A method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection

Abstract

The present invention relates to a multi-stage motor initial position estimation method based on low-frequency square wave signal injection. The method sequentially applies square wave voltage low-frequency voltage signals in four directions equivalent to the α-axis, α-axis, β-axis and β-axis of the stator winding of the main motor, and collects the stator winding current of the two-phase exciter in real time. At the end of the injection of the square wave voltage in the four directions within a low-frequency voltage cycle, the estimated main motor excitation current is sequentially collected and saved, and then the sine and cosine signals containing the main motor rotor position information are extracted. The obtained sine and cosine signals are compensated by the inverse tangent and according to the positive and negative value relationship of the sine and cosine signals, and the initial position of the main motor rotor can be uniquely estimated. The method proposed by the present invention is not affected by the salient polarity of the main motor and the initial position of the exciter rotor, has high estimation accuracy, and the signal extraction and position estimation methods are simple, with low requirements on processor performance.

Description

Technical Field

[0001] The present invention belongs to the field of motor technology and relates to a multi-stage motor initial position estimation method based on low-frequency square wave signal injection, and specifically relates to a multi-stage brushless electrically excited synchronous motor rotor initial position estimation method based on low-frequency square wave signal injection and independent of motor salient polarity and without the need for secondary identification of magnetic poles. Background Art

[0002] Three-stage brushless synchronous motors offer excellent power generation quality and high reliability, and are widely used as generators in aircraft power systems. This allows the motor to operate in a motoring mode to start the aircraft engine. Once started, the engine then drives the motor in a generating mode to power onboard electrical equipment. This achieves integrated starting and power generation for the three-stage brushless synchronous motor, eliminating the need for a dedicated engine starter, reducing system size and weight, and improving system integration. This is of great significance to aircraft power systems.

[0003] Three-stage brushless synchronous motors require accurate rotor position information when operating in electric mode to start aircraft engines. Traditional mechanical position sensors for acquiring rotor position suffer from low reliability and increased system size and weight. Online rotor position estimation for three-stage brushless synchronous motors eliminates the need for mechanical position sensors, enabling sensorless starting control. This improves system reliability and reduces system size and weight.

[0004] The three-stage brushless synchronous motor consists of a coaxially mounted permanent magnet auxiliary exciter, an exciter and a main motor. The structural diagram is shown in the figure. Figure 1 As shown in Figure 2, in some applications, the permanent magnet auxiliary exciter can be omitted, resulting in a two-stage brushless synchronous motor. Because the core components and operating principles are essentially 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 primarily utilize the motor's saliency or magnetic saturation characteristics, inject an auxiliary voltage signal, and extract the response signal containing rotor position information to estimate the rotor position. Methods utilizing motor saliency also require identifying the rotor's magnetic pole direction during zero speed. The steps include: 1. Injecting a rotating or high-frequency signal; 2. Using a bandpass filter to extract the response signal; 3. Signal demodulation; 4. Using a low-pass filter to obtain the signal envelope; 5. Using a phase-locked loop or inverse tangent method to estimate the rotor position; and 6. Initial rotor position estimation requires secondary identification of the magnetic pole polarity. Methods utilizing magnetic saturation characteristics primarily include: 1. Injecting pulse voltages into the main stator winding at fixed intervals within the range of 0-2π, collecting and storing the response current at the end of the pulse voltage injection; 2. Fitting and smoothing the stored data; and 3. Selecting the angular position corresponding to the maximum value as the rotor's initial position. This method processes a large amount of data, and the estimation accuracy is significantly affected by the arc interval between pulse injections and the accuracy of the current pulse response detection.

[0006] As can be seen from the above, the traditional method of identifying the initial position of the rotor has the following disadvantages: 1. The method used requires large amounts of data calculation and storage, and has high requirements for the processor; 2. The magnetic pole direction needs to be identified twice in the zero-speed static state; 3. The estimation accuracy is easily affected. Summary of the Invention

[0007] Technical problems to be solved

[0008] In order to avoid the shortcomings of the existing technology, the present invention proposes a multi-stage motor initial position estimation method based on low-frequency square wave signal injection, which overcomes the shortcomings of the existing rotor initial position estimation technology, such as reliance on saliency, large data processing volume, and the need for secondary identification of magnetic poles.

