JP2025001101A - Correlation signal generation method, phase velocity estimation device, and sensor-less vector control device - Google Patents

Abstract

To provide a correlation signal generation method for a carrier high-frequency voltage application method using axial element amplitude extraction that has the same degree of freedom of design as that of a conventional high-frequency voltage application method and can achieve the method while suppressing increase in an operation load.SOLUTION: In a correlation signal generation method for a carrier high-frequency voltage application method using axial element amplitude extraction, an axial element amplitude included in a high-frequency current (a stator current) is extracted, an intermediate signal is synthesized using the extracted axial element amplitude, and a positive correlation signal is adapted to be generated using the synthesized intermediate signal.SELECTED DRAWING: Figure 3

Description

The present invention relates to a correlation signal generation method, a phase speed estimation device, and a sensorless vector control device.

More specifically, the present invention relates to a correlation signal generation method used to estimate the rotor phase (synonymous with position) and speed in an AC motor (such as a permanent magnet synchronous motor (PMSM), which is a synchronous motor that uses a permanent magnet (ferromagnetic material) in the rotor (field magnet), a synchronous reluctance motor, an induction motor, etc.) whose rotor exhibits salient pole characteristics when a high-frequency voltage with a higher frequency than the drive frequency is applied, the correlation signal generation method being a carrier high-frequency voltage application method using shaft element amplitude extraction (the "carrier high-frequency voltage application method using shaft element amplitude extraction" will be described in detail later), a phase speed estimation device that estimates the rotor phase speed using the correlation signal generation method, and a sensorless vector control device equipped with the phase speed estimation device.

Highly efficient and highly responsive control of AC motors can be achieved by vector control. This vector control method requires position information of the motor rotor, but traditionally, position sensors such as encoders have been used to obtain this position information.

However, this type of position sensor has problems such as reduced reliability, increased volume, and increased costs, so research into sensorless vector control methods that do not require position sensors has been conducted for many years.

One of the most effective sensorless vector control methods is a high-frequency voltage application method in which a high-frequency voltage with a higher frequency than the drive frequency is forcibly applied, the high-frequency current resulting from the forced application is processed, and an estimate of the salient pole phase of an AC motor whose rotor exhibits salient pole characteristics in response to the application of a high-frequency voltage with a higher frequency than the drive frequency is obtained (see Non-Patent Documents 1 and 2).

In the high-frequency voltage application method, the rotor phase to be estimated can be determined arbitrarily, but it is common to select either the negative or positive salient pole phase of the rotor as the rotor phase. There is an electrical phase deviation of ±π/2 (rad) between the negative and positive salient pole phases, but if one of the phases is identified, the other can be determined automatically.

Taking the above into consideration, in the following explanation, unless otherwise specified, the rotor's negative salient pole phase will be described as the rotor phase.

The high-frequency voltage application method is characterized by how it extracts a signal component that contains phase information contained in the stator current and synthesizes a signal that is correlated with the rotor phase (in this specification and the claims, the "signal that is correlated with the rotor phase" will be referred to as the "correlated signal" as appropriate).

Typical correlation signals in conventional high-frequency voltage application methods include positive-phase and negative-phase amplitude correlation signals, shaft element amplitude correlation signals, and high-frequency current correlation signals (see Non-Patent Documents 2 to 6).

Here, the positive-phase/negative-phase amplitude correlation signal is a correlation signal that is synthesized using the extracted positive-phase/negative-phase amplitudes, extracted by filtering the amplitudes of the positive-phase and negative-phase components that make up the high-frequency current generated by applying a high-frequency voltage, making it possible to synthesize correlation signals with a wide variety of characteristics (see Non-Patent Documents 2 to 5).

The axis element amplitude correlation signal is a correlation signal that is synthesized using the extracted axis element amplitudes, extracted by filtering the amplitudes of each γδ axis element of the high-frequency current (see Non-Patent Documents 2 to 5).

There is a correspondence between the positive and negative phase amplitudes and the axis element amplitudes, and it is possible to synthesize a correlation signal with the same characteristics as the correlation signal synthesized in the positive and negative phase amplitude correlation signal in the axis element amplitude correlation signal (see Non-Patent Documents 2 to 5).

On the other hand, the high frequency current correlation signal is a correlation signal synthesized using each element of the high frequency current itself (see Non-Patent Document 2 and Non-Patent Documents 4 to 6). This high frequency current correlation signal does not involve the extraction of any special signals compared to other signals, so it is possible to reduce the computational load (see Non-Patent Document 2 and Non-Patent Documents 4 to 6).

However, since noise components are generated in the high-frequency current correlation signals during synthesis, it is practical to perform filtering to remove these noise components (see Non-Patent Document 2 and Non-Patent Documents 4 to 6). In addition, the characteristics of the synthesizable correlation signals seem to have less flexibility compared to others (see Non-Patent Document 2 and Non-Patent Documents 4 to 6).

Here, the main problems with the above-mentioned high-frequency voltage application method have been pointed out as the need to improve the speed of phase estimation and the need to reduce audible acoustic noise caused by the applied voltage.

The reason why the speed of response of phase estimation in the high-frequency voltage application method is limited is basically due to the large time constant (generally narrow bandwidth) filters used for demodulation. For this reason, there is a demand for proposals for demodulation methods that can use filters with high speed response, or demodulation methods that require as little as possible of this type of filter.

On the other hand, when it comes to reducing audible acoustic noise in the high-frequency voltage application method, even in the audible range of frequencies from 4 kHz to 16 kHz, the higher the frequency, the lower the audibility, so there is a demand for proposals for demodulation methods that use higher frequencies.

