A High Performance Direct Torque Control Method with PWM Approach of PMSMs

Abstract

The present invention proposes a DTC-PWM method that uses a torque and flux error level selector in a 12-sector voltage vector to select the voltage vector most suitable for the torque and flux conditions, and determines the duty cycle of the voltage vector using an approximate voltage function according to the angle within the sector of the selected voltage vector. In the proposed method, the approximate voltage function is approximated as a quadratic function for each dq-axis and used as the denominator to calculate the duty ratio. The numerator of the duty ratio calculation is derived from torque, magnetic flux error, and speed information, and, unlike conventional DTC-PWM, is used to obtain an appropriate duty ratio by referring to the motor speed. The validity of the proposed control method was verified through actual experiments using a commercial PMSM, and the experiments confirmed that the proposed control method has superior control performance compared to the conventional method.

Description

Technology Field

[0001] The present invention relates to a DTC control method for a PMSM using a PWM method, and more specifically, to a high-performance direct torque control method using high-performance approximate voltage function-based PWM (pulse width modulation). Background Technology

[0002] As is well known, PMSMs (Permanent Magnet Synchronous Motors) are the most widely used electric motors in industrial robotics and high-performance control systems, and high-performance torque control methods are essential for the improved control performance of PMSMs.

[0003] Among these torque control methods, Field Oriented Control (FOC) and Direct Torque Control (DTC), known as vector control, are commonly and widely used.

[0004] As is known, vector control methods can separate torque and flux components by dq-axis transformation, and torque and flux can be directly obtained from the motor's torque constant and the transformed q-axis and d-axis currents.

[0005] However, in the case of torque and flux current control methods, dq-axis voltage commands must be used to generate appropriate dq-axis current, and current control performance may depend on the gain of the controller and the precision of the motor parameters.

[0006] In addition, in the case of some controllers such as PI and PID controllers (Proportional-Integral-Derivative controllers), saturation may occur due to the integral term.

[0007] To address the saturation of these integral terms, a well-designed anti-windup controller may be required separately.

[0008] In particular, the actual implementation of the voltage output can be performed by PWM techniques such as the SVPWM (Space Vector Pulse Width Modulation) method.

[0009] However, as is known, SVPWM has the disadvantage of being computationally complex due to dead-time compensation.

[0010] On the other hand, DTC (Direct Torque Control) has the advantage of being implementable with simple switching rules, but it has the problem that the same sampling switching frequency is not constant and the torque and magnetic flux ripple are relatively high.

[0011] Various DTC methods are being studied to solve these problems. DTC-SVM (Direct Torque Control-Space Vector Modulation) is a method designed to reduce torque ripple by combining DTC and SVPWM techniques. However, while DTC-SVM offers excellent control performance, it has the drawback of a very complex control structure due to the mutual coupling of DTC and SVPWM. Furthermore, because DTC-SVM generates voltage commands using a conventional PI controller, it has the disadvantage that, in addition to the saturation problem, its control performance is dependent on the controller's gain and motor parameters.

[0012] DTC-PWM is another approach to improve CDTC performance, and in DTC-PWM, the method of calculating the duty cycle is very important.

[0013] To reduce torque ripple, various duty cycle measurement methods have been proposed in existing papers; however, the majority of these papers have used a method that utilizes torque and magnetic flux errors based on the sampling period to optimize torque ripple.

[0015] Meanwhile, the basic principles of conventionally used DTC can be briefly explained as follows.

[0016] [Basic Principles of Conventional DTC]

[0017] A. Analysis of DTC using the 6-Sector Method

[0018] Figure 1 shows the control block diagram and voltage vector selection of a conventional DTC. As shown in Figure 1, the control block of the CDTC is very simple. Only torque and flux estimators are used to obtain the torque and flux of the PMSM. The torque and flux errors are used to determine the voltage vector between the six RMS voltage vectors of the VSI (voltage-type inverter), as shown in Figure 1(b). As shown in Figure 1(b), the voltage space can be divided into four regions according to the torque and flux characteristics. (1,1), (1,-1), (-1,1), and (-1,-1) represent the torque error and flux error. , It is an area divided according to each.

[0019] class is determined according to CDTC as shown in Fig. 1(a).

[0020] (1)

[0021] (2)

[0022] In each region, only one voltage vector is selected to generate the specified magnetic flux and torque shown in Table 1 below.

[0023] Rotor position (sector number) S1 S2 S3 S4 S5 S6 1 1 V2 V3 V4 V5 V6 V1 -1 V6 V1 V2 V3 V4 V5 -1 1 V3 V4 V5 V6 V1 V2 -1 V5 V6 V1 V2 V3 V4

[0024] In Table 1, the voltage vector of the inverter selected according to torque, magnetic flux, and sector is the dq-axis voltage component class It is analyzed as follows. Figure 2 shows the dq-axis analysis voltage of the voltage vector preceding Table 1. As shown in Figure 2, the q-axis and d-axis voltages are not constant in magnitude within the same sector area and change in a sinusoidal manner. The d-axis voltage fluctuation is nearly 70% of the DC-link voltage, and the q-axis voltage fluctuation is approximately 40% of the DC-link voltage. As shown in Figure 2, the d-axis voltage around the sector edge is nearly zero. And this zero voltage is insufficient to generate an adequate magnetic flux current under certain conditions. Due to voltage variations within the same sector, torque and magnetic flux cannot be maintained at constant values ​​by the selected voltage vector in CDTC.

