Multi-stage motor initial position estimation method based on excitation signal square wave modulation

Abstract

This invention relates to a method for estimating the initial position of a multi-stage motor based on square wave modulation of an excitation signal. The excitation voltage is applied as a carrier signal to the stator winding of the exciter. By inducing three-phase low-frequency components in the stator winding of the main motor, these components are converted into d-axis current. The variance of the d-axis current changes with the position angle and reaches its maximum value at the corresponding actual d-axis angle, thus obtaining the initial rotor position θ without magnetic polarity. 0_est1 , θ 0_est1 Used for converting three-phase induced current to d-axis current before and after excitation power failure, and adjusting θ based on the magnitude of the d-axis current before and after power failure. 0_est1 Polarity identification is performed. The method does not involve other complex filtering, phase-locked loop and other processing, so the requirements for processor performance are low. At the same time, it does not require secondary identification of magnetic poles. It uses low-frequency induction signal as the data source for rotor initial position estimation. The damping winding, exciter winding and main motor excitation winding have small attenuation of low-frequency induction signal, so it has a high signal-to-noise ratio.

Description

Technical Field

[0001] This invention belongs to the field of motor technology and relates to a method for estimating the initial position of a multi-stage motor based on square wave modulation of the excitation signal. Specifically, it relates to a method for estimating the initial position of the rotor of a multi-stage brushless electrically excited synchronous motor based on square wave modulation of the excitation signal, which does not depend on the salient polarity of the motor, does not require secondary identification of magnetic poles, and is unaffected by the characteristics of the main motor damping winding and the excitation frequency characteristics of the exciter. Background Technology

[0002] Three-stage brushless synchronous motors offer advantages such as high power quality and reliability, and are widely used as generators in aircraft power systems. By operating the three-stage brushless synchronous motor in electric mode to start the aircraft engine, and then having the engine drive it in generator mode to supply power to onboard electrical equipment after starting, the three-stage brushless synchronous motor achieves integrated starting and generating. This eliminates the need for a dedicated engine starter, reduces system size and weight, and increases integration, which is of great significance to aircraft power systems.

[0003] Accurate rotor position information is required when a three-stage brushless synchronous motor drives an aero-engine in electric mode. Traditional mechanical position sensors for acquiring rotor position suffer from low reliability and increased system size and weight. Online rotor position estimation for a three-stage brushless synchronous motor can eliminate the need for mechanical position sensors, enabling sensorless start-up control. This improves system reliability and reduces size and weight. Furthermore, the accuracy of the rotor's initial position directly affects the magnitude of the starting current and even the success of the start-up; therefore, research on rotor initial position estimation is crucial.

[0004] A three-stage brushless synchronous motor consists of a coaxially mounted permanent magnet auxiliary exciter, an exciter, and a main motor. Its structural diagram is shown below. Figure 1 As shown. In some applications, the permanent magnet auxiliary exciter can be omitted, forming a two-stage brushless synchronous motor. Since the core components and operating principles are basically the same, three-stage and two-stage brushless synchronous motors are collectively referred to as multi-stage brushless synchronous motors.

[0005] Traditional methods for estimating the initial rotor position at zero speed and stationary conditions primarily utilize the salient polarity or magnetic saturation characteristics of the motor, injecting an auxiliary voltage signal and extracting the response signal containing rotor position information to estimate the rotor position. The method using the motor's salient polarity for initial rotor position estimation at zero speed and stationary conditions also includes identifying the rotor's magnetic pole direction. The main steps include: 1. Injecting a rotating or high-frequency signal; 2. Extracting the response signal using a bandpass filter; 3. Signal demodulation; 4. Obtaining the signal envelope using a low-pass filter; 5. Estimating the rotor position using a phase-locked loop or arctangent method; 6. A secondary identification of the magnetic pole polarity is required during initial rotor position estimation. The method utilizing magnetic saturation characteristics mainly includes: 1. Injecting pulse voltages into the main stator windings at fixed intervals of electric arc within the range of 0-2π, and collecting and storing the response current at the end of the pulse voltage injection; 2. Performing fitting and smoothing processing on the stored data; 3. Selecting the angular position corresponding to the maximum value point as the initial rotor position. This method processes a large amount of data, and the estimation accuracy is significantly affected by the electric arc interval between the two pulse injections and the accuracy of the current pulse response detection. In addition, methods based on high-frequency signal injection from the main motor stator to the exciter stator signal detection, and vice versa, have also been studied to some extent. However, these methods based on high-frequency signal injection are not only complex in processing, but also face some common problems with methods based on motor salient polarity. That is, none of them consider the high-frequency filtering characteristics of the damping winding, exciter winding, and main motor excitation winding, which limits their application. When these factors are considered, the signal-to-noise ratio is low due to the severe attenuation of the high-frequency signal amplitude, which poses a risk of estimation failure.