As an effective method for solving the major problems of the high frequency voltage application method described above, several methods have been proposed to apply a high frequency voltage having a frequency range (frequency ratio (frequency ratio between applied high frequency voltage and PWM carrier wave frequency. In other words, "frequency ratio = applied high frequency voltage / PWM carrier wave frequency") of 1 to 1/10 of the PWM carrier wave (frequency range referred to as "carrier high frequency range" in this specification and claims) (in this specification and claims, "method of applying a high frequency voltage having a carrier high frequency range" is referred to as "carrier high frequency voltage application method") similar to that of the PWM carrier wave (carrier wave used in PWM (pulse width modulation)) (see Non-Patent Documents 3 to 17).

Here, the high-frequency current, which is the response of the high-frequency voltage, begins to show a differential discontinuous response as the voltage frequency approaches the PWM carrier (see Non-Patent Documents 3 to 17). In conventional high-frequency voltage application methods that apply a high-frequency voltage with a frequency ratio of approximately 1/20 or less compared to the PWM carrier, analysis and processing methods that assume a differential continuous response have been widely used (see Non-Patent Documents 1 and 2).

However, this premise and method can no longer be applied to the carrier high-frequency voltage application method, and analysis and processing methods based on differential processing of the stator current have been used (see Non-Patent Documents 7 to 17). However, this differential processing has the inherent problem of being vulnerable to noise.

In Non-Patent Document 3, a new carrier high-frequency voltage application method is proposed in which a carrier high-frequency voltage is applied, but the carrier high-frequency current, which is the response, is processed without differentiation to obtain a phase estimate.

This method assumes the use of an inverter (power converter with PWM), and finds an analytical solution for sample values of a differentially discontinuous high-frequency current corresponding to the discrete-time application of a carrier high-frequency voltage. Based on this current analytical solution, a demodulation method, i.e., a phase estimation method, is constructed. According to this current analytical solution, the difference between a case where the high-frequency voltage frequency is sufficiently low compared to the carrier frequency and a case where it is close appears as a spatial phase difference between the high-frequency voltage and the high-frequency current, and the amplitude of the high-frequency current. Taking these differences into consideration, the conventional high-frequency voltage application method (see Non-Patent Documents 1 and 2), which assumes that the applied high-frequency voltage frequency is sufficiently low compared to the carrier frequency, can be applied in the carrier high-frequency region (see Non-Patent Documents 3, 5, and 6).

However, in reality, in addition to these differences, errors appear to occur in the above current analytical solution due to the effects of the inverter's short circuit prevention period (dead time) and the time delay in the current detection value due to AD conversion (see Non-Patent Documents 3 and 5).

Therefore, in the above-mentioned phase estimation method, an error appears to occur between the true value and the estimated value of the rotor phase due to the influence of the above-mentioned errors (see Non-Patent Document 5), but it appears that part of the error can be captured as a change in the spatial phase difference between the amplitude of the current analytical solution and the high-frequency voltage and high-frequency current, and by generating a phase deviation equivalent value (correlation signal) that is less affected by these, it appears that the influence of the high-frequency voltage error can be sufficiently suppressed (see Non-Patent Documents 3 and 5).

In order to apply the conventional high-frequency voltage application method to the carrier high-frequency voltage application method, Non-Patent Documents 18 and 19 propose a method for extracting this phase difference, based on the recognition that the error affecting the phase estimation in the error of the current analytical solution described above appears as a phase difference (high-frequency voltage phase error) when the high-frequency voltage command value and the true value are converted into unit vectors. However, although it is easy to realize a phase correction function using the phase error extracted by this extraction method, an increase in the calculation load is expected.

Non-patent document 20 focuses on a carrier high-frequency voltage application method using positive-phase and negative-phase amplitude extraction, and minimizes the increase in computational load by reconstructing the realization of the above-mentioned phase correction function as a correlation signal synthesis method.

Shinji Shinnaka: "Vector Control Technology of Permanent Magnet Synchronous Motors, Volumes 1 and 2 (The Essence of Sensorless Vector Control)", Dempa Shimbunsha (2008-12) Shinji Shinnaka: "Control of Permanent Magnet Synchronous Motors (Sensorless Vector Control Technology)", University of Tokyo Press (September 2013) R. Hosooka, S. Shinnaka, N. Nakamura: “New Sensorless Vector Control of PMSM by Discrete-Time Voltage Injection of PWM Carrier Frequency”, IEEJ Trans. IA, vol. 136, No. 11, pp. 837-850 (2016-11)Ryu Hosooka, Shinnaka Shinji, Nakamura Naoto: “Discrete-Time Carrier High-Frequency Voltage Injection Method for Sensorless Permanent Magnet Synchronous Motor”, Journal of Electrical Engineering D, 136, 11, pp. 837-850 (2016-11) Ryu Hosooka: "Study on the carrier high frequency voltage application method for sensorless permanent magnet synchronous motors", Kanagawa University Academic Institutional Repository, B0700-02-01工甲226 (2018-3) Shinji Shinnaka: "Detailed Explanation of Vector Control Technology for Synchronous Motors", University of Tokyo Press (June 2019) R. Hosooka, N. Nakamura, S. Shinnaka: “Sensorless Vector Control of PMSM using Positive- and Negative-Phase High-Frequency Currents Correlation “Signal by Discrete-Time Voltage Injection of PWM Carrier Frequency”, IEEJ Trans. IA, vol. 138, No. 2, pp. 150-163 (2018-2) Ryu Hosooka, Naoto Nakamura, Shinji Shinnaka: "A discrete-time carrier high-frequency voltage application method using positive-phase and negative-phase high-frequency current correlation for sensorless permanent magnet synchronous motors", IEICE Theory D, 138, 2, pp. 150-163 (2018-2) R. Masaki, S. Kanekko, Y. Sakurai, and M. Hombu, "Position Sensorless Control System of IPM Motor Based on Voltage Injection Synchronized with PWM Carrier", IEEJ Trans. IA, vol. 134, No. 1, pp. 37-43 (2002-2) 37-43 (2002-2) S. Shinnaka, "New Sensorless Vector Control of PMSM by Pulsation Voltage Injection of PWM Carrier Frequency", IEEJ Trans. IA, vol. 134, No. 6, pp. 596-605 (2014-6) Shinnaka, S.: "Sensorless Vector Control of Permanent Magnet Synchronous Motor by Linear PWM Carrier High Frequency Voltage Injection", Electrical Engineering Society D, 134, 6, pp. 596-605 (2014-6) Ryu Hosooka and Shinji Shinnaka, "Actual verification of linear PWM carrier high frequency voltage application method for sensorless permanent magnet synchronous motors", 2014 National Convention of the Institute of Electrical Engineers of Japan, 4-100 (March 2014)

