituting Φdc in equation (12) for Φdc in equation (7), the term for the axis error θe in equation (7) can be expressed by the following equation (15).

[0081]

[0082] From the relationship between equations (14) and (15), the axis error θe can be calculated by the following equation (16).

[0083]

[0084] Therefore, the dc-axis voltage vdc and the qc-axis voltage vqc are not required to calculate the axis error θe. This has the advantage of being less susceptible to the influence of output voltage errors due to dead time in the power converter, etc. Furthermore, the only characteristic of the synchronous motor used is the static q-axis inductance Lq, which improves robustness against synchronous motor parameters. When using equation (16), the dc-axis magnetic flux Φdc and the qc-axis magnetic flux Φqc can be calculated using the method using the table described above or a method based on the voltage equation.

[0085] 5 is a diagram showing the configuration of the speed calculation unit 130 in the position estimation unit 100A. The speed calculation unit 130 receives the position error θe as an input value and includes a subtractor 131, a proportional unit 132, an integrator 133, and an adder. A command value CMD is set to set the position error θe to a predetermined value, and the estimated speed ω^ is calculated by making the position error θe follow the CMD.

[0086] The subtractor 131 calculates the deviation between the command value CMD and the axis error θe, and sends it as an input value to the proportional unit 132 and the integrator 133 .

[0087] The proportional unit 132 amplifies the deviation between the command value CMD and the axis error θe and sends it to the adder 134 .

[0088] The integrator 133 time-integrates and amplifies the deviation between the command value CMD and the axis error θe, and sends the amplified deviation to the adder 134 .

[0089] The adder 134 receives the output signals from the proportional unit 132 and the integrator 133 as input values, and calculates the speed estimate value ω^ by adding up the outputs of each.

[0090] When CMD and θe coincide with each other, the speed calculation unit 130 controls θe to follow CMD so that the estimated speed value ω^ and the true speed value ω coincide with each other.

[0091] CMD is, for example, zero. If CMD is set to zero and θe is made to follow zero, the dc axis and the d axis coincide, making it possible to perform position sensorless drive equivalent to control using rotor position (= d axis) detection by a resolver.

[0092] CMD may use a value other than zero. For example, there is a method of defining a coordinate system so that the d-axis and q-axis currents always have a maximum torque per ampere (MTPA). When driving the synchronous motor 200 in this coordinate system, it is sufficient to set CMD to a value that results in MTPA, and CMD does not need to be constant. Alternatively, CMD may be set to a coordinate system known in the art.

[0093] Furthermore, although the speed calculation unit 130 in this embodiment is configured to make the axis error θe follow the CMD by proportional-integral control, it does not necessarily have to be proportional-integral control. Various known technology configurations can be used, such as providing differential control or feedforward compensation.

[0094] The position calculation unit 140 calculates an estimated value θest of the rotor position of the synchronous motor 200 by integrating the estimated speed ω^.

[0095] In this embodiment, a position estimation unit 100A in a synchronous motor control device is provided with a dc-qc axis magnetic flux calculation unit 110 that calculates dc-qc axis magnetic fluxes Φdc, Φqc. An axis error calculation unit 120 is configured to receive as inputs dc-qc axis voltages vdc, vqc, dc-qc axis magnetic fluxes Φdc, Φqc, and an estimated speed ω^, and calculates an axis error θe from a voltage equation based only on magnetic flux.

[0096] The dc-qc axis magnetic flux calculation unit is configured to have a relationship between the dc-axis and qc-axis currents i dc and i qc and the dc-axis and qc-axis magnetic flux, making it possible to easily consider inductance changes due to magnetic saturation. Furthermore, by providing a differential calculation unit 150, it is possible to consider not only static inductance but also dynamic inductance.

[0097] Furthermore, the axis error calculation unit 120 is configured to calculate the axis error θe based on a voltage equation on dc-qc coordinates expressed in magnetic flux without using current. This allows the axis error θe to be calculated without using the approximation that the true speed value and the estimated speed value are equal, as was previously necessary, while taking into account the aforementioned magnetic saturation and dynamic / static inductance. This makes it possible to estimate the position with high accuracy even when there is a change in the characteristics of the synchronous motor due to a change in current or a sudden change in rotation speed.