[0006] As can be seen from the above, the traditional method for identifying the initial position of the rotor has the following disadvantages: 1. The method requires a large amount of data processing and storage, which places high demands on the processor; 2. The magnetic pole direction needs to be identified again in the zero-speed stationary state; 3. When considering the high-frequency filtering characteristics of the damping winding, exciter winding and main motor excitation winding, the signal-to-noise ratio is severely attenuated, and there is a risk of failure in the initial position estimation. Summary of the Invention

[0007] Technical problems to be solved

[0008] To avoid the shortcomings of existing technologies, this invention proposes a multi-stage motor initial position estimation method based on square wave modulation of excitation signal. This method overcomes the deficiencies of existing rotor initial position estimation techniques, such as reliance on salient polarity, large data processing volume, and the need for secondary identification of magnetic poles, especially when considering the high-frequency filtering characteristics of damping windings, exciter windings, and main motor excitation windings.

[0009] Technical solution

[0010] A method for estimating the initial position of a multi-stage motor based on square-wave modulation of an excitation signal, characterized in that: the multi-stage brushless electro-excited synchronous motor includes a main motor and a two-phase exciter coaxially installed, and the rotor position of the multi-stage brushless electro-excited synchronous motor refers to the rotor position of the main motor in the multi-stage brushless electro-excited synchronous motor; the steps of the estimation method are as follows:

[0011] Step 1: Apply a symmetric excitation voltage modulated by a two-phase low-frequency square-wave modulation signal to the stator windings of the two-phase exciter, where the period of the low-frequency square-wave modulation signal is T n1 , MMB_2 , MMC_1 , MMA_1 , d0 , <​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​,I d0_2 ...I d0_n1 After n = 1, θ1 = 1 * Δθ, the three-phase induced currents I1, I2…I in group n1 are... n1 Transform each data point to a virtual d-axis and denote the transformed data set as I. d1 =[I d1_1 ,I d1_ 2...I d1_n1 ], ... Finally, n = n² - 1, The three-phase induced currents I1, I2…In1 are transferred to the n1 group. n1 Transform each data point to the virtual d-axis, and denote the transformed data as follows:

[0015] Step 5: Calculate each set of data The variance, and denoted as

[0016] Step 6: Search The maximum value v in dk k∈[0, n²-1], then θ 0est1 =kΔθ is the initial position of the rotor without magnetic polarity;

[0017] Step 7: Disconnect the two-phase symmetrical excitation voltage modulated by a low-frequency square wave signal applied to the stator windings of the two-phase exciter mentioned in Step 1, and record the three-phase induced current of the main motor stator at two acquisition times before and after disconnection, respectively, and record them as two sets of data I. MMA1 I MMB1 I MMC1 and I MMA2 I MMB2 and I MMC2 ;

[0018] Step 8: Use θ from step 6 0est1 =kΔθ is the transformation angle, which represents the two sets of currents I before and after disconnecting the symmetrical excitation voltage in step 7. MMA1 I MMB1 I MMC1 and I MMA2 I MMB2 and I MMC2 Transform to the d-axis and denote them as I. MMd1 and I MMd2 And according to the following rules, for θ without magnetic polarity 0est1 Processing:

[0019] If I MMd1 MMd2 , then θ 0est1 =θ 0est1 ,

[0020] If I​MMd1 >I MMd2 , then θ 0est1 =θ 0est1 +π;

[0021] Step 9: The final estimated initial position of the main motor is θ. 0est =θ 0est1 .

[0022] The expression for the symmetrical excitation voltage modulated by the low-frequency square wave modulation signal in the two phases is:

[0023]

[0024] Among them, U es This is the maximum voltage value. ω represents the angular velocity of the excitation carrier voltage, t represents time, k represents the modulation coefficient of the square wave modulation signal, and n represents the nth period.

[0025] The expression for the fundamental current with the same excitation frequency induced in the three-phase windings of the exciter rotor by the symmetrical excitation voltage modulated by the two-phase low-frequency square wave modulation signal is:

[0026]

[0027] Among them, I esr This represents the maximum induced current when the modulation coefficient k = 1. The signal transmission phase delay caused by inductive reactance is given by τ, which is the equivalent time constant of the exciter rotor winding at the low-frequency square wave frequency. sgn(k) is the sign function, which is 1 when k>0 and -1 when k<0.

