Initial position estimation method for embedded permanent magnet synchronous motor
By comparing the rate of change of response current and using moving average filters, the problem that the high-frequency square wave voltage injection method cannot accurately obtain the initial position of the embedded permanent magnet synchronous motor is solved, and the accurate judgment of the polarity of the magnetic pole and the initial position recognition are achieved, which improves the Detection accuracy and stability.
Patent Information
- Application Number
- CN202210073110.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-21
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2042-01-21
AI Technical Summary
In the prior art, the high-frequency square wave voltage injection method cannot accurately obtain the initial position of the embedded permanent magnet synchronous motor, and the signal-to-noise ratio of the second harmonic detection method is too low, resulting in misjudgment of the polarity determination of the magnetic pole.
By comparing the rate of response current change, using high-frequency square wave voltage injection and moving average filter to obtain accurate magnetic pole polarity, inject additional DC current into the straight axis to determine the magnetic pole position, and using proportional integral observer and moving average filter Improve detection accuracy.
It realizes simple and fast magnetic pole polarity judgment and initial position identification, overcomes the accuracy defects of the high-frequency square wave voltage injection method, improves the stability and detection accuracy of the current rate of change, and is suitable for stationary and free-operating states.
Smart Images

Figure CN114531076B_ABST
Abstract
Description
Technical Field
[0001] The invention discloses an initial position estimation method of an embedded permanent magnet synchronous motor, relates to motor drive technology, and belongs to the technical field of power generation, power transformation or power distribution. Background Art
[0002] During the startup phase of an embedded permanent magnet synchronous motor, due to the small back electromotive force and low current signal-to-noise ratio, an additional high-frequency voltage signal needs to be injected to obtain the rotor position. Among the commonly used high-frequency injection methods, high-frequency square wave voltage injection has been widely studied for its simplicity and ease of implementation. However, during the startup phase of the motor, the rotor position identification method based on high-frequency square wave voltage injection can only obtain the position of the rotor magnetic pole, but cannot obtain the polarity of the magnetic pole corresponding to this position, that is, it cannot obtain the accurate initial position of the rotor.
[0003] The second harmonic detection method is generally used to determine the magnetic pole polarity. The second harmonic method uses the second harmonic component of the high-frequency current to determine the magnetic pole polarity. There is no need to increase the pulse voltage on the basis of the original high-frequency injection method. However, the signal-to-noise ratio of the second harmonic component containing the rotor magnetic pole polarity is too low, which can easily lead to misjudgment.
[0004] In order to overcome the defects of the second harmonic detection method in judging the magnetic pole polarity, a magnetic pole identification technology mainly uses the difference in d-axis current amplitude caused by the injection of high-frequency square wave voltages of different polarities to judge the magnetic pole of the rotor. However, the d-axis amplitude change caused by the high-frequency square wave voltage is limited. This type of magnetic pole identification method has high requirements on the accuracy of the current sensor; a magnetic pole position observation technology injects low-frequency square wave current and high-frequency square wave voltage into the d-axis at the same time, and estimates the rotor position based on the high-frequency response current. The accumulated values of the d-axis response current amplitudes of different polarities are compared by the observer to judge the rotor magnetic pole. Although this method can overcome the defects of the second harmonic detection method, it is complicated to operate and still has the defect that the high-frequency square wave voltage injection method cannot obtain the accurate initial position of the rotor.
[0005] The present invention aims to provide an initial position estimation method for an embedded permanent magnet synchronous motor, which can simply realize accurate judgment of magnetic pole polarity and overcome the inherent defect that a high-frequency square wave voltage injection method cannot accurately obtain the initial position of the rotor. Summary of the invention
[0006] The purpose of the present invention is to address the deficiencies of the above-mentioned background technology and provide an initial position estimation method for an embedded permanent magnet synchronous motor. By comparing the response current change rate, accurate judgment of the magnetic pole polarity can be achieved, and the initial position of the embedded permanent magnet synchronous motor system in a stationary or free-running state can be simply and quickly identified, thereby solving the technical problem that the method of identifying the magnetic pole polarity by high-frequency square wave voltage injection cannot accurately obtain the initial position of the rotor.