[0098] Therefore, by using the synchronous motor control device of this embodiment as a drive motor for an electric vehicle, a position sensorless system can be realized, resulting in lower costs and a smaller system size.Since the position estimation accuracy is improved, the control performance of the synchronous motor (torque accuracy, torque response, etc.) is improved, and a comfortable ride can be provided to passengers.

[0099] 8 is a configuration diagram of a drive device for a synchronous motor according to the second embodiment. The differences from the first embodiment are that a magnetic flux command calculation unit 800 replaces the current command calculation unit 400, a magnetic flux control unit 900 replaces the current control unit 600, a position estimation unit 100B, and a dc-qc axis magnetic flux calculation unit 110 in the position estimation unit 100A are provided subsequent to a coordinate conversion unit 700B.

[0100] The main function of the dc-qc-axis magnetic flux calculation unit 110 is the same as in the first embodiment, and it calculates the dc-axis magnetic flux Φdc and the qc-axis magnetic flux Φqc using the dc-axis current idc and the qc-axis current iqc obtained from the coordinate conversion unit 700B as inputs. Note that the calculation method of the dc-qc-axis magnetic flux calculation unit 110 in Figure 8 shows the configuration shown in Figure 3 as an example, as in the first embodiment. Alternatively, it may be configured to add the dc-axis voltage vdc and the qc-axis voltage vqc as inputs and use Figure 6 or equations (12) and (13).

[0101] The magnetic flux command calculation unit 800 receives a torque command T* as a drive command for the synchronous motor 200 from a higher-level control device (not shown), calculates a dc-axis magnetic flux command Φdc* and a qc-axis magnetic flux command Φqc*, which are target values ​​of the magnetic flux to be generated in the synchronous motor 200, and sends them to the magnetic flux control unit 900.

[0102] The magnetic flux control unit 900 receives the magnetic flux commands Φdc* and Φqc* from the magnetic flux command calculation unit 800 and the magnetic fluxes Φdc and Φqc from the dc-qc axis magnetic flux calculation unit 110, and adjusts and controls the voltage commands vdc and vqc so that Φdc* and Φdc, and Φqc* and Φqc, respectively, match. By controlling based on the magnetic flux, dynamic inductance and static inductance can be taken into account in the drive control, and the torque responsiveness and torque accuracy of the drive control of the synchronous motor 200 can be improved. Furthermore, the magnetic flux control unit 900 outputs p(Φdc) and p(Φqc), which are the differential values ​​of Φdc and Φqc, based on the deviation between Φdc* and Φdc and the deviation between Φqc* and Φqc, which will be described later, and transmits these to the position estimation unit 100B.

[0103] The position estimation unit 100B receives Φdc and Φqc, which are the outputs of the dc-qc axis magnetic flux calculation unit 110, and vdc and vqc, p(Φdc) and p(Φqc), which are the outputs of the magnetic flux control unit 900, and calculates the rotor position of the synchronous motor 200 as its estimated value θest based on these.

[0104] Other than that, the coordinate inverse conversion unit 700A, the coordinate conversion unit 700B, the power converter 300, and the synchronous motor 200 have the same configuration as in the first embodiment.

[0105] 9 is a diagram showing the configuration of a magnetic flux control unit 900 according to this embodiment. The magnetic flux control unit 900 includes a dc-axis magnetic flux control unit 910 and a qc-axis magnetic flux control unit 920. These units have similar configurations, so only the dc-axis magnetic flux control unit 910 is shown in detail. The dc-axis magnetic flux control unit 910 receives a dc-axis magnetic flux command Φdc* and a dc-axis magnetic flux Φdc as inputs, and outputs a dc-axis voltage vdc and p(Φdc), which is the differential value of the dc-axis magnetic flux. The dc-axis magnetic flux control unit 910 includes a subtractor 911, a proportional unit 912, an integrator 913, an adder 914, and a differential value calculation unit 915.