[0028] The expression for the current in the main motor excitation winding is:

[0029]

[0030] Among them, I dc I is the DC component value of the induced current when the modulation coefficient k = 1. ωes_6v v = 2, 3... represents the amplitude of the harmonic current signal of the excitation carrier voltage signal in the main motor. The phase delay of the excitation carrier voltage signal in the harmonic current signal of the main motor is denoted as .

[0031] The expression for the current induced along the d-axis of the main motor stator winding is:

[0032]

[0033] Where K is the gain ratio resulting from the stator and rotor winding structure of the main motor. For time differential operators, Id_MM This represents the amplitude of the induced current.

[0034] Beneficial effects

[0035] This invention proposes a method for estimating the initial position of a multi-stage motor based on square wave modulation of the excitation signal. By using the excitation voltage as a carrier signal and modulating it with a low-frequency square wave before applying it to the exciter stator winding, the high inductance and low-pass filtering characteristics of the exciter and main motor excitation windings effectively filter out the harmonic components of the six-times excitation voltage carrier signal on the main motor excitation winding, leaving only the low-frequency modulation signal frequency and low-order harmonic components. This further induces three-phase low-frequency components in the main motor stator winding through the mutual inductance between the stator and rotor. When these components are transformed to the actual d-axis, the corresponding current fluctuation will be maximized, and the actual q-axis current fluctuation will be minimized. Therefore, by increasing the position information interval for rotational transformation by a fixed angle and obtaining the corresponding d-axis current, the d-axis current variance will change with the position angle, reaching its maximum value at the corresponding actual d-axis angle. This yields the rotor's initial position θ, which is free of magnetic polarity. 0_est1 Finally, θ 0_est1 Used for converting three-phase induced current to d-axis current before and after excitation power failure, and adjusting θ based on the magnitude of the d-axis current before and after power failure. 0_est1 Polarity identification is performed. This method only requires simple mathematical processing such as equal-interval sampling and variance calculation of the three-phase induced current, as well as logical operations such as finding the maximum value. It does not involve other complex processing such as filtering and phase-locked loops, so the requirements for processor performance are low. At the same time, it does not require secondary identification of magnetic poles. Finally, since the low-frequency signal is used as the modulation wave and the low-frequency induced signal is ultimately used as the data source for rotor initial position estimation, the damping winding, exciter winding and main motor excitation winding have small attenuation of the low-frequency induced signal, thus having a high signal-to-noise ratio.

[0036] The beneficial effects of this invention include:

[0037] 1) Traditional methods for estimating the initial rotor position, which rely on the salient polarity of the main motor, require injecting a high-frequency rotating signal into the stator side of the main motor, extracting the signal using a filter, multiplying it with a signal of the same injection frequency for demodulation, and then sending it into a phase-locked loop to obtain the initial rotor position without magnetic polarity. Finally, additional pulses are needed to identify the magnetic polarity for a second time. This process is not only complex in terms of parameter matching, but also places high demands on the performance of the controller. This invention only requires low-frequency modulation of the excitation voltage signal and its application to the exciter stator winding. After collecting the induced current in the main motor stator winding, the maximum variance is calculated according to the steps shown. Based on the relationship between the current magnitude after transforming the two sampling values ​​before and after the excitation is cut off to the d-axis, the initial rotor position of the main motor can be obtained in one go. This does not require complex parameter matching and has lower requirements on processor performance.

[0038] 2) Traditional rotor initial position estimation methods based on magnetic saturation characteristics require injecting signals into the stator windings of the main motor at fixed intervals and storing the response current. Then, the stored data is fitted and the rotor initial position is estimated by finding the maximum value. This method has high requirements for processor storage and computing power. The method shown in this invention requires extremely limited storage space and does not require numerical fitting.

[0039] 3) The present invention applies the excitation voltage after low-frequency modulation to the stator winding of the exciter, so that the required effective signal is less affected by the exciter, the main motor excitation winding and the damping winding, and the signal-to-noise ratio is higher.

[0040] 4) In the actual measurement, the main motor rotor was adjusted every 10°, and the estimation was performed using the method proposed in this invention. The estimation error under different actual positions is as follows: Figure 7 As shown, the estimation errors are all within 2°, which meets the requirement of within 10° in this field;

[0041] 5) In actual measurements, based on the high-frequency injection method, the signal attenuation reached 12.4dB and 10.3dB respectively when the frequencies were 1200Hz and 400Hz, according to the proportion of the DC signal. However, the signal attenuation of the method shown in this invention is almost negligible, showing a higher signal-to-noise ratio. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of a three-stage brushless synchronous motor with damping windings.

[0043] Figure 2 This is a flowchart of the method for estimating the initial position of the rotor of a multi-stage brushless synchronous motor proposed in this invention.

