A three-stage synchronous machine wide speed range normalized rotor position estimation method

Abstract

The application discloses a three-stage synchronous motor wide-speed-range normalized rotor position estimation method. The three-stage brushless alternating-current synchronous motor combines an alternating-current excitation indirect injection of a second harmonic rotor position estimation method in a zero-low-speed stage and a back electromotive force analytical calculation method in a medium-high-speed stage. The two times frequency signals containing rotor position information extracted by the two position estimation methods are normalized, and a hybrid position estimation error signal is formed after normalization processing. The motor speed and position angle estimation information are calculated, and the rotor position estimation in a wide-speed-range is realized. The position estimation method in the wide-speed-range disclosed by the application moves the composite process of the two position estimation methods forward, can effectively avoid the oscillation generated when the rotor position is zero in the switching speed interval, and improves the position estimation precision in the wide-speed-range.

Description

Technical Field

[0001] This invention relates to the field of motor control, and more particularly to a normalized rotor position estimation method for a three-stage synchronous motor with a wide speed range. Background Technology

[0002] Three-stage brushless synchronous motors are currently the most widely used aircraft starter generators, offering advantages such as high reliability and ease of maintenance. However, the method of using mechanical sensors to obtain the motor rotor position has many problems, such as increased cost, reduced motor operating accuracy, and low utilization rate. Therefore, research into sensorless technology is necessary.

[0003] Currently, positionless control methods applicable to three-stage brushless synchronous motors can be divided into two categories: (1) position estimation methods based on harmonic injection; and (2) position estimation methods based on fundamental wave models. The position estimation method based on harmonic injection estimates the rotor position by injecting, extracting, and analyzing signals, and is suitable for sensorless control in the zero-speed stage. The position estimation method based on fundamental wave models estimates the rotor position by calculating or observing the flux linkage, back EMF, etc. in the fundamental wave model of the motor, and is suitable for sensorless control in the medium-speed stage.

[0004] In summary, a single control algorithm struggles to accurately estimate rotor position over a wide speed range. Therefore, switching between two position estimation methods can establish a composite position estimation method for a broad speed range. Furthermore, traditional switching methods, which use simple weighted averaging based on speed, may experience oscillations when the rotor position crosses zero, leading to switching failure. Therefore, designing a smooth and stable switching process is also of significant research value. Summary of the Invention

[0005] The purpose of this invention is to provide a normalized rotor position estimation method for a three-stage synchronous motor with a wide speed range, which adopts the following technical solution:

[0006] The three-stage brushless AC synchronous motor combines the rotor position estimation method of indirectly injecting second harmonics through AC excitation in the zero-low speed stage with the back EMF analytical calculation method in the medium-high speed stage. The double-frequency signals containing rotor position information extracted by the two position estimation methods are normalized and then normalized to form a hybrid position estimation error signal. The motor speed and position angle estimation information are calculated to achieve rotor position estimation over a wide speed range.

[0007] The aforementioned indirect second harmonic injection method in the zero-low speed stage includes: the main exciter uses single-phase AC excitation; the nonlinear characteristics of the rotating rectifier generate a second harmonic voltage in the main generator excitation winding, which is directly used as a high-frequency signal indirectly injected into the main generator excitation winding; a bandpass filter extracts the two-phase high-frequency response signal of the main generator in the stationary coordinate system; after mathematical calculation and processing, a low-pass filter extracts a set of two-phase double-frequency signals containing rotor position information.

[0008]

[0009] in,

[0010] The analytical calculation method for back electromotive force in the medium-to-high speed stage includes the following steps:

[0011] (1) Acquire the phase voltage u of the main generator armature winding. a u b u c and phase current i a i b i c The two-phase voltage u in the two-phase stationary coordinate system is obtained by performing Clark transformation. α u β and two-phase current i α i β ;

[0012] (2) The flux linkage in the two-phase stationary coordinate system calculated using the improved second-order generalized integrator can be expressed as:

[0013]

[0014] Among them, e α With e β These are the back electromotive forces in a two-phase stationary coordinate system, R. s The resistance of the main generator armature winding.