[0009] Technical Solution

[0010] A method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection is characterized in that the rotor position of the multi-stage brushless electrically excited synchronous motor refers to the rotor position of the main motor in the multi-stage brushless electrically excited synchronous motor, and the estimation steps are as follows:

[0011] Step 1: Apply two-phase symmetrical voltages to the stator windings of the two-phase exciter, and collect the stator currents of the two-phase exciter to estimate the excitation current of the main motor in real time. f_est ;

[0012] Step 2: After the main motor excitation current is stable, square wave voltages are applied in the four directions of the main motor stator winding equivalent to the α axis, -α axis, β axis and -β axis in sequence; within a low-frequency voltage cycle, at the end of the injection of the square wave voltages in the above four directions, the main motor excitation current is sequentially collected as I f_est_α , I f_est_-α , I f_est_β , I f_est_-β ;

[0013] Step 3: Calculate the initial position without magnetic pole polarity:

[0014] in:

[0015] Step 4: Compensate the initial position without magnetic polarity according to the following rules:

[0016] When θ0<0, and A<0, B>0, then θ1=θ0+2π,

[0017] When θ0<0, and A>0, B<0, then θ1=θ0+π,

[0018] When θ0>0, and A<0, B<0, then θ1=θ0,

[0019] When θ0>0, and A>0, B>0, then θ1=θ0+π;

[0020] Step 5: Final estimated rotor initial position θ 0est =θ1.

[0021] The multi-stage brushless electric excitation synchronous motor comprises a main motor and a two-phase exciter which are coaxially mounted.

[0022] In step 1, two symmetrical voltages are applied. Among them, u esα and u esβ is the stator winding voltage of the two-phase exciter, U es is the voltage amplitude, ω esf is the voltage fundamental angular velocity, and t is the time.

[0023] The step 1 estimates the excitation current I f_est for: in, and are the estimated three-phase currents of the two-phase exciter rotor, and:

[0024]

[0025] Among them, θ er ∈[0,2π) and can be selected as any fixed value, M emis the mutual inductance of the stator and rotor of the exciter, R es is the stator winding resistance, dt is the time differential, i esα and i esβ is the stator current of the two-phase exciter collected in real time in step 1, L es is the stator inductance of the exciter.

[0026] In step 2, the low-frequency square wave voltages applied to the stator winding of the main motor in the four directions of equivalent α axis, -α axis, β axis and -β axis are respectively:

[0027]

[0028] Where U is the voltage amplitude, t is the time, T is the low-frequency voltage period, k1 and k2 are positive integers greater than 8 and satisfy:

[0029] The ending time of the square wave voltage injection in the four directions in step 2 is:

[0030]

[0031] Among them, t0, t1, t2, and t3 are also the four saved time points for estimating the excitation current.

[0032] The multi-stage brushless electric excitation synchronous motor also includes a coaxially mounted main motor and a three-phase exciter. The three-phase voltage or the three-phase current is subjected to Clark transformation to obtain an equivalent two-phase exciter as described in 2. esα and u esβ ,i esα and i esβ .

[0033] Beneficial effects

[0034] The present invention proposes a multi-stage motor initial position estimation method based on low-frequency square wave signal injection. This method sequentially applies square wave voltage low-frequency voltage signals in four directions (equivalent to the α-axis, -α-axis, β-axis, and -β-axis) to the stator winding of the main motor, and collects the stator winding current of the two-phase exciter in real time. At the end of the square wave voltage injection in the four directions within a low-frequency voltage cycle, the estimated main motor excitation current is sequentially collected and saved, and the sine and cosine signals containing the main motor rotor position information are extracted. The obtained sine and cosine signals are then compensated based on the positive and negative value relationship of the sine and cosine signals through inverse tangent, and the initial position of the main motor rotor can be uniquely estimated. The method proposed in the present invention is not affected by the saliency of the main motor and the initial position of the exciter rotor, and has high estimation accuracy. Furthermore, the signal extraction and position estimation methods are simple, and the processor performance requirements are low.