[0007] The present invention adopts the following technical solutions to achieve the above-mentioned invention object:
[0008] The initial position estimation method of the embedded permanent magnet synchronous motor includes steps A to E:
[0009] Step A: Inject a high-frequency square wave voltage signal into the estimated direct axis, and use a proportional integral observer to obtain the position θ of the rotor magnetic pole of the motor. p ;
[0010] Step B, while maintaining the high-frequency square wave voltage injection in step A and obtaining the rotor magnetic poles in real time, using a moving average filter to obtain and estimate the change rate M0 of the high-frequency response current on the direct axis;
[0011] Step C, while maintaining the high-frequency square wave voltage injection in step A and obtaining the rotor magnetic poles in real time, injecting additional direct current into the estimated direct axis, and using a moving average filter to obtain the high-frequency response current change rate M1 on the estimated direct axis;
[0012] Step D: while maintaining the high-frequency square wave voltage injection in step A and obtaining the rotor magnetic pole in real time, remove the additional current injection in step C, and then compare the magnitudes of M0 and M1 to determine whether the magnetic pole position estimated in step A corresponds to the south pole or the north pole;
[0013] Step E: While maintaining the high-frequency square wave voltage injection in step A and obtaining the rotor magnetic pole in real time, the accurate rotor position θ is obtained using the result of the polarity judgment in step D. e .
[0014] Furthermore, the initial position estimation method of the embedded permanent magnet synchronous motor uses a proportional integral observer to obtain the position θ of the rotor magnetic pole of the motor. p The specific method is: obtain the q-axis high-frequency response current by subtracting the q-axis current from its value delayed by one beat, and obtain the position of the rotor magnetic pole of the motor from the q-axis high-frequency response current.
[0015] Furthermore, the initial position estimation method of the embedded permanent magnet synchronous motor uses a moving average filter to obtain the estimated change rate M0 of the high-frequency response current on the direct axis. The specific method is: extract the median of the d-axis current change rate from the sampling sequence of the d-axis current change rate, and the sampling sequence of the d-axis current change rate is obtained by subtracting the d-axis current from its value delayed by one beat.
[0016] Furthermore, in the initial position estimation method of the embedded permanent magnet synchronous motor, the injected DC current can be of positive polarity or negative polarity, and the amplitude of the current is a fixed value.
[0017] Furthermore, in the initial position estimation method of the embedded permanent magnet synchronous motor, in step A, a high-frequency square wave voltage signal is continuously injected into the estimated direct axis for a period of time T1 until the rotor magnetic pole position estimated by the proportional integral observer converges.
[0018] Furthermore, in the method for estimating the initial position of the embedded permanent magnet synchronous motor, in step B, the high-frequency response current change rate is continuously obtained for a period of time T2 until T2 = N·T s , N is the filter window of the moving average filter, T s is the sampling period.
[0019] Further, in the method for estimating the initial position of the embedded permanent magnet synchronous motor, in step C, an additional DC current is continuously injected into the d-axis for a period of time T3 until T3 = N·T s .
[0020] Furthermore, in the method for estimating the initial position of the embedded permanent magnet synchronous motor, the specific basis for determining whether the magnetic pole position estimated in step A corresponds to the south pole or the north pole is:
[0021] When the polarity of the current injected in step C is positive: if M0<M1, the magnetic pole position corresponds to the north pole, otherwise, the magnetic pole position corresponds to the south pole;
[0022] When the polarity of the current injected in step C is negative: if M0>M1, the magnetic pole position corresponds to the north pole, otherwise, the magnetic pole position corresponds to the south pole.
[0023] Furthermore, the initial position estimation method of the embedded permanent magnet synchronous motor uses the result of the polarity judgment in step D to obtain the accurate rotor position θ e The specific method is:
[0024] When the magnetic pole position corresponds to the North Pole: The exact rotor position is the same as the rotor magnetic pole position obtained in step A, that is, θ e =θ p ;
[0025] When the magnetic pole position corresponds to the south pole: the exact rotor position is 180 degrees different from the rotor magnetic pole position obtained in step A, expressed in radians as θ e =θ p +π.