[0106] The subtractor 911 calculates the deviation between the dc-axis magnetic flux command value Φdc* and the dc-axis magnetic flux Φdc, and sends the deviation to the proportional unit 912, the integrator 913, and the differential value calculation means 915, respectively.

[0107] The proportional unit 912 amplifies the deviation between the dc-axis magnetic flux command value Φdc* and the dc-axis magnetic flux Φdc, and sends it to the adder 914 .

[0108] The integrator 913 time-integrates and amplifies the deviation between the dc-axis magnetic flux command value Φdc* and the dc-axis magnetic flux Φdc, and sends the amplified deviation to the adder 914 .

[0109] The adder 914 receives the output signals from the proportional unit 912 and the integrator 913 as input values, and calculates the dc-axis voltage vdc by adding up the outputs of each.

[0110] Incidentally, the response of the feedback value Φdc to the command value Φdc* is generally controlled by adjusting and designing the amplification gains of the proportional unit 912 and the integrator 913 to achieve a desired response. For example, the dc-axis magnetic flux Φ is often designed with a first-order lag for the dc-axis magnetic flux command Φdc*. In this case, the relationship between Φdc* and Φdc is expressed by the following equation (17):

[0111]

[0112] Here, s is the Laplace operator, and T is the response time constant of Φdc. By modifying equation (17), the differential value of Φdc can be calculated as follows:

[0113]

[0114] Since the Laplace operator s represents differentiation, it is possible to perform a differential calculation of Φdc using equation (18). In this way, by finding the time differential from the deviation between the command value and the feedback value, time differentiation is not performed directly, and therefore noise amplification can be suppressed. In this way, the differential value calculation means 915 is configured, and the dc-axis magnetic flux differential value p(Φdc) is output.

[0115] Although the dc-axis magnetic flux control unit 910 is shown as being configured to perform proportional-integral control as an example, it may take various known forms, such as adding a term to compensate for the armature reaction of the synchronous motor 200.

[0116] 10 is a diagram showing the configuration of the position estimation unit 100B. This embodiment differs from the position estimation unit 100A in the first embodiment in that a dc-qc axis magnetic flux calculation unit 110 is provided outside the position estimation unit in order to perform control based on magnetic flux, and in that there is no differential calculation unit 150 because the magnetic flux differential value is calculated by the magnetic flux control unit.

[0117] Other points are the same as those in the first embodiment.

[0118] In the figure, the axis error is calculated based on the dc-axis voltage vdc, the qc-axis voltage vqc, the dc-axis magnetic flux Φdc, the qc-axis magnetic flux Φqc, the differential value p(Φdc) of the dc-axis magnetic flux, and the differential value p(Φqc) of the qc-axis magnetic flux. However, various applications are possible, such as the configuration of FIG. 7 and equation (16).

[0119] The main features of the first and second embodiments can be summarized as follows.

[0120] The control device 1 of the synchronous motor 200 defines the rotor position of the synchronous motor 200 as the d-axis, the axis 90 degrees ahead of the d-axis as the q-axis, the axis having a phase difference θe with the d-axis as the dc-axis, and the axis 90 degrees ahead of the dc-axis as the qc-axis, and controls the synchronous motor 200 using the phase of the dc-axis (the rotor position estimated value θest). The processor of the control device 1 of the synchronous motor 200 calculates the phase difference θe using the voltages of the dc-axis and qc-axis (voltage commands vdc, vqc), the angular velocity of the dc-axis (the electrical angular velocity estimated value ω^), the characteristic values ​​of the synchronous motor 200 (the resistance value R, the static q-axis inductance Lq), and the magnetic fluxes of the dc-axis and qc-axis (Φdc, Φqc) without using an approximation of the angular velocity of the d-axis (the true electrical angular velocity ω) ( FIG. 2 , position estimator 100A, equations (8) and (9)).