[0035] The beneficial effects of the present invention include:

[0036] 1) Compared with the traditional rotor position initialization method based on saliency, this method does not rely on saliency and does not require the injection of high-frequency signals and the resulting complex filtering, demodulation, phase-locked loop processing, and secondary pulse injection for magnetic pole direction identification. Instead, the present invention only requires applying low-frequency square wave voltage signals in sequence to the four directions of the main motor stator winding equivalent to the α-axis, -α-axis, β-axis, and -β-axis. After simple addition and subtraction operations and inverse tangent processing on the collected signals at specific moments, the rotor initial position can be estimated at one time using the positive and negative values ​​of the collected signals. This reduces the amount of computation and places lower demands on the processor.

[0037] 2) Compared with the traditional rotor position initialization method based on magnetic pole saturation characteristics, it does not require fixed-angle injection signals and large amounts of data collection. It also avoids the steps of maximum value processing after fitting the data and the estimation accuracy is easily affected by the injection signal interval angle. Therefore, when only low-frequency square waves are injected in four directions, the processor memory requirements of this method are lower. Because the estimation principle uses arc tangent numerical calculation rather than approximate search based on magnetic saturation characteristics, the estimation accuracy is also higher.

[0038] 3) In the actual measurement, the rotor position is estimated every 15° electrical angle, and the estimation error results are as follows: Figure 9 As shown, the estimation errors are all less than 3°, and the error fluctuation is small. The estimation accuracy is much higher than the requirement of 10° error in this field.

[0039] 4) The method proposed in the present invention is also applicable to single-unit electromagnetic excitation motors. In this case, the same method can be used to collect and process the excitation current. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a schematic diagram of the structure of a three-stage brushless synchronous motor.

[0041] Figure 2 This is a flow chart of the method for estimating the initial position of the rotor of a multi-stage brushless synchronous motor proposed in the present invention.

[0042] Figure 3 The waveforms of the actual and estimated steady-state excitation currents are shown when there is no injection signal into the stator winding of the main motor of the present invention and the exciter has an ideal error.

[0043] Figure 4 The waveforms are the actual value and estimated value of the steady-state excitation current when there is no injection signal in the stator winding of the main motor of the present invention and the exciter has non-ideal errors.

[0044] Figure 5 This is the waveform of the low-frequency square wave voltage signal injected into the stator winding of the main motor of the present invention.

[0045] Figure 6 This is the stator current waveform of the exciter after injecting low-frequency square wave voltage in the present invention.

[0046] Figure 7 These are the estimation results of the main motor of the present invention under different actual rotor initial positions.

[0047] Figure 8 The actual value and estimated value waveforms of the excitation current of the main motor rotor in different sectors at the initial position of the present invention.

[0048] Figure 9 This is a diagram of the estimated error of the main motor rotor under different initial positions measured by the present invention. DETAILED DESCRIPTION

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

[0050] In the embodiment, the exciter of the multi-stage brushless synchronous motor is a two-phase exciter, that is, the stator excitation winding of the exciter is a two-phase winding with a 90 electrical angle difference. The exciter has 6 pairs of poles, and the main motor has 3 pairs of poles. In the following, the actual position of the main motor is θ 0real =30° and θ 0real =240° is used as an example for detailed explanation.