S. Shinnaka, "New Sensorless Vector Control of PMSM by Rotating Voltage Injection of PWM Carrier Frequency", IEEJ Trans. IA, vol. 134, No. 6, pp. 606-617 (2014-6) Shinnaka, S.: "Sensorless Vector Control of Permanent Magnet Synchronous Motor by Rotating Voltage Injection of PWM Carrier Frequency", Journal of Electrical Engineering D, 134, 6, pp. 606-617 (2014-6) Y. D. Yoon, S. K. Sul, S. Morimoto, and K. Ide: “High-Bandwidth Sensorless Algorithm for AC Machines Based on Square-Wave-Type Voltage Injection”, IEEE Trans Ind. Appl. , Vol. 47, No. 3, pp. 1361-1370 (2011-5/6) S. Murakami, T. Shiota, M. Ohta, and K. Ide: “Encoderless Servo Drive With Adequately Designed IPMSM for Pulse-Voltage-Injection-Based Position Detection”, IEEE Trans Ind. Appl. , Vol. 48, No. 6, pp. 1922-1930 (2012-11/12) S. Kim, and S. K. Sul: “High Performance Position Sensorless Control Using Rotating Voltage Signal Injection in IPMSM”, EPE 2011, pp. 1-10 (2011-9) S. Kim, J. I. Ha, and S. K. Sul: “PWM Switching Frequency Signal Injection Sensorless Method in IPMSM”, IEEE Trans Ind. Appl. , Vol. 48, No. 5, pp. 1576-1587 (2012-9/10) D. Kaneko, Y. Iwaji, K. Sakamoto, T. Endoh "Initial Rotor Position Estimation of Interior Permanent Magnet Synchronous Motor", IEEJ Trans. IA, vol. 123, No. 2, pp. 140-148 (2003-2) J. Lara, and A. Chandra: “Performance Study of Switching Frequency Signal Injection Algorithm in PMSMs for EV Propulsion: A Comparison in Stator and Rotor Coordinates”, ISIE 2014, pp. 865-870 (2014-6) D. Kim, Y. C Kwon, and S. K Sul: “Suppression of injection voltage disturbance for High Frequency square-wave injection sensorless “drive with regulation of induced High Frequency current ripple”, IPEC-Hiroshima 2014 - ECCE-ASIA, pp. 925-932 (2014-5) Ryu Hosooka: "High-frequency voltage phase error extraction method in carrier high-frequency voltage application method based on positive-phase and negative-phase amplitude correlation for sensorless permanent magnet synchronous motors", 2022 IEEJ Industrial Applications Division Conference 3-10 (2022-8) Akihiro Nakamura, Hiroki Kurumatani, Yuta Miyamoto, Ryu Hosooka: "High-frequency error angle estimation for carrier high-frequency voltage application method using elliptical voltage", 2023 National Conference of the Institute of Electrical Engineers of Japan 5-095 (September 2023) Ryu Hosooka: "Correlation signal synthesis method in carrier high frequency voltage application method using positive and negative phase amplitude extraction for sensorless permanent magnet synchronous motors", 2023 National Conference of the Institute of Electrical Engineers of Japan 5-094 (2023-9)

The present invention has been made in consideration of the above-mentioned various problems with the conventional technology and the demands related to these problems, and aims to provide a correlation signal generation method targeted at a carrier high-frequency voltage application method using axial element amplitude extraction, which has design freedom equal to or greater than that of conventional high-frequency voltage application methods, and can be realized while suppressing an increase in the computational load.

The present invention also aims to provide a phase velocity estimation device that uses the correlation signal generation method according to the present invention.

Furthermore, the present invention aims to provide a sensorless vector control device equipped with the phase speed estimation device according to the present invention.

To achieve the above object, the correlation signal generation method according to the present invention has the following characteristics (1) to (4).

(1) Four intermediate signals that have a phase error correction function in the current analytical solution are synthesized from the four axial element amplitudes, and then used to synthesize a correlation signal.

(2) The phase error correction function in the current analysis solution eliminates the need for phase correction processing, which is required when extracting the axis element amplitude.

(3) The four intermediate signals that are synthesized enable the synthesis of a wide variety of correlation signals depending on the selection of design parameters.

(4) By selecting specific values as design parameters, the four types of intermediate signals to be synthesized can be used to synthesize a positive correlation signal that has the same correlation characteristics as a conventional axis element amplitude correlation signal. In other words, it has the same degree of design freedom as a conventional axis element amplitude correlation signal.

The correlation signal generation method according to the present invention, which has these characteristics, is a correlation signal generation method used to estimate the rotor phase speed in an AC motor in which the rotor exhibits salient pole characteristics when a high-frequency voltage with a frequency higher than the driving frequency is applied, and in the correlation signal generation method in the carrier high-frequency voltage application method using shaft element amplitude extraction, the shaft element amplitude contained in the high-frequency current is extracted based on the following equations (26a) and (26b), and an intermediate signal is synthesized based on the following equations (28a) to (28d) using the extracted shaft element amplitude, and the synthesized intermediate signal is used to

p _c =f(C _2γ , S _2γ , S _2δ , C _2δ )
The positive correlation signal p _c is generated based on the above.