[0121] In equations (8) and (9), the approximation that the true electrical angular velocity value ω (speed of the d-axis) = the estimated electrical angular velocity value ω^ (speed of the d-axis) is unnecessary, so the accuracy of the phase difference θe (axis error) is ensured even if the synchronous motor suddenly accelerates or decelerates.

[0122] Since equations (6) and (7) do not include a dynamic inductance term, θe can be easily calculated while taking static and dynamic inductance into consideration. The voltage equations of equations (6) and (7) are expressed in terms of magnetic flux instead of current, so they are highly compatible with motor control using magnetic flux command values ​​(magnetic flux commands Φdc*, Φqc*). For example, a processor derives magnetic flux command values ​​(magnetic flux commands Φdc*, Φqc*) from a torque command value (torque command T*) or a current command value using a table (map) or a formula, etc., and performs feedback control.

[0123] Since the characteristic values ​​(parameters) of the synchronous motor 200 used in equations (8) and (9) are only the resistance value R and the static q-axis inductance Lq, the capacity of the storage device (memory) that stores the characteristic values ​​can be reduced, thereby reducing the manufacturing cost of the control device 1.

[0124] The processor may further calculate the phase difference θe using the magnetic flux differential values ​​(pΦdc, pΦqc) of the dc-axis and qc-axis (FIG. 4, position estimation unit 100A, equation (8)). Since the magnetic flux differential values ​​of the dc-axis and qc-axis are used in equation (8), the accuracy of θe is ensured even if the magnetic flux changes while taking into account dynamic inductance.

[0125] In the first embodiment described above, the processor derives the magnetic fluxes (Φdc, Φqc) on the dc and qc axes using a table (map) that defines the correspondence between the currents on the dc and qc axes (current commands i dc* and i qc*) and the magnetic fluxes (Φdc, Φqc) on the dc and qc axes (FIG. 3, dc-qc axis magnetic flux calculation unit 110). The table shown in FIG. 3 contains information on the gradient of the magnetic flux relative to the current (the gradient of the magnetic flux near the current operating point and the gradient of the magnetic flux between the current operating point and the origin), so it is possible to derive the magnetic fluxes (Φdc, Φqc) on the dc and qc axes from the currents on the dc and qc axes while taking into account dynamic and static inductances.

[0126] On the other hand, the processor may calculate the magnetic fluxes (Φdc, Φqc) of the dc and qc axes based on voltage equations (Equations (10) and (11)) of the voltages (voltage commands vdc, vqc) of the dc and qc axes represented by the magnetic fluxes (Φdc, Φqc) of the dc and qc axes and the characteristic value (resistance value R) of synchronous motor 200 (FIG. 6, dc-qc axis magnetic flux calculation unit 110, Equations (12) and (13)). The calculations of Equations (12) and (13) are composed of only simple arithmetic operations, and therefore the calculation load on the processor can be reduced.

[0127] In the first embodiment described above, the differential calculation unit 150 calculates the magnetic flux differential values ​​(pΦdc, pΦqc) by pure time differentiation or filtering, but the processor may also calculate the magnetic flux differential values ​​(pΦdc, pΦqc) of the dc-axis and qc-axis using the difference between the magnetic flux command values ​​of the dc-axis and qc-axis (flux commands Φdc*, Φqc*) that are drive commands for the synchronous motor 200 and the magnetic flux of the dc-axis and qc-axis (Φdc, Φqc) ( FIG. 9 , magnetic flux control unit 900, equation (18), etc.). Because no direct time differentiation is performed, it is possible to suppress noise amplification, etc.

[0128] In the first embodiment described above, the axis error calculation unit 120 calculates θe using equation (8) or (9). However, the processor may also calculate the phase difference θe using the product of the qc-axis current (iqc) and the q-axis inductance (static q-axis inductance Lq), the product of the dc-axis current (idc) and the q-axis inductance (static q-axis inductance Lq), and the magnetic fluxes of the dc and qc axes (Φdc, Φqc) (see FIG. 7, axis error calculation unit 120; for example, an equation in which the numerator and denominator of the argument of the arctangent function of equation (16) are multiplied by Lq).