[0015] (3) The magnetic flux linkage in the two-phase stationary coordinate system can be analytically calculated as follows:

[0016]

[0017] Among them, L q Main generator quadrature axis inductance, L d Main generator direct-axis inductance, M df For the mutual inductance between the excitation winding and the direct-axis winding, M dD For the mutual inductance between the direct-axis winding and the damping winding, i f For the excitation current, i D This represents the damping winding current.

[0018] Obtain a set of signals containing rotor position information:

[0019]

[0020] Among them, K θ =(L d -L q )i d +M df i f +M dD i D .

[0021] (4) The two-phase double-frequency signal containing rotor position information is obtained by analysis as follows:

[0022]

[0023] in,

[0024] Furthermore, the double-frequency signals containing rotor position information extracted by the two methods are normalized by dividing the two-phase double-frequency signals obtained in the zero-low speed stage and the medium-high speed stage by the signal amplitude, resulting in:

[0025]

[0026]

[0027] Furthermore, the rotor position estimation over the wide speed range includes the following steps:

[0028] (1) Two sets of normalized double-frequency signals were obtained by using the indirect second harmonic injection method in the zero-low speed stage and the back EMF analytical calculation method in the medium-high speed stage. The position angle was estimated by the output of the phase-locked loop, and their position error signals sin(2Δθ1) and sin(2Δθ2) were obtained respectively.

[0029] (2) The two position error signals are weighted and averaged to obtain the mixed position estimation error signal, which is:

[0030] f(Δθ)=ε·sin(2Δθ1)+(1-ε)·sin(2Δθ2)

[0031] Where ε is the weighted average coefficient related to the estimated rotational speed.

[0032] (3) The mixed position estimation error signal is converged to 0 by a PI controller. When the corresponding Δθ approaches 0, sin(2Δθ)≈2Δθ, the rotor estimated position is obtained, completing the rotor position estimation over a wide speed range. Furthermore, when the estimated speed satisfies... Under certain conditions, the indirect second harmonic injection method in the zero-low speed stage is adopted; when the estimated rotational speed meets the requirements... Under certain conditions, the back electromotive force is calculated analytically during the medium-to-high speed stage; when the estimated rotational speed meets the requirements... Under certain conditions, the two position estimation methods are combined. As the estimated rotational speed increases, the estimation accuracy of the indirect second harmonic injection method decreases, and its weight decreases accordingly until it reaches 0. Conversely, the estimation accuracy of the back electromotive force analytical calculation method increases, and its weight increases accordingly until it reaches 1. (ω) L ,ω H This refers to switching the speed range, where ω L To switch the lower limit of the speed range, ω H To switch the upper limit of the speed range.

[0033] By adopting the above solution, the present invention has the following beneficial effects:

[0034] (1) The position estimation method under a wide speed range given by the present invention moves the process of combining the two position estimation methods forward, which can effectively avoid the oscillation caused when the rotor position crosses zero in the switching speed range and improve the position estimation accuracy over a wide speed range.

[0035] (2) The method of the present invention only detects the induced current on the stator side of the main generator and the voltage of the stator armature winding. It has a simple structure, is easy to implement, and reduces the complexity of system control. Attached Figure Description

[0036] Figure 1 A system structure diagram of an aviation three-stage synchronous motor;

[0037] Figure 2 This is a block diagram illustrating the starting control principle of a three-stage synchronous motor for aviation based on the method of this invention.

[0038] Figure 3 Block diagram of the improved second-order generalized integrator;

[0039] Figure 4 Simulation waveform of the output flux linkage signal of the improved second-order generalized integrator;

[0040] Figure 5 (a) is a two-phase double-frequency signal of the indirect second harmonic injection method in the zero low-speed stage;

[0041] Figure 5 (b) is the two-phase double-frequency harmonic signal calculated by the analytical method of back EMF in the medium-to-high speed stage;

[0042] Figure 6 (a) is a schematic diagram showing the actual rotor position and the estimated rotor position;

[0043] Figure 6 (b) is a schematic diagram of rotor position estimation error;

[0044] Figure 6 (c) is the simulated waveform of the rotational speed. Detailed Implementation

[0045] The present invention will now be further described in conjunction with the embodiments and accompanying drawings.