[0026] Furthermore, the initial position estimation method of the embedded permanent magnet synchronous motor, when the magnetic pole position corresponds to the south pole, obtains the accurate rotor position while reversing the polarity of the high-frequency square wave voltage injected into the d-axis.
[0027] The present invention adopts the above technical solution, which has the following beneficial effects: The present invention aims to provide a simple and fast initial position identification method for embedded permanent magnet synchronous motors. First, compared with the classic position-free control based on high-frequency square wave voltage injection, the initial position identification can be completed by only injecting a DC current into the d-axis for an additional period of time. This solution greatly simplifies the complex d-axis injection current generator in the previous initial position identification; second, the d-axis high-frequency response current change rate is utilized in the absence of additional d-axis current injection, which further simplifies the identification process compared to the method of injecting d-axis currents of two polarities; third, the d-axis high-frequency response current change rate has the characteristics of large values and large variation range compared to the d-axis high-frequency response current amplitude used in the past, which is more conducive to subsequent filtering and comparison; fourth, the moving average filter is used to obtain the d-axis high-frequency response current change rate. This method is relatively For the solution of directly using the d-axis high-frequency response current change rate or low-pass filter, a more accurate and stable current change rate can be obtained. The d-axis high-frequency response current change rate is an important indicator for judging the rotor polarity. Therefore, the use of a moving average filter can improve the accuracy of polarity identification. The use of a moving average filter can suppress the current change rate fluctuation caused by sampling noise and rotor position error, thereby improving the detection accuracy of the rotor magnetic pole; fifthly, the rotor position is constantly updated during the polarity identification process. This operating mode enables the present invention to be used not only for motors in a stationary state, but also for motors in a free-running state; sixthly, the d-axis current mutation caused by the estimated rotor position change is also compensated by changing the polarity of the injected square wave. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 It is a system block diagram of the initial position identification of the embedded permanent magnet synchronous motor provided by the present invention.
[0029] Figure 2 It is a timing diagram of initial position identification of the embedded permanent magnet synchronous motor provided by the present invention.
[0030] Figure 3 This is an experimental comparison diagram of the moving average filter used in the present invention and the low-pass filter used to process the d-axis high-frequency response current change rate without using a filter.
[0031] Figure 4 The method for identifying the initial position of an embedded permanent magnet synchronous motor provided by the present invention identifies a waveform diagram showing that the initial magnetic pole position is the south pole when the motor is in a stationary state.
[0032] Figure 5 The initial position identification method of the embedded permanent magnet synchronous motor provided by the present invention identifies the waveform diagram of the initial magnetic pole position corresponding to the south pole when the motor is in a free running state.
[0033] Figure 6 The waveform diagram is a waveform diagram of identifying the initial position of an embedded permanent magnet synchronous motor when the compensation method provided by the present invention is not used and when the compensation method provided by the present invention is used. DETAILED DESCRIPTION
[0034] The technical solution of the invention is described in detail below with reference to the accompanying drawings.
[0035] The present invention provides a system block diagram of the initial position identification of an embedded permanent magnet synchronous motor. Figure 1 As shown in the figure, the specific execution process of the initial position identification method is as follows Figure 2 As shown, the steps include:
[0036] Step A: After the controller is turned on, a high-frequency square wave voltage signal is injected into the d-axis, and the q-axis high-frequency response current i is obtained by subtracting the q-axis current from its value delayed by one beat. qh , where the period of one-beat delay is the sampling period T s After that, the proportional integral observer is used to obtain the position of the motor's rotor magnetic pole from the q-axis high-frequency response current. p , the expression of the proportional integral observer used in the continuous domain is:
[0037]
[0038] Among them, s is the Laplace operator, k p is the proportionality coefficient, k i is the integration coefficient,
[0039] Injecting a high-frequency square wave voltage signal into the d-axis needs to be executed for a period of time T1 until the rotor magnetic pole position estimated by the proportional-integral observer converges before executing the subsequent steps.