In addition, the phase speed estimation device according to the present invention is a correlation signal generation device used to estimate the rotor phase speed in an AC motor in which the rotor exhibits salient pole characteristics in response to application of a high-frequency voltage having a frequency higher than the driving frequency, and includes a high-frequency voltage application means for generating a discrete high-frequency voltage command value, and a means for extracting shaft element amplitudes contained in the high-frequency current based on the following equations (26a) and (26b), synthesizing intermediate signals based on the following equations (28a) to (28d) using the extracted shaft element amplitudes, and synthesizing intermediate signals based on the following equations (28a) to (28d), and using the synthesized intermediate signals.

p _c =f(C _2γ , S _2γ , S _2δ , C _2δ )
a correlation signal generating means for generating a positive correlation signal p _c based on the positive correlation signal p _c generated by the correlation signal generating means; and and generating means.

The phase velocity estimation device according to the present invention is the above-described phase velocity estimation device according to the present invention, further comprising a phase compensation means for correcting steady-state phase deviation.

The phase velocity estimation device according to the present invention is the above-described phase velocity estimation device according to the present invention, further comprising a bandpass filter that extracts high-frequency components.

The sensorless vector control device according to the present invention is a sensorless vector control device that controls the drive of an AC motor whose rotor exhibits salient pole characteristics in response to the application of a high-frequency voltage with a frequency higher than the drive frequency, and is equipped with the phase speed estimation device according to the present invention described above.

The correlation signal generation method according to the present invention is configured as described above, and is a correlation signal generation method targeted at a carrier high-frequency voltage application method using axial element amplitude extraction, and has the excellent effect of being able to provide a correlation signal generation method that has the same design freedom as conventional high-frequency voltage application methods, while suppressing an increase in the computational load.

In addition, the phase velocity estimation device according to the present invention and the sensorless vector control device equipped with the phase velocity estimation device have the excellent effect of making it possible to realize each device while suppressing an increase in the computational load.

FIG. 1 is an explanatory diagram showing the relationship between three types of coordinate systems and the rotor salient pole phase. FIG. 2 is a graph showing an example of characteristics of a positive correlation signal. FIG. 3 is a block diagram illustrating the configuration of a phase-speed estimator according to an embodiment of the present invention. FIG. 4 is a block diagram illustrating the configuration of a correlation signal generator in the phase velocity estimator shown in FIG. FIG. 5 is a block diagram illustrating the configuration of an amplitude extractor in the correlation signal generator shown in FIG. FIG. 6 is a block diagram illustrating the configuration of a sensorless vector control device according to an embodiment of the present invention.

Below, we will explain in detail an example of application of the correlation signal generation method, phase speed estimation device, and sensorless vector control device of the present invention to a permanent magnet synchronous motor, with reference to the attached drawings.

(I) Description of a correlation signal generation method according to an embodiment of the present invention

Figure 1 shows an explanatory diagram showing the relationship between three types of coordinate systems and the rotor salient pole phase.

More specifically, Figure 1 shows an explanatory diagram showing the relationship between three types of coordinate systems, namely, the dq synchronous coordinate system consisting of the d-axis and q-axis, the αβ fixed coordinate system consisting of the α-axis and β-axis, and the γδ quasi-synchronous coordinate system consisting of the γ-axis and δ-axis, and the rotor salient pole phase. The d-axis direction in the diagram is defined as the rotor negative salient pole phase direction.

In addition, in an AC motor that uses a permanent magnet in the rotor (such as a permanent magnet synchronous motor), if the effects of nonlinear characteristics such as saturation of the magnetic circuit and inter-axis magnetic flux interference between the d and q axes can be ignored, the d-axis direction is also the north pole direction of the rotor magnet.

First, as shown in FIG. 1, in a γδ quasi-synchronous coordinate system rotating at a speed _ωγ , it is assumed that the d-axis of a rotor 10 of a permanent magnet synchronous motor forms a phase _θγ with respect to the γ-axis of the main shaft at a certain instant.

Then, a mathematical model (circuit equation) of a permanent magnet synchronous motor on the γδ quasi-synchronous coordinate system can be written as the following equations (1) to (8).

In the above formulas (1) to (8), the 2-row, 1-column vectors v ₁ , ι ₁ , and φ ₁ respectively represent the stator voltage, current, and (interlinkage) magnetic flux. The 2-row, 1-column vectors φ _ι and φ _m represent the components that make up the stator magnetic flux φ ₁ , where φ _ι is the stator reaction flux (armature reaction flux) caused by the stator current ι ₁ , and φ _m is the rotor magnetic flux caused by the rotor permanent magnet. I is a 2-row, 2-column unit matrix, and J is a 2-row, 2-column alternating matrix defined by the following formula (9).

In addition, _ω2n is the electrical speed of the rotor, _R1 is the resistance of the stator winding, _{Lι and Lm} are the in-phase inductance and mirror-phase inductance of the stator, and have the relationship with the d-axis and q-axis inductances expressed by the following equation (3).

Also, s is the differential operator d/dt.

When a high frequency voltage is superimposed on the driving voltage, the following equation (11) holds for the voltage, current, and magnetic flux of the stator.

Here, the subscripts f and h in equation (11) refer to the driving component and high-frequency component, respectively.

Next, we will explain the discrete-time high-frequency voltage and the discrete-time high-frequency current based on the mathematical model (circuit equation) of the permanent magnet synchronous motor in the γδ quasi-synchronous coordinate system described above.