[0044] Figure 3 This is the excitation voltage signal modulated by a low-frequency square wave according to the present invention.

[0045] Figure 4 This is the response current waveform of the main motor excitation winding of the present invention.

[0046] Figure 5 The virtual angle θ of this invention n Current fluctuation diagrams of the d-axis and q-axis as the current increments from 0 to π.

[0047] Figure 6 This is a graph showing the trend of d-axis current variation of the main motor without magnetic pole polarity when the excitation is disconnected at different actual positions according to the present invention.

[0048] Figure 7 This invention provides the estimation error under different actual locations measured in actual measurements. Detailed Implementation

[0049] The present invention will now be further described in conjunction with the embodiments and accompanying drawings:

[0050] In this embodiment, the exciter of the multi-stage brushless synchronous motor is a two-phase exciter, meaning the exciter stator windings are two-phase windings with a 90° electrical angle difference. The exciter has 6 pole pairs, and the main motor has 3 pole pairs. The following description uses the actual position of the main motor as... and Let's take an example to explain in detail.

[0051] 1. A two-phase symmetrical excitation voltage modulated by a low-frequency square wave signal is applied to two phase windings of the exciter stator (denoted as the α-phase winding and the β-phase winding, respectively), wherein the period of the low-frequency square wave modulation signal is T. L =100ms, take k=-0.5, and the excitation carrier voltage period is T. esf =5ms, excitation carrier voltage amplitude U es =50V, then:

[0052]

[0053] 2. After the main motor excitation current stabilizes, at fixed intervals T sample = Real-time acquisition of the three-phase induced current of the main motor stator connected to the three-phase inverter in 1ms. The three-phase induced current is denoted as I. MMA I MMB and I MMC Then a low-frequency square wave period T L The number of sampling points within the time period is n1 = T L / T sample =100, and the set of three-phase currents corresponding to each sampling point is denoted as I1 = [I MMA_1 ,I MMB_1 ,I MMC_1 ],I2=[I MMA_2 ,I MMB_2 ,I MMC_2 ...I 100 =[I MMA_100 ,I MMB_100 ,I MMC_100 ];

[0054] 3. Set Δθ=pi / 60, n2=π / Δθ=60;

[0055] 4. Starting from n = 0, θ0 = 0 * Δθ, set n1 = 100 sets of three-phase induced currents I1, I2…I 100 Transform each data point to a virtual d-axis and denote the transformed data set as I. d0 =[I d0_1 ,I d0_2 ...I d0_100After n = 1, θ1 = 1 * Δθ, n1 = 100 sets of three-phase induced currents I1, I2…I 100 Transform each data point to a virtual d-axis and denote the transformed data set as I. d1 =[I d1_1 ,I d1_2 ...I d1_100 ], ... Finally, n = n² - 1 = 59, θ 59 =59*Δθ, which will be the number of three-phase induced currents I1, I2…I1 = 100. n1 Transform each data point to a virtual d-axis and denote the transformed data set as I. d59 =[I d59_1 ,I d59_2 ...I d59_100 The transformation formula used to convert the three-phase induced current to the virtual d-axis is:

[0056]

[0057] 5. Calculate I for each set of data. d0 I d1 …I d59 The variance, denoted as v d0 v d1 …v d59 ;

[0058] 6. Sort and search v d0 v d1 …v d59 The maximum value v in dk k∈[0, n²-1], when the actual initial positions are respectively and When, find the largest element that is v d20 and v d15 ,but and These are the initial positions of the rotor without magnetic polarity;

[0059] 7. Disconnect the two-phase symmetrical excitation voltage modulated by a low-frequency square wave signal applied to the stator windings of the two-phase exciter mentioned in step 1, and record the three-phase induced current of the main motor stator at two acquisition times before and after disconnection, denoted as I. MMA1 I MMB1 I MMC1 and I MMA2 I MMB2 and I MMC2 ;

[0060] 8. When At that time, the estimated position without magnetic polarity is used. To change the angle, the two sets of currents I before and after disconnecting the symmetrical excitation voltage in step 7 are...MMA1 I MMB1 I MMC1 and I MMA2 I MMB2 and I MMC2 Transform to the d-axis and denote them as I. MMd1 and I MMd2 And there is I MMd1 MMd2 ,but

[0061] when At that time, the estimated position without magnetic polarity is used. To change the angle, the two sets of currents I before and after disconnecting the symmetrical excitation voltage in step 7 are... MMA1 I MMB1 I MMC1 and I MMA2 I MMB2 and I MMC2 Transform to the d-axis and denote them as I. MMd1 and I MMd2 And there is I MMd1 >I MMd2 ,but