[0040] Step B: While injecting high-frequency square wave voltage in step A and measuring the rotor magnetic pole position by using proportional integral observer, obtain the d-axis high-frequency response current i by subtracting the d-axis current from its value delayed by one beat. dh Then, the high-frequency response current change rate M on the d-axis is calculated. The calculation formula of the high-frequency response current change rate M on the d-axis is:
[0041]
[0042] Among them, i dh [k-1], i dh [k] is the d-axis high-frequency response current at time k-1 and time k;
[0043] Afterwards, a moving average filter is used to extract the median M0 of the high-frequency response current change rate on the d-axis from the d-axis high-frequency response current sequence of length N and suppress its fluctuation. The moving average filter used in the present invention can be written as:
[0044]
[0045] Among them, M[i] and M[k] are the high-frequency response current change rates on the d-axis at time i and time k respectively, and N is the filter window of MAF;
[0046] The execution time for obtaining the median value M0 of the high-frequency response current change rate on the d-axis is T2. The setting of the execution time T2 mainly depends on the filtering effect of MAF, that is, the filtering window N. Therefore, the relationship between the execution time T2 and the filtering window N is: T2 = N·T s .
[0047] Step C: While injecting the high-frequency square wave voltage in step A and measuring the rotor pole position using a proportional integral observer, inject an additional DC current into the d-axis, and obtain the d-axis high-frequency response current i by subtracting the d-axis current from its value delayed by one beat. dh ;
[0048] Afterwards, the high-frequency response current change rate M1 on the d-axis is obtained by using MAF. Similar to step B, the execution time of step C to inject additional DC current into the d-axis is T3. T3 is also related to the filter window N of MAF. The relationship between the two is T3 = N·T s .
[0049] In addition, the DC current injected into the d-axis can be positive or negative, but the amplitude of the current is kept constant.
[0050] Step D, while maintaining the high-frequency square wave voltage injected in step A and using a proportional-integral observer to measure the rotor position, removing the additional current injection in step C;
[0051] Afterwards, the magnitudes of the high-frequency response current change rates M0 and M1 on the d-axis obtained in steps B and C are compared to determine the magnetic pole position θ estimated in step A. p The corresponding rotor position is the South Pole or the North Pole. The specific judgment basis is as follows:
[0052] The polarity of the DC current injected into the d-axis in step C is positive: if M0<M1, θ p The corresponding rotor position is the north pole, otherwise, it is the south pole;
[0053] The polarity of the DC current injected into the d-axis in step C is negative: if M0>M1, θ p The corresponding rotor position is the north pole, otherwise, it is the south pole.
[0054] Step E: While injecting high-frequency square wave voltage in step A and measuring the rotor magnetic pole position using a proportional integral observer, obtain the accurate rotor position θ according to the result of the rotor polarity judgment in step D. e , get the rotor position θ e The way is as follows:
[0055] If the rotor position is north pole: The exact rotor position is the same as the rotor magnetic pole position obtained in step A, i.e., θ e =θ p ;
[0056] If the rotor position is south pole: The exact rotor position is 180 degrees different from the rotor magnetic pole position obtained in step A, expressed as θ in radians e =θ p +π.
[0057] If the rotor position is the south pole, then the estimated position is changed from θ p Convert to θ e When the d-axis is moved, the polarity of the high-frequency square wave voltage injected into the d-axis needs to be reversed to reduce the d-axis current impact caused by the change in the estimated position.
[0058] like Figure 3 As shown, by comparing the d-axis high-frequency response current change rates obtained under the three conditions of no filter processing, low-pass filter processing and moving average filter processing, it can be seen that using the moving average filter can obtain a more stable d-axis high-frequency response current change rate.
[0059] like Figure 4 The figure shows the waveform of the initial position identification process of the motor in a stationary state. After a simple d-axis current injection, the rotor magnetic poles are obtained and the estimated initial position is compensated. The entire identification process takes only 11ms.
[0060] like Figure 5 As shown, it is a waveform diagram of the process of identifying the initial position of the motor in a free motion state. It can be seen that the algorithm disclosed in this patent can also quickly identify the initial position when the motor is running at a low speed.