First, the discrete-time high-frequency voltage will be explained. When the frequency of the high-frequency voltage/current is sufficiently high relative to the driving voltage/current, the following equation approximately holds:

As the above formulas (12a), (12b), (13a), and (13b) clearly show, the high frequency magnetic flux φ _ιh has an integral dynamic relationship with the high frequency voltage v _ιh , while the high frequency current ι _ιh has a static relationship with the high frequency magnetic flux φ _ιh .

It is assumed that the average frequency _ωh of the applied high frequency voltage _v1h satisfies the relationship of the following equation (14) in terms of the relative ratio to the coordinate system velocity _ωγ .

When the above-mentioned formula (14) is satisfied, the formula (13a) is approximated as the following formula (15).

Here, the continuous-time high-frequency voltage v _1h on the right-hand side of equation (15) is also applied via a power converter modeled as a zero-order holder.

Also,

Time t = _kTs to (k+1) _Ts
The discrete-time high-frequency voltage corresponding to the continuous-time high-frequency voltage v _1h is expressed as v _1h,k . As the discrete-time high-frequency voltage v _1h,k , a constant elliptical high-frequency voltage with a high-frequency period T _h and an average speed _ωh and the following elliptical coefficient K is considered.

Here, the subscript k denotes the sampling time at t= _kTs . θ _h0 is the initial phase of the applied high-frequency voltage, and in this specification and claims, unless otherwise specified, θ _h0 =0. For simplicity, the high-frequency period T _h is selected so as to hold the following relationship using the discrete time period T _s and a positive integer N _h ≧2.

When N _h =2 is selected, the voltage waveform becomes linear regardless of the ellipse coefficient K.

As for the error in the process from the voltage command to the current detection, external factors such as the time delay of the power conversion to generate the voltage according to the voltage command value and the AD conversion accompanying the current detection are considered. In this specification and the claims, the errors associated with all external factors are modeled as being summarized in the error of the power conversion when generating the voltage according to the voltage command value. For this purpose, the following is considered as the command value v ^* _1h,k corresponding to the discrete-time high-frequency voltage v1h _,k shown in the formulas (16a) to (16c).

Here, θ _he is a phase error between a discrete-time high-frequency voltage command and its true value defined by the following equation (20):

The discrete-time high-frequency voltage command v ^* _1h,k is sent to a power converter to generate a discrete-time high-frequency voltage, which is then applied to a permanent magnet synchronous motor. For simplicity, it is assumed that the change in the true value and the command value can be ignored for the elliptical shape of the discrete-time high-frequency voltage. That is,

K ^* = K (21)
Let us assume that.

Next, the discrete-time high-frequency current response will be explained. When the elliptic coefficient K is constant, the discrete-time high-frequency current ι 1h, _{k in response to the application of the discrete-time high-frequency voltage v 1h ,} k in equation (16a) is given by the following equation:

The high frequency currents in equations (22a), (22b) and (22c) are re-expressed as follows:

As shown in the above equation, the effect of the phase error θ _he of the discrete-time high-frequency voltage appears as a phase change of the unit vector that constitutes the high-frequency current.

Next, we will explain amplitude extraction and phase estimation of discrete-time currents, but first we will explain amplitude extraction and correlation signal synthesis.

First, the amplitude extraction will be explained. The axis element amplitude contained in the high frequency current l _1h,k is extracted based on the following equation.

は、正規化周波数ゼロで減衰ゼロを、正規化周波数

で十分な減衰を示すディジタルローパスフィルタを意味する。

は、（２５ａ）式および（２５ｂ）式で定義された高周波電圧位相と同じ平均速度ω_ｈで変化する振幅抽出用高周波位相であり、簡単には、

として高周波電圧位相指令をそのまま持ち得ればよい。 In the above formula,

gives zero attenuation at normalized frequency zero, and

This means a digital low-pass filter that exhibits sufficient attenuation at

is a high-frequency phase for amplitude extraction that changes at the same average speed _ωh as the high-frequency voltage phase defined in equations (25a) and (25b), and can be expressed simply as

It is sufficient to have the high frequency voltage phase command as it is.

の偏差である。 Δθ ^* _h is the high-frequency voltage phase command θ ^* _h,k-1 and the high-frequency phase for amplitude extraction

is the deviation.

を用いて、中間信号Ｃ_２γ，Ｓ_２γ，Ｓ_２δ，Ｃ_２δを次式に基づき合成する。

Next, the synthesis of the intermediate signal will be described. As a preliminary step for synthesizing the positive correlation signal p _c , the axis element amplitudes extracted based on the formulas (26a) and (26b) are

Using this, intermediate signals C _2γ , S _2γ , S _2δ , and C _2δ are synthesized based on the following equations.

に含まれる離散的位相特性

、位相誤差θ_ｈｅ、位相偏差Δθ^＊ _ｈの影響を受けない信号となる。 The intermediate signal synthesized based on equations (28a) to (28d) has the amplitude

Discrete phase characteristics contained in

, the signal is not affected by the phase error θ _he and the phase deviation Δθ ^* _h .

Furthermore, K _γc , K _γs , and K _δc are arbitrary design parameters left to the discretion of the designer. Representative selection examples of the design parameters are as follows.

In equations (28a) to (28d), the intermediate signals C _2γ , S _2γ , S _2δ , and C _2δ , which are selected such that K _γc = K _γs = K _δc = K, and the axis element amplitudes c _γ , s _γ , s _δ , and c _δ are in a proportional relationship as shown in the following equation. By taking this difference into consideration, it becomes possible to apply the conventional correlation signal synthesis method using axis element amplitudes to the carrier high frequency voltage application method.

Next, synthesis of a positive correlation signal will be explained. A method of synthesizing a positive correlation signal p _c using the intermediate signals C _2γ , S _2γ , S _2δ , and C _2δ synthesized as described above will be explained.