[0046] Figure 1 This is a system structure diagram of an aviation three-stage synchronous motor. The permanent magnet auxiliary exciter does not participate in the starting process. The stator side of the main exciter is powered by a single-phase AC source for single-phase AC excitation, and the rotor side outputs three-phase AC power which is rectified by a rotating rectifier to provide DC excitation for the main generator.

[0047] Figure 2 This is a block diagram of the starting control of a three-stage synchronous motor for aviation based on the method of this invention. The control method is vector control, and the estimated rotor position angle is used to realize coordinate transformation. In the sensorless rotor position measurement process of the three-stage synchronous motor for aviation, the rotor position estimation method of AC excitation indirect injection of second harmonics is used in the zero-low speed stage, and the rotor position is estimated by back electromotive force analytical calculation method in the medium-high speed stage. The double frequency signals containing rotor position information extracted by the two position estimation methods are normalized and then normalized to form a hybrid position estimation error signal. The motor speed and position angle estimation information are calculated to realize rotor position estimation over a wide speed range. The specific steps are as follows:

[0048] Step 1: Rotor position estimation method using AC excitation indirect injection of second harmonics in the zero-low speed stage

[0049] Reference Figure 2 The main exciter uses single-phase AC excitation. The nonlinear characteristics of the rotating rectifier cause a second harmonic voltage to be generated in the excitation winding of the main generator, which is:

[0050]

[0051] Among them, u n Let ω be the amplitude of the 2nth harmonic voltage. ex The excitation angular frequency of the main exciter. The phase of the 2nth harmonic;

[0052] The second harmonic voltage can be directly used as a high-frequency signal indirectly injected into the main generator excitation winding. Therefore, the high-frequency response signal containing rotor position information generated in the main generator armature winding can be expressed in a two-phase stationary coordinate system as follows:

[0053]

[0054] Among them, u h The amplitude of the high-frequency response signal. The phase of the response signal is θ, where θ is the rotor position angle of the motor.

[0055] The two-phase high-frequency response signal of the main generator in the stationary coordinate system is extracted by using a bandpass filter, and the signal f1 is obtained by calculating the square difference of the two-phase voltages.

[0056]

[0057] The signal f2 is obtained by multiplying the product of the two-phase voltages by two:

[0058]

[0059] A set of two-phase, double-frequency signals containing rotor position information is extracted using a low-pass filter:

[0060]

[0061] in,

[0062] Step 2: Analytical calculation of back electromotive force is used in the medium-to-high speed stage.

[0063] Reference Figure 2 Collect the phase voltage u of the main generator armature winding. a u b u c and phase current i a i b i c The two-phase voltage u in the two-phase stationary coordinate system is obtained by performing Clark transformation. α u β and two-phase current i α i β .

[0064] Reference Figure 3 An improvement to the second-order generalized integrator is proposed, which combines integration and high-pass filtering functions. The flux linkage in the two-phase stationary coordinate system calculated using the improved second-order generalized integrator can be expressed as:

[0065]

[0066] Among them, e α With e β These are the back electromotive forces in a two-phase stationary coordinate system, R. s The resistance of the main generator armature winding.

[0067] The analytical calculation of flux linkage in a two-phase stationary coordinate system can be expressed as:

[0068]

[0069] Among them, L q Main generator quadrature axis inductance, Ld Main generator direct-axis inductance, M df For the mutual inductance between the excitation winding and the direct-axis winding, M dD For the mutual inductance between the direct-axis winding and the damping winding, i f For the excitation current, i D This represents the damping winding current.

[0070] Obtain a set of signals containing rotor position information:

[0071]

[0072] Among them, K θ =(L d -L q )i d +M df i f +M dD i D .

[0073] The fundamental frequency signal above is processed and analyzed to obtain a two-phase double-harmonic signal containing rotor position information:

[0074]

[0075] in,

[0076] Step 3: Normalize the signal to generate a mixed position estimation error signal and estimate the rotor position.

[0077] The double-frequency signals containing rotor position information obtained by the above two methods are normalized by dividing the two-phase double-frequency signals by the signal amplitude, resulting in:

[0078]

[0079]

[0080] By combining the two sets of normalized double-frequency signals with the estimated position angle output by the phase-locked loop, their position error signals sin(2Δθ1) and sin(2Δθ2) are obtained respectively.