[0051] 1. Apply two-phase symmetrical voltages to the stator windings of the two-phase exciter and collect the stator current of the two-phase exciter to estimate the excitation current of the main motor in real time. The real-time estimated current is recorded as I f_est ;

[0052] The symmetrical voltage applied to the two phases is:

[0053]

[0054] Among them, u esα and u esβ is the stator winding voltage of the two-phase exciter, U es is the voltage amplitude, ω esf is the voltage fundamental angular velocity, t is the time;

[0055] And the estimated main motor excitation current is:

[0056]

[0057] in, and are the estimated three-phase currents of the two-phase exciter rotor, and:

[0058]

[0059] Among them, θ er ∈[0,2π) and can be selected as any fixed value, M em is the mutual inductance of the stator and rotor of the exciter, R esis the stator winding resistance, dt is the time differential, i esα and i esβ is the stator current of the two-phase exciter collected in real time in step 1, L es is the stator inductance of the exciter;

[0060] In the embodiment, two-phase symmetrical voltages are applied to the two-phase windings of the exciter stator (respectively denoted as α-phase winding and β-phase winding), the excitation voltage amplitude is 50V, the excitation frequency is 200Hz, and the two-phase currents of the stator of the two-phase exciter are collected and denoted as i esα and i esβ , then:

[0061]

[0062] Arbitrarily choose θ er =π / 12, the main motor excitation current can be estimated according to equations (2) and (3). er The error between the actual position of the exciter and When the excitation machine ideal error is abbreviated, the estimated steady-state excitation current of the main motor contains only DC quantity, which can be expressed as I f_dc ,like Figure 3 As shown. When θ er When the error between the actual position of the exciter and the exciter is other values, it is referred to as the exciter non-ideal error. The estimated steady-state excitation current of the main motor contains DC and periodic AC quantities, and the frequency of the periodic AC quantity is 6 times the frequency of the exciter stator voltage. The estimated excitation current can be expressed as I f_dc +I(6ω esf t), such as Figure 4 As shown;

[0063] 2. After the main motor excitation current is stable, square wave voltages are applied in sequence in the four directions of the main motor stator winding equivalent to the α axis, -α axis, β axis and -β axis. Within one low-frequency voltage cycle, at the end of the injection of the square wave voltages in the four directions, the estimated main motor excitation currents are sequentially collected and saved, and are recorded as I f_est_α , I f_est_-α , I f_est_β , I f_est_-β ;

[0064] The expressions for applying low-frequency square wave voltages in the four directions of the main motor stator winding equivalent to the α axis, -α axis, β axis and -β axis are:

[0065]

[0066] Where U is the voltage amplitude, t is the time, T is the low-frequency voltage period, k1 and k2 are positive integers greater than 8 and satisfy:

[0067]

[0068] In a low-frequency voltage cycle, the square wave voltage injection in the four directions ends at:

[0069]

[0070] Among them, t0, t1, t2, and t3 are also the four saved time points for estimating the excitation current.

[0071] In the embodiment, the low frequency square wave voltage amplitude U is injected L =3V, T=40ms, that is Take k1=k2=80, then the injected square wave low frequency signal is as follows Figure 5 As shown, it can be expressed as:

[0072]

[0073] The stator current of the exciter is as follows Figure 6 As shown, under the more complex non-ideal error conditions of the exciter, the estimated current expressions corresponding to t0, t1, t2 and t3 are:

[0074]

[0075] where R e is the equivalent resistance of the low-frequency square wave voltage loop, θ 0real is the actual value of the rotor initial position, τ is the equivalent time constant of the low-frequency square wave voltage loop;

[0076] 4. According to the estimated value of the main motor excitation current saved above, let:

[0077]

[0078] This results in the following initial position without magnetic pole polarity:

[0079]

[0080] In the embodiment, the expressions of A and B can be obtained from formula (10) as follows:

[0081]

[0082] The position information without magnetic polarity can be obtained from formula (11) and can be expressed as follows:

[0083]

[0084] 5. According to the compensation rule, perform phase compensation on θ0 to obtain θ1;

[0085] If θ0<0, and A<0, B>0, then θ1=θ0+2π,

[0086] If θ0<0, and A>0, B<0, then θ1=θ0+π,

[0087] If θ0>0, and A<0, B<0, then θ1=θ0,

[0088] If θ0>0, and A>0, B>0, then θ1=θ0+π;

[0089] 6. The final estimated initial position of the main motor rotor is:

[0090] θ 0est =θ1 (14)

[0091] The actual position of the main motor is θ 0real =30° and θ 0real =240° when the estimated result is as follows Figure 7 shown.