In other words, there are many different possible correlation signal synthesis methods based on the following equation (33), similar to the conventional positive correlation signal synthesis method using axial element amplitudes.

p _c =f(C _2γ , S _2γ , S _2δ , C _2δ ) (33)

In this specification and the claims, a positive correlation signal given by the following equation will be shown as an example of the correlation signal synthesis method based on equation (33).

Here, atan2(·) means four-quadrant arctangent processing. _Kc is an arbitrary design parameter left to the designer. The positive correlation signal synthesized based on the above formulas (34) and (35) allows for a very high degree of freedom in designing the positive phase correlation characteristics and the calculation load, including the selection of the design parameters _Kγc , _Kγs , and _Kδc of the intermediate signal.

Representative parameters are as follows:

および位相誤差θ_ｈｅの値に関わることなく±π／２の範囲にて優れた直線近似特性を有することが確認される。 2 shows an example of the characteristics of a positive correlation signal based on the formula (36a) under the conditions of an ellipse coefficient K=0.5, a phase error θ _he =0.5 [rad], and a design parameter K _c =K. From this FIG. 2, it can be seen that the phase characteristic of this positive correlation signal varies depending on the design parameters when the ellipse coefficient K=0.5.

It is also confirmed that the linear approximation characteristic is excellent in the range of ±π/2, regardless of the value of the phase error θ _he .

In addition, by converting the intermediate signal according to the present invention described above based on the following equation (40), it is possible to synthesize an intermediate signal having the same characteristics as the intermediate signal introduced in Non-Patent Document 20.

Therefore, since there exists a correspondence relationship similar to that between the positive-phase and negative-phase amplitudes and the shaft element amplitudes used in the conventional high-frequency voltage application method, it is possible to synthesize a positive-phase-to-positive signal having the same characteristics as the positive-phase-to-positive signal that can be synthesized using the intermediate signal introduced in Non-Patent Document 20, using the intermediate signal according to the present invention described above. In other words, it is also possible to synthesize a positive-phase-to-positive signal having characteristics equivalent to the positive-phase-to-positive signal using conventional positive-phase and negative-phase amplitude extraction.

に変換される。併せて、回転子の速度推定値

が生成される。一般化積分形ＰＬＬ法（離散時間版）は、以下のように記述される。

Next, the generalized integral PLL method will be described. The synthesized positive correlation signal p _c is processed based on the generalized integral PLL method to obtain the rotor phase estimate with respect to the α axis.

In addition, the rotor speed estimate

The generalized integral PLL method (discrete time version) is described as follows:

は、γδ準同期座標系速度をそのまま用いてもよいし、ローパスフィルタ処理して用いてもよい。 Here, Δθ _s on the right side of equation (38a) is a compensation signal for the stationary phase deviation caused by inter-axis magnetic flux interference, high-frequency residual disturbance, etc. Also, the rotor speed estimate

The γδ quasi-synchronous coordinate system velocity may be used as is, or may be subjected to low-pass filtering before use.

In addition, equation (38c) shows two examples: direct use and using a first-order low-pass filter.

(II) Description of a phase velocity estimation device according to an embodiment of the present invention

Figure 3 shows a block diagram of a phase-speed estimator according to an embodiment of the present invention.

In addition, in Figure 3 and Figures 4 to 6 described below, vector signals in two rows and one column are shown with a single thick signal line to improve visibility.

That is, FIG. 3 shows a phase speed estimation device that generates a discrete-time high-frequency voltage command value and processes the discrete-time high-frequency current to generate rotor phase estimates and speed estimates.

The phase speed estimator 100 shown in FIG. 3 includes a high-frequency voltage commander (HFVC) 102 as a high-frequency voltage application means that generates a discrete high-frequency voltage command value based on equations (19a) and (19b), and a correlation signal generator as a correlation signal generation means that extracts the positive-phase amplitude and the negative-phase amplitude contained in the high-frequency current (stator current) based on equations (26a) and (26b), synthesizes an intermediate signal based on equations (28a) to (28d) using the extracted positive-phase amplitude and the negative-phase amplitude, and generates a positive correlation signal based on equation (33) using the synthesized intermediate signal. The system is configured with a phase synchroniser 104 and a phase synchronizer 106, which is an estimation generating means that generates a rotor phase estimate and a speed estimate as viewed from the α-axis of the rotor based on equations (38a), (38b), (38c) and (38d).

The reference numeral 108 denotes a band-pass filter that extracts high-frequency components, but this is not necessarily required in the phase velocity estimation device 100, depending on the design of the amplitude extractor (see FIG. 4), which is the amplitude extraction means constituting the correlation signal generator 104. Taking this into consideration, it is indicated by a dashed block in FIG. 3.

Also, reference numeral 110 denotes a phase compensator serving as a phase compensation means. This phase compensator 110 plays a role of generating a phase correction signal K _θ Δθ _s for correcting a stationary phase deviation, but since it is not necessarily required for implementing the phase velocity estimation device 100 according to the present invention, it is indicated by a dashed block similarly to the band pass filter 108.

Here, FIG. 4 shows a block diagram of the correlation signal generator in the phase velocity estimation device shown in FIG. 3.

Figure 5 also shows a block diagram of the amplitude extractor in the correlation signal generator shown in Figure 4.

The correlation signal generator 104 is configured to include an amplitude extractor 112, which is an amplitude extraction means that extracts the positive-phase amplitude and the negative-phase amplitude contained in the high-frequency current (stator current) based on equations (26a) and (26b), and a correlation signal synthesizer 114, which is a positive correlation signal synthesis means that synthesizes a positive correlation signal based on equations (28a) to (28d) and equation (33).

In the above configuration, the phase speed estimation device 100 receives as input the stator current in the γδ quasi-synchronous coordinate system, which is the output signal of the vector rotator (see vector rotator 212 shown in Figure 6), and outputs a rotor phase estimate, speed estimate, and discrete high-frequency voltage command value.