[0129] In other words, the processor calculates the phase difference θe using the dc-axis and qc-axis currents (idc, iqc), the q-axis inductance (static q-axis inductance Lq), and the dc-axis and qc-axis magnetic flux (Φdc, Φqc) (FIG. 7, axis error calculation unit 120, equation (16)). This has the advantage of being less susceptible to output voltage errors caused by dead time in the power converter. Furthermore, the only characteristic of the synchronous motor used is the static q-axis inductance Lq, which improves robustness against synchronous motor parameters.

[0130] As shown in FIG. 8 , the synchronous motor 200 may include a position sensor 200A (e.g., a resolver) for detecting the rotor position of the synchronous motor 200. The processor then calculates an estimate of the dc-axis phase (rotor position estimate θest) from the phase difference θe (position estimator 100B in FIG. 10 ). The processor compares the estimate of the dc-axis phase (rotor position estimate θest) with the rotor position θ detected by the position sensor while the synchronous motor 200 is running, and determines whether or not an abnormality exists in the position sensor. For example, the processor determines that an abnormality (failure) has occurred in the position sensor when the difference between the rotor position θ and the rotor position estimate θest is equal to or greater than a threshold. The high accuracy of the dc-axis phase (rotor position estimate θest) calculated from the highly accurate θe improves the accuracy of determining an abnormality (failure) in the position sensor.

[0131] The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0132] Furthermore, the above-described configurations, functions, etc. may be realized in part or in whole by hardware, for example, by designing them as integrated circuits. Furthermore, the above-described configurations, functions, etc. may be realized in software by a processor (microcomputer) interpreting and executing a program that realizes each function. Information such as the program, table, and file that realizes each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0133] The embodiment of the present invention may have the following aspects.

[0134] (C1) A synchronous motor control device that controls the synchronous motor using phase information of the dc axis, where the rotor position of the synchronous motor is the d-axis, the axis that is 90 degrees ahead of the d-axis is the q-axis, the axis that has a phase difference of θe with the d-axis is the dc-axis, and the axis that is 90 degrees ahead of the dc-axis is the qc-axis, and the synchronous motor is controlled using the voltage of the dc-axis and the qc-axis, the angular velocity of the dc-axis, a characteristic value of the synchronous motor, and the magnetic flux of the dc-axis and the qc-axis, and the magnetic flux of the dc-axis and the qc-axis is calculated by a dc-qc-axis magnetic flux calculation unit that has a table based on the relationship between the dc-axis and the qc-axis currents and the dc-axis and the qc-axis magnetic flux.

[0135] (C2) A control device for a synchronous motor, wherein θe is calculated using the voltages of the dc axis and the qc axis, the angular velocity of the dc axis, a characteristic value of the synchronous motor, the magnetic flux of the dc axis and the qc axis, and the magnetic flux differential values ​​of the dc axis and the qc axis.

[0136] (C3) The dc-qc axis magnetic flux calculation unit calculates the magnetic flux of the dc axis and the qc axis based on a voltage equation in which the voltages of the dc axis and the qc axis are expressed by the magnetic flux of the dc axis and the qc axis and characteristic values ​​of the synchronous motor.

[0137] (C4) A control device for a synchronous motor, characterized in that the magnetic flux differential values ​​of the dc-axis and the qc-axis are calculated using a difference between magnetic flux command values ​​of the dc-axis and qc-axis, which are drive commands for the synchronous motor, and the magnetic flux of the dc-axis and the qc-axis.

[0138] (C5) A control device for a synchronous motor, wherein the θe is calculated using the product of the qc-axis current and the q-axis inductance of the synchronous motor, the product of the dc-axis current and the q-axis inductance, and the magnetic fluxes of the dc axis and the qc axis.

[0139] (C6) In the control device for a synchronous motor according to any one of (C1) to (C5), the synchronous motor is provided with a position sensor for detecting a rotor position of the synchronous motor, and when the synchronous motor is being driven, the dc-axis phase information obtained from θe is compared with the rotor position obtained by the position sensor to determine whether or not there is an abnormality in the position sensor.