[0081] The two position error signals are weighted and averaged to obtain a mixed position estimation error signal, which is:

[0082] f(Δθ)=ε·sin(2Δθ1)+(1-ε)·sin(2Δθ2)

[0083] Where ε is the weighted average coefficient related to the estimated rotational speed, it can be expressed as:

[0084]

[0085] in, To estimate the rotational speed, ω L To switch the lower limit of the speed range, ω H To switch the upper limit of the speed range.

[0086] The mixed position estimation error signal is converged to 0 by a PI regulator. When the corresponding Δθ approaches 0, sin(2Δθ)≈2Δθ, the rotor estimated position is obtained, thus completing the rotor position estimation over a wide speed range.

[0087] Based on the above scheme, when the estimated rotational speed satisfies Under certain conditions, the indirect second harmonic injection method in the zero-low speed stage is adopted; when the estimated rotational speed meets the requirements... Under certain conditions, the back electromotive force is calculated analytically during the medium-to-high speed stage; when the estimated rotational speed meets the requirements... Under certain conditions, the two position estimation methods are combined. As the estimated rotational speed increases, the estimation accuracy of the indirect second harmonic injection method decreases, and its weight decreases accordingly until it reaches 0. Conversely, the estimation accuracy of the back electromotive force analytical calculation method increases, and its weight increases accordingly until it reaches 1. (ω) L ,ω H This indicates the switching speed range.

[0088] To verify the normalized rotor position estimation method for a three-stage brushless synchronous motor over a wide speed range, a mathematical model of the three-stage brushless synchronous motor and a sensorless control model were built for simulation verification. The operating conditions are as follows: excitation frequency f... ex =200Hz, initial position angle is 0.1rad.

[0089] Figure 4 The waveform of the output flux linkage signal of the improved second-order generalized integrator is shown. The solid line represents the α-axis flux linkage and the dashed line represents the β-axis flux linkage. As can be seen from the waveform, the improved second-order generalized integrator can integrate the back EMF to calculate the flux linkage and effectively filter out the DC bias caused by the integration.

[0090] Figure 5 The simulation waveforms are normalized versions of two sets of two-phase double-harmonic signals containing rotor position information. (a) represents the two-phase double-harmonic signal obtained by the indirect second harmonic injection method in the zero-low speed stage, and (b) represents the two-phase double-harmonic signal obtained by the back EMF analytical calculation method in the medium-high speed stage. In both sets of simulation waveforms, solid lines represent α-axis signals, and dashed lines represent β-axis signals.

[0091] Figure 6 Simulation waveforms for the actual rotor position, estimated rotor position, position estimation error, and rotational speed. Figure 6 In (a), the solid line represents the actual rotor position, and the dashed line represents the estimated rotor position. Figure 6 (b) represents the rotor position estimation error. Figure 6 (c) represents the motor speed. As can be seen from the waveform, the estimated rotor position can quickly converge to the actual value. With the switching speed range set to (150rpm, 180rpm), the rotor position estimation error remains within 0.1rad after starting from 0.46s, and the switching process is smooth and stable, demonstrating good position estimation performance.

[0092] The simulation results of the sensorless control system of the three-stage brushless synchronous motor show that the position estimation method given in this invention has good performance over a wide speed range. It can effectively avoid the oscillation caused when the rotor position crosses zero in the switching speed range and has high position estimation accuracy.