More specifically, the high-frequency voltage command generator 102 generates a discrete high-frequency voltage command value based on equations (19a) and (19b) and outputs it to the outside of the phase speed estimation device 100.

が抽出される（図５を参照する。）。 Then, the amplitude extractor 112 of the correlation signal generator 104 extracts the axial element amplitude from the discrete-time high-frequency current ι _1h,k based on the equations (26a) and (26b).

is extracted (see FIG. 5).

は、相関信号合成器１１４へ出力される。 The axis element amplitude extracted by the amplitude extractor 112

is output to the correlation signal combiner 114.

The correlation signal combiner 114 combines the positive correlation signal p 1 _c based on equations (28a) to (28d) and equation (33), and outputs the combined positive correlation signal p 1 _c to the phase synchronizer 106 .

と、回転子の速度推定値

とを生成して、位相速度推定装置１００の外部へ出力する。 The phase synchronizer 106 executes equations (38a), (38b), (38c), and (38d) using the positive correlation signal p _c to obtain an estimated phase value of the rotor of the permanent magnet synchronous motor as viewed from the α-axis of the rotor.

and the rotor speed estimate

and output it to the outside of the phase velocity estimation apparatus 100.

(III) Description of a sensorless vector control device according to an embodiment of the present invention

Figure 6 shows a block diagram of a sensorless vector control device according to an embodiment of the present invention.

The sensorless vector control device 200 according to an embodiment of the present invention includes the phase speed estimation device 100 according to the present invention described above, and controls the driving of the permanent magnet synchronous motor 300.

In FIG. 6, a block ^ST designated by reference numeral 202 is a three-phase to two-phase converter, and a block S designated by reference numeral 204 is a two-phase to three-phase converter.

In Figure 6, the stator voltage and stator current are expressed with the subscripts r (γδ quasi-synchronous coordinate system), s (αβ fixed coordinate system), and t (uvw coordinate system) to clearly indicate the coordinate systems in which these signals are defined.

The sensorless vector control device 200 includes the phase speed estimation device 100 described above, and the application of the stator voltage and the detection and control of the stator current are performed in discrete time after being synchronized.

In the sensorless vector control device 200, the block F _bs (z ⁻¹ ) indicated by reference numeral 206 in the current feedback is a high-frequency component removal filter (digital filter for removing high-frequency components) for removing high-frequency components contained in the stator current, but may not be used. Taking this into consideration, it is indicated by a dashed line block in FIG. 6.

Generally, a band-stop filter is used as the digital filter F _bs (z ⁻¹ ). However, when the frequency of the high-frequency current is particularly high, a low-pass filter with a wide bandwidth may be used.

More specifically, the sensorless vector control device 200 is configured to include a power converter (inverter) 208, a discrete-time current detector 210, a three-phase to two-phase converter 202, a two-phase to three-phase converter 204, an inverse vector rotator 212, a vector rotator 214, a current controller 216, a command converter 218, a speed controller 220, a high-frequency component removal filter 206, the phase speed estimation device 100 according to the present invention described above, a coefficient unit 222, and a cosine/sine signal generator 224.

In the sensorless vector control device 200, the three-phase stator currents detected by the discrete-time current detector 210 are converted by the three-phase to two-phase converter 202 into two-phase currents in an αβ fixed coordinate system, and then converted by the inverse vector rotator 212 into two-phase currents in a γδ quasi-synchronous coordinate system that aims for phase synchronization with zero phase deviation to the rotor phase (same as the phase in the dq synchronous coordinate system).

The two-phase current converted by the three-phase to two-phase converter 202 has high-frequency components contained in the stator current removed by the high-frequency component removal filter 206, and the two-phase current from which the high-frequency components have been removed is output to the current controller 216.

The current controller 216 generates two-phase drive voltage command values in the γδ quasi-synchronous coordinate system so that the two-phase drive currents in the γδ quasi-synchronous coordinate system follow the current command values of each phase.

Here, in the sensorless vector control device 200, the discrete high-frequency voltage command value generated by the high-frequency voltage commander 102 of the phase speed estimation device 100 is superimposed on the two-phase voltage command value for driving, and the superimposed and synthesized two-phase voltage command value is output to the vector rotator 214.

The vector rotator 214 converts the superimposed voltage command value in the γδ quasi-synchronous coordinate system into a two-phase voltage command value in the αβ fixed coordinate system, and outputs it to the two-phase to three-phase converter 204.

Then, the two-phase to three-phase converter 204 converts the two-phase voltage command value into a three-phase voltage command value and outputs it as the final voltage command value to the power converter 208.

The power converter 208 generates a voltage according to the final voltage command value and applies this voltage to the permanent magnet synchronous motor 300 to control the drive of the permanent magnet synchronous motor 300.

As described above, the phase speed estimation device 100 receives as input the stator current in the γδ quasi-synchronous coordinate system, which is the output signal of the vector rotator 212, and outputs a rotor phase estimate, a speed estimate, and a discrete high-frequency voltage command value.

The rotor phase estimate output from the phase speed estimation device 100 is converted to a cosine-sine signal in the cosine-sine signal generator 224, and then output to the inverse vector rotator 212 and vector rotator 214, which determine the γδ quasi-synchronous coordinate system. This means that the rotor phase estimate is the phase of the γδ quasi-synchronous coordinate system (equivalent to the phase of the γ axis).

Moreover, two-phase current command values on the γδ quasi-synchronous coordinate system are obtained by converting the torque command value through a command converter 218. The rotor speed estimate, which is one of the output signals from the phase speed estimator 100, is multiplied by the reciprocal of the number of pole pairs _Np , which is a constant value, through a coefficient multiplier 22 and converted into a mechanical speed estimate, and then sent to the speed controller 220.