[0092] When the present invention estimates the excitation current of the main motor, if the exciter is in the ideal error state, I(6ω esf t) is always 0.

[0093] The waveforms of the actual and estimated excitation currents when the exciter has non-ideal errors at different sectors of the main motor rotor initial position are as follows: Figure 8 As shown, the estimation process is as follows Figure 2 shown.

Claims

1. A method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection, characterized in that: The rotor position of a multi-stage brushless electrically excited synchronous motor refers to the rotor position of the main motor in the multi-stage brushless electrically excited synchronous motor. The estimation steps are as follows: Step 1: Apply two-phase symmetrical voltages to the stator windings of the two-phase exciter, and collect the stator currents of the two-phase exciter to estimate the excitation current of the main motor in real time. f_est ; Step 2: After the main motor excitation current is stable, square wave voltages are applied in the four directions of the main motor stator winding equivalent to the α axis, -α axis, β axis and -β axis in sequence; within a low-frequency voltage cycle, at the end of the injection of the square wave voltages in the above four directions, the main motor excitation current is sequentially collected as I f_est_α , I f_est_-α , I f_est_β , I f_est_-β ; Step 3: Calculate the initial position without magnetic pole polarity: in: Step 4: Compensate the initial position without magnetic polarity according to the following rules: When θ0<0, and A<0, B>0, then θ1=θ0+2π, When θ0<0, and A>0, B<0, then θ1=θ0+π, When θ0>0, and A<0, B<0, then θ1=θ0, When θ0>0, and A>0, B>0, then θ1=θ0+π; Step 5: Final estimated rotor initial position θ 0est =θ1.

2. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: The multi-stage brushless electric excitation synchronous motor comprises a main motor and a two-phase exciter which are coaxially mounted.

3. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: In step 1, two symmetrical voltages are applied. Among them, u esα and u esβ is the stator winding voltage of the two-phase exciter, U es is the voltage amplitude, ω esf is the voltage fundamental angular velocity, and t is the time.

4. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: The step 1 estimates the excitation current I f_est for: in, and are the estimated three-phase currents of the two-phase exciter rotor, and: Among them, θ er ∈[0,2π) and can be selected as any fixed value, M em is the mutual inductance of the stator and rotor of the exciter, R es is the stator winding resistance, dt is the time differential, i esα and i esβ is the stator current of the two-phase exciter collected in real time in step 1, L es is the stator inductance of the exciter, u esα and u esβ is the stator winding voltage of the two-phase exciter.

5. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: In step 2, the low-frequency square wave voltages applied to the stator winding of the main motor in the four directions of equivalent α axis, -α axis, β axis and -β axis are respectively: Where U is the voltage amplitude, t is the time, T is the low-frequency voltage period, k1 and k2 are positive integers greater than 8 and satisfy: Where: esf is the voltage fundamental angular velocity.

6. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: The ending time of the square wave voltage injection in the four directions in step 2 is: Among them, t0, t1, t2, and t3 are also four storage time points for estimating the excitation current, T is the low-frequency voltage period, and k1 is a positive integer greater than 8.

7. The method for estimating the initial position of a multi-stage motor based on low-frequency square wave signal injection according to claim 1, characterized in that: The multi-stage brushless electric excitation synchronous motor also includes a coaxially mounted main motor and a three-phase exciter. The three-phase voltage or the three-phase current is subjected to Clark transformation to obtain the u of the equivalent two-phase exciter. esα and u esβ ,i esα and i esβ , where: u esα and u esβ is the stator winding voltage of the two-phase exciter, i esα and i esβ It is the stator current of the two-phase exciter collected in real time in step 1.

Images