[0062] 9: The estimated final initial position values ​​of the main motor are as follows: and ​

Claims

1. A method for estimating the initial position of a multi-stage motor based on square wave modulation of an excitation signal, characterized in that: A multi-stage brushless electrically excited synchronous motor includes a coaxially mounted main motor and a two-phase exciter. The rotor position of a multi-stage brushless electrically excited synchronous motor refers to the rotor position of the main motor within the multi-stage brushless electrically excited synchronous motor. The estimation method and steps are as follows: Step 1: Apply two-phase symmetrical excitation voltages modulated by a low-frequency square wave signal to the stator windings of the two-phase exciter. The period of the low-frequency square wave modulation signal is... The excitation carrier voltage period is ,and Then, the symmetrical excitation voltage modulated by the low-frequency square wave modulation signal will induce the fundamental current and high-order harmonic current with the same excitation frequency in the three-phase winding of the exciter rotor. After the three-phase winding current passes through the rotating rectifier, it forms the excitation winding current of the main motor. Due to the low-frequency modulation effect, the high-order harmonics are filtered out, and the low-frequency pulsation will induce current in the d-axis of the stator winding of the main motor. Step 2: After the main motor excitation current stabilizes, at fixed intervals... Real-time acquisition of the three-phase induced current of the main motor stator connected to the three-phase inverter. The three-phase induced currents are denoted as follows: , and Then one low-frequency square wave period The number of sampling points within the time period is The three-phase current corresponding to each sampling point is denoted as follows: , … ; Step 3: Set the virtual angle used during the transformation to... And for Integers, where , The value can be freely set, and its size will affect the estimation accuracy and estimation time. Step 4: From start, ,Will Group three-phase induced current , … Transform to virtual respectively The axis is defined, and the transformed set of data is denoted as... After that , ,Will Group three-phase induced current , … Transform to virtual respectively The axis is defined, and the transformed set of data is denoted as... ,…at last , ,Will Group three-phase induced current , … Transform to virtual respectively The axis is defined, and the transformed set of data is denoted as... ; Step 5: Calculate each set of data , … The variance, and denoted as , … ; Step 6: Search , … The maximum value in ,but That is, the initial position of the rotor without magnetic polarity; Step 7: Disconnect the two-phase symmetrical excitation voltage modulated by a low-frequency square wave signal applied to the stator windings of the two-phase exciter mentioned in Step 1, and record the three-phase induced current of the main motor stator at two acquisition times before and after disconnection, respectively, and record them as two sets of data. , , and , and ; Step 8: Use the method described in step 6 To change the angle, the two sets of currents before and after disconnecting the symmetrical excitation voltage in step 7 are respectively... , , and , and Transform to Axis, and respectively denoted as and And according to the following rules for non-magnetic polarity Processing: if ,So , if ,So ; Step 9: The final estimated initial position of the main motor is... .

2. The method for estimating the initial position of a multi-stage motor based on square wave modulation of excitation signal according to claim 1, characterized in that: The expression for the symmetrical excitation voltage modulated by the low-frequency square wave modulation signal in the two phases is: (1) in, This is the maximum voltage value. The excitation carrier voltage angular velocity, For time, The modulation coefficients of the square wave modulated signal are... Indicates the first One cycle.

3. The method for estimating the initial position of a multi-stage motor based on square wave modulation of excitation signal according to claim 1, characterized in that: The expression for the fundamental current with the same excitation frequency induced in the three-phase windings of the exciter rotor by the symmetrical excitation voltage modulated by the two-phase low-frequency square wave modulation signal is: （2） in, Modulation coefficient The maximum value of the induced current at that time The signal transmission phase delay is caused by inductive reactance. This is the equivalent time constant of the exciter rotor winding at a low-frequency square wave frequency. For a sign function, when hour, ,when hour, .

4. The method for estimating the initial position of a multi-stage motor based on square wave modulation of excitation signal according to claim 1, characterized in that: The expression for the current in the main motor excitation winding is: (2) in, Modulation coefficient The DC component of the induced current at that time This represents the amplitude of the harmonic current signal of the excitation carrier voltage signal in the main motor. The phase delay of the excitation carrier voltage signal in the harmonic current signal of the main motor is denoted as .

5. The method for estimating the initial position of a multi-stage motor based on square wave modulation of excitation signal according to claim 1, characterized in that: The expression for the current induced along the d-axis of the main motor stator winding is: (3) in The gain ratio is determined by the structure of the main motor's stator and rotor windings. For time differential operators, This represents the amplitude of the induced current.

Images