[0061] like Figure 6 As shown, after the rotor magnetic pole is identified as the south pole, the change of the d-axis current after the rotor position is changed is estimated. In comparison, the compensation method proposed in the present invention can effectively suppress the current impact caused by the estimated rotor position change.
Claims
1. An initial position estimation method for an embedded permanent magnet synchronous motor, characterized in that: Inject a high-frequency square wave voltage signal into the direct axis, estimate the real-time magnetic pole position of the rotor, and calculate the median value of the high-frequency response current change rate on the direct axis; A positive polarity DC current or a negative polarity DC current with a constant amplitude is injected into the direct axis, and the median value of the high frequency response current change rate on the direct axis is calculated; The operation of injecting DC current into the direct axis is removed, and the median value of the change rate of the high-frequency response current on the direct axis before and after the DC current is injected is compared; The rotor position corresponding to the rotor real-time magnetic pole position is determined according to the comparison result. The specific method is as follows: When a positive polarity DC current is injected into the direct axis, if the median value of the high-frequency response current change rate on the direct axis before the DC current is injected is less than the median value of the high-frequency response current change rate on the direct axis after the DC current is injected, the rotor position corresponding to the real-time magnetic pole position of the rotor is the North Pole; otherwise, the rotor position corresponding to the real-time magnetic pole position of the rotor is the South Pole. When a negative polarity direct current is injected into the direct axis, when the median value of the change rate of the high-frequency response current on the direct axis before the direct current is injected is greater than the median value of the change rate of the high-frequency response current on the direct axis after the direct current is injected, the rotor position corresponding to the real-time magnetic pole position of the rotor is the north pole, otherwise, the rotor position corresponding to the real-time magnetic pole position of the rotor is the south pole; The final rotor position is obtained according to the rotor position corresponding to the real-time magnetic pole position of the rotor. The specific method is as follows: When the rotor position corresponding to the rotor real-time magnetic pole position is the North Pole, the rotor position corresponding to the rotor real-time magnetic pole position is the final rotor position. When the rotor position corresponding to the rotor real-time magnetic pole position is the south pole, 180 degrees is accumulated on the rotor position corresponding to the rotor real-time magnetic pole position to obtain the final rotor position.
2. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 1, characterized in that: The proportional integral observer is used to estimate the real-time magnetic pole position of the rotor. The expression of the proportional integral observer in the continuous domain is: s is the Laplace operator, θ p is the real-time magnetic pole position of the rotor, i qh is the high frequency response current on the quadrature axis, k p is the proportionality coefficient, k i is the integration coefficient.
3. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 1, characterized in that: The moving average filter is used to calculate the median value of the high frequency response current change rate on the direct axis. Among them, M[i] and M[k] are the change rates of the high-frequency response current on the direct axis at time i and time k respectively, and N is the filtering window of the moving average filter. i dh [k-1], i dh [k] is the high-frequency response current on the direct axis at time k-1 and time k, T s is the sampling period.
4. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 2, characterized in that: A high-frequency square wave voltage signal is injected into the direct axis for a period of time until the real-time rotor magnetic pole position estimated by the proportional-integral observer converges.
5. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 3, characterized in that: Before injecting DC current into the direct axis, obtain the time series of the high-frequency response current change rate on the direct axis in the T2 time period, and use the moving average filter to process the time series of the high-frequency response current change rate on the direct axis in the T2 time period to obtain the median value of the high-frequency response current change rate on the direct axis, T2 = N·T s , N is the filter window, T s is the sampling period.
6. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 1, characterized in that: The time for injecting DC current into the direct axis is T3, and the median value of the high-frequency response current change rate on the direct axis is calculated using a moving average filter, T3 = N·T s , N is the filter window of the moving average filter, T s is the sampling period.
7. The method for estimating the initial position of an embedded permanent magnet synchronous motor according to claim 1, characterized in that: When the rotor position corresponding to the real-time magnetic pole position of the rotor is the south pole, the polarity of the high-frequency square wave voltage signal injected into the direct axis is reversed while obtaining the final rotor position.