Figure 6 shows an example in which a speed control system is configured, so a torque command value is obtained as the output of the speed controller 220. Note that if the control purpose is torque control and a speed control system is not configured, the speed controller 220 is not necessary. In this case, the torque command value is applied directly from the outside.

(IV) Effects of the correlation signal generation method, phase speed estimation device, and sensorless vector control device according to an embodiment of the present invention

The above explanation has shown a correlation signal generation method targeted at the carrier high frequency voltage application method using shaft element amplitude extraction. In this correlation signal generation method, four intermediate signals corresponding to four shaft element amplitudes are synthesized as a preliminary step to synthesizing the correlation signal, and these are then used to synthesize positive phase-to-phase signals. When synthesizing the intermediate signals, the selection of two arbitrary parameters makes it possible to synthesize positive correlation signals with a wide variety of characteristics, while in the carrier high frequency voltage application method, it is also possible to synthesize positive correlation signals with characteristics equivalent to those of the conventional high frequency current amplitude correlation method. In addition, the phase correction function of the intermediate signals makes it unnecessary to carry out the phase correction process that was previously required when extracting shaft element amplitudes.

(V) Other implementation forms

The above-described embodiment is merely an example, and the present invention can be implemented in various other forms. In other words, the present invention is not limited to the above-described embodiment, and various omissions, substitutions, and modifications can be made as appropriate without departing from the gist of the present invention.

That is, in the above, a detailed description has been given of the case where the present invention is applied to a permanent magnet synchronous motor as an example of an embodiment of the correlation signal generation method, phase speed estimation device, and sensorless vector control device according to the present invention.

However, the configurations of the correlation signal generation method, phase velocity estimation device, and sensorless vector control device according to the present invention are not limited to the embodiments described above, and various omissions, substitutions, or modifications can be made as appropriate without departing from the spirit of the present invention.

In the above, in order to facilitate understanding of the present invention, a permanent magnet synchronous motor is used as an example of an AC motor, and the correlation signal generation method, phase speed estimation device, and sensorless vector control device related thereto are described in detail. However, the correlation signal generation method, phase speed estimation device, and sensorless vector control device of the present invention are not limited to application to permanent magnet synchronous motors, and can be applied to any AC motor (in addition to permanent magnet synchronous motors, there are, for example, synchronous reluctance motors and induction motors) in which the rotor exhibits salient pole characteristics when a high-frequency voltage with a frequency higher than the driving frequency is applied.

The present invention can be used to control AC motors whose rotors exhibit salient pole characteristics when high-frequency voltages with frequencies higher than the driving frequency are applied.

10 Rotor
100 Phase-speed estimator
102 High-Frequency Voltage Commander (HFVC) (high-frequency voltage application means)
104 Correlation signal generator (correlation signal generating means)
106 Phase synchronizer (estimated value generation means)
108 Band-pass filter
110 Phase compensator (phase compensation means)
112 Amplitude Extractor (amplitude extraction means)
114 Correlation signal synthesizer (correlation signal synthesizer means)
200 Sensorless vector control device 202 Three-phase to two-phase converter 204 Two-phase to three-phase converter 206 High-frequency component removal filter (digital filter for removing high-frequency components)
208 Power converter (inverter)
210 Discrete-time current detector 212 Inverse vector rotator 214 Vector rotator 216 Current controller
218 Command converter
220 Speed controller
222 Coefficient multiplier 224 Cosine-sine signal generator 300 Permanent magnet synchronous motor (SM)

Claims (7)

A correlation signal generating method used to estimate a rotor phase speed in an AC motor in which a rotor exhibits salient pole characteristics in response to application of a high frequency voltage having a frequency higher than a driving frequency, comprising:
In a correlation signal generation method in a carrier high frequency voltage application method using axial element amplitude extraction,
Extracting the shaft element amplitude contained in the high frequency current based on the following formula (26a) and the following formula (26b),
Using the extracted axis element amplitude, an intermediate signal is synthesized based on the following equations (28a) to (28d):
Using the synthesized intermediate signal,
p _c =f(C _2γ , S _2γ , S _2δ , C _2δ )
and generating a positive correlation signal p _{c based on the correlation signal p c} .

A phase speed estimation device used to estimate a rotor phase speed in an AC motor in which a rotor exhibits salient pole characteristics in response to application of a high frequency voltage having a frequency higher than a driving frequency,
A high frequency voltage application means for generating discrete high frequency voltage command values;
An axis element amplitude included in the high frequency current is extracted based on the following formula (26a) and the following formula (26b), and an intermediate signal is synthesized based on the following formulas (28a) to (28d) using the extracted axis element amplitude, and the synthesized intermediate signal is used to
p _c =f(C _2γ , S _2γ , S _2δ , C _2δ )
A correlation signal generating means for generating a positive correlation signal p _c based on
and an estimate generating means for generating a rotor phase estimate and a rotor speed estimate using the positive correlation signal p _c generated by the correlation signal generating means.

The phase velocity estimation device according to claim 2, further comprising:
and a phase compensation means for correcting a steady-state phase deviation. The phase velocity estimation device according to claim 2 or 3, further comprising:
A phase velocity estimation device comprising: a bandpass filter that extracts high-frequency components. A sensorless vector control device controls the drive of an AC motor whose rotor exhibits salient pole characteristics in response to application of a high-frequency voltage having a higher frequency than a drive frequency,
A sensorless vector control device comprising the phase speed estimation device according to claim 2. A sensorless vector control device controls the drive of an AC motor whose rotor exhibits salient pole characteristics in response to application of a high-frequency voltage having a higher frequency than a drive frequency,
A sensorless vector control device comprising the phase speed estimation device according to claim 3. A sensorless vector control device controls the drive of an AC motor whose rotor exhibits salient pole characteristics in response to application of a high-frequency voltage having a higher frequency than a drive frequency,
A sensorless vector control device comprising the phase speed estimation device according to claim 4.

Images