[0140] According to (C1)-(C6), the dc-qc-axis magnetic flux is calculated by a dc-qc-axis magnetic flux calculation unit that has the relationship between the dc-axis and qc-axis currents and the dc-axis and qc-axis magnetic flux. This dc-qc-axis magnetic flux includes changes in inductance due to magnetic saturation, and by expressing it in terms of magnetic flux, static and dynamic inductances can also be taken into account, making it possible to estimate position with high accuracy without a position sensor. In addition, there is no need for approximation, such that the d-axis speed, which is the true speed value, is equal to the dc-axis speed, which is the estimated speed value. This makes it possible to estimate position with high accuracy even when the rotation speed of the synchronous motor changes suddenly.

[0141] 1, 2...Control device 100A, 100B...Position estimation unit 110...dc-qc axis magnetic flux calculation unit 111...dc axis magnetic flux calculation unit 112...qc axis magnetic flux calculation unit 120...Axis error calculation unit 130...Speed ​​calculation unit 131...Subtractor 132...Proportional unit 133...Integrator 134...Adder 140...Position calculation unit 150...Differential calculation unit 200...Synchronous motor 200A...Position sensor 300...Power converter 400...Current command calculation unit 500...Current detection unit 600...Current control unit 700A...Coordinate inverse transformation unit 700B...Coordinate transformation unit 800...Flux command calculation unit 900...Flux control unit

Claims

1. A synchronous motor control device in which the rotor position of a synchronous motor is the d-axis, the axis 90 degrees ahead of the d-axis is the q-axis, the axis having a phase difference θe with the d-axis is the dc-axis, and the axis 90 degrees ahead of the dc-axis is the qc-axis, and which controls the synchronous motor using the phase of the dc-axis, and which has a processor that calculates the phase difference θe using the voltages of the dc-axis and the qc-axis, the angular velocity of the dc-axis, characteristic values ​​of the synchronous motor, and the magnetic flux of the dc-axis and qc-axis, without using an approximation of the angular velocity of the d-axis.

2. A control device for a synchronous motor according to claim 1, wherein the processor further calculates the phase difference θe using magnetic flux differential values ​​of the dc axis and the qc axis.

3. A control device for a synchronous motor according to claim 1, characterized in that the processor calculates the magnetic fluxes of the dc axis and the qc axis based on a voltage equation for the magnetic fluxes of the dc axis and the qc axis and the voltages of the dc axis and the qc axis represented by characteristic values ​​of the synchronous motor.

4. A control device for a synchronous motor according to claim 1, characterized in that the processor calculates magnetic flux differential values ​​for the dc axis and the qc axis using the difference between the magnetic flux command values ​​for the dc axis and the qc axis, which are drive commands for the synchronous motor, and the magnetic fluxes for the dc axis and the qc axis.

5. A control device for a synchronous motor according to claim 1, wherein the processor calculates the phase difference θe using the product of the qc-axis current and the q-axis inductance, the product of the dc-axis current and the q-axis inductance, and the magnetic fluxes of the dc-axis and the qc-axis.

6. A control device for a synchronous motor as described in claim 1, wherein the synchronous motor is equipped with a position sensor for detecting a rotor position of the synchronous motor, and the processor calculates an estimate of the phase of the dc axis from the phase difference θe, and during the period when the synchronous motor is running, compares the estimate of the phase of the dc axis with the rotor position detected by the position sensor, and determines whether or not there is an abnormality in the position sensor.

7. A control device for a synchronous motor according to claim 1, characterized in that the processor derives the magnetic fluxes of the dc axis and the qc axis using a table that defines the correspondence between the currents of the dc axis and the qc axis and the magnetic fluxes of the dc axis and the qc axis.

8. A control device for a synchronous motor according to claim 1, wherein the processor calculates the phase difference θe using the currents on the dc axis and the qc axis, the q-axis inductance, and the magnetic fluxes on the dc axis and the qc axis.

Images