[0093] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solutions based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A method for estimating the normalized rotor position of a three-stage synchronous motor over a wide speed range, characterized in that: The three-stage synchronous motor includes a main exciter, a rotating rectifier, and a main generator arranged coaxially in sequence; The estimation method includes: the three-stage synchronous motor uses the indirect second harmonic injection method to estimate the rotor position in the zero-low speed stage; and uses the back EMF analytical calculation method to estimate the rotor position in the medium-high speed stage; when the three-stage synchronous motor is between the zero-low speed stage and the medium-high speed stage, the two-phase double-frequency signals containing rotor position information obtained by the two methods are normalized respectively, and a hybrid position estimation error signal is formed after normalization. The rotor speed and position estimation of the three-stage synchronous motor are obtained by analytical calculation, thereby realizing rotor position estimation over a wide speed range. The process of estimating rotor position using the indirect second harmonic injection method in the zero-low speed stage is as follows: The excitation winding of the main exciter is a single-phase excitation winding. A single-phase AC power supply with a constant frequency is passed through the single-phase excitation winding, so that a three-phase AC power supply is induced on the rotor side of the main exciter. The three-phase AC power supply is input to the rotating rectifier. The second harmonic generated by the rotating rectifier is used as a high-frequency signal indirectly injected into the main exciter. The two-phase high-frequency response signal of the main generator in the stationary coordinate system is extracted by a bandpass filter. Finally, a set of two-phase double-harmonic signals containing rotor position information is extracted by a low-pass filter. Where K1 is the amplitude of the two-phase double-frequency signal in the zero-low speed stage, θ is the rotor position angle of the main generator, and u h The amplitude of the high-frequency response signal; The process of estimating rotor position using the back electromotive force analytical calculation method in the medium-to-high speed stage is as follows: (1) Based on the collected armature winding voltage and current of the main generator, the flux linkage ψ in the two-phase stationary coordinate system is obtained by analytical calculation. α ψ β : Among them, L q Main generator quadrature axis inductance, L d Main generator direct-axis inductance, M df The mutual inductance between the excitation winding and the direct-axis winding of the main generator, M dD For the mutual inductance between the direct-axis winding and the damping winding, i d Let i be the d-axis current. α i β Let i be the current in a stationary two-phase coordinate system. f For the excitation current, i D For the damping winding current; This yields a set of signals containing rotor position information: Wherein, parameter K θ =(L d -L q )i d +M df i f +M dD i D ; (2) A set of two-phase double-frequency signals containing rotor position information is obtained by analysis: Wherein, parameter K2 is the amplitude of the two-phase double-frequency signal in the medium-to-high speed stage, and θ is the rotor position angle of the main generator.

2. The method for estimating the normalized rotor position of a three-stage synchronous motor over a wide speed range as described in claim 1, characterized in that, The obtained two-phase double-frequency signals are divided by the signal amplitude and then normalized.

3. The method for estimating the normalized rotor position of a three-stage synchronous motor over a wide speed range as described in claim 2, characterized in that, (1) Two sets of normalized two-phase double-frequency signals obtained by the indirect second harmonic injection method in the zero-low speed stage and the back EMF analytical calculation method in the medium-high speed stage are used to estimate the position angle through the phase-locked loop output. The position error signals sin(2Δθ1) and sin(2Δθ2) are obtained, where Δθ1 is the position angle error obtained by the indirect second harmonic injection method and Δθ2 is the position angle error obtained by the back electromotive force analytical calculation method. (2) The position error signals sin(2Δθ1) and sin(2Δθ2) are weighted and averaged to obtain the mixed position estimation error signal f(Δθ), which is: f(Δθ)=ε·sin(2Δθ1)+(1-ε)·sin(2Δθ2) Where ε is the weighted average coefficient; And there are: in, The estimated rotational speed of the main generator rotor, ω L To switch the lower limit of the speed range, ω H To switch the upper limit of the speed range; (3) The mixed position estimation error signal is converged to 0 by the PI regulator. When the corresponding Δθ approaches 0, i.e. sin(2Δθ)≈2Δθ, the rotor estimated position is obtained, and the rotor position estimation under a wide speed range is completed.

4. The method for estimating the normalized rotor position of a three-stage synchronous motor over a wide speed range as described in claim 3, characterized in that, When the estimated rotational speed of the main generator rotor meets the following conditions: At that time, the rotor position is estimated using the indirect second harmonic injection method in the zero-low speed stage; when the estimated speed of the main generator rotor meets the following conditions... At that time, the rotor position is estimated using the back EMF analytical calculation method in the medium-to-high speed stage; when the estimated speed of the main generator rotor satisfies At that time, the rotor position is estimated by combining the indirect second harmonic injection method in the zero-low speed stage and the back EMF analytical calculation method in the medium-high speed stage. As the estimated speed of the main generator rotor increases, the estimation accuracy of the indirect second harmonic injection method decreases, and the weighted average coefficient decreases accordingly until it becomes 0. The estimation accuracy of the back EMF analytical calculation method increases, and the weighted average coefficient increases accordingly until it becomes 1.

Images