[0025] Meanwhile, regarding conventional torque control for PMSMs, the Direct Torque Control (DTC) method receives significant attention due to its ease of implementation and fast dynamic response. However, it has the disadvantage of being difficult to apply to precision systems due to high torque ripple and random switching frequencies. For this reason, DTC-PWM (Direct Torque Control Pulse Width Modulation) is used to reduce torque ripple and fix the switching frequency; however, even with the DTC-PWM method, the output voltage varies depending on the rotor position of the selected voltage vector, and this issue leads to an increase in torque ripple.

[0026] In order to solve this problem, the present invention proposes a DTC-PWM control technique having an approximate voltage function, a DTC with 12 sectors, a torque and flux error level selector, and a duty cycle that considers back EMF. Prior art literature

[0027] [1] M. Depenbrock, "Direct self-control of inverter-fed machine," IEEETrans. Power Electron., vol. 3, pp. 420-429, Oct. 1988.[2] I. Takahashi and T. Naguchi, "A new quick-response and high efficiency control strategy of an induction motor," IEEE Trans.Ind. Applicat., vol.IA-22, pp. 820-827, Sept. / Oct. 1986.[3] C. French and P. Acarnley, "Direct torque control of permanent magnet drives," IEEE Trans. Ind. Applicat., vol. IA-32, pp. 1080 - 1088, Sept. / Oct. 1996.[4] L. Zhong, M. F. Rahman, W. Y. Hu, and K. W. Lim, "Analysis of direct torque control in permanent magnet synchronous motor drives," IEEE Trans. Power Electron., vol. 12, pp. 528-536, May 1997.[5] M. F. Rahman, L. Zhong, and K. W. Lim, "A direct torque controlledinterior permanent magnet synchronous motor drive incorporating field weakening," IEEE Trans. Ind. Applicat., vol.34, pp. 1246-1253, Nov. / Dec. 1998.[6] C. G. Mei, S. K. Panda, J. X. Xu, and K. W.Lim, "Direct torque control of induction motor-variable switching sectors," in Proc. IEEE-PEDS Conf., Hong Kong, July 1999, pp. 80-85.[7] A. Tripathi, A. M. Khambadkone, and S. K. Panda, "Space-vector based, constant frequency, direct torque control and dead beat stator flux control of AC machines," in Proc. IEEE-IECON'01 Conf., Nov. 2001, pp. 1219-1224.[8] Yongchang Zhang, Jianguo Zhu, "Direct Torque Control of Permanent Magnet Synchronous Motor With Reduced Torque Ripple and Commutation Frequency", Power Electronics, IEEE Transactions , Jan. 2011 ,vol.26 ,pp. 235-248[9] Ke Wang, Yaohua Li, Liming Shi, Qiongxuan Ge, "A Novel Switching Scheme for Direct Thrust Control of LIM with Reduction of Thrust Ripple", Electrical Machines and Systems (ICEMS), 2010 International Conference, Oct. 2010 , 1491-1494.

[10] Chintan Patel, Rijil Ramchand, Rajeevan.P.P,K Sivakumar, Anandarup Das, K.Gopakumar, Marian P.Kazmierkowski, "Direct Torque Control Scheme of IM Drive with 12-sided Polygonal Voltage Space Vectors", Power Electronics and Applications (EPE 2011), Proceedings of the 2011-14th European Conference, Sept 2011, 1-11. The problem to be solved

[0028] The present invention proposes a DTC-PWM method that uses a torque and flux error level selector in a 12-sector voltage vector to select the voltage vector most suitable for the torque and flux conditions, and determines the duty cycle of the voltage vector using an approximate voltage function according to the angle within the sector of the selected voltage vector.

[0029] In the present invention, a new DTC-PWM control method for a PMSM based on an approximate voltage function is proposed to reduce torque angle deviation based on the analysis of the supply voltage in the two-part division.

[0030] The voltage vector selection is determined by the magnetic flux and torque error level selector, and unlike conventional DTC-PWM, the voltage vector can be selected from three different voltage vectors according to the magnetic flux and torque error levels.

[0031] Among the three different voltage vectors, the most suitable voltage vector is determined by an error level selector designed to reduce magnetic flux ripple in torque and fast dynamic response.

[0032] The present invention proposes an approximate voltage function based on the dq voltage analysis of a selected voltage vector for calculating the PWM duty ratio.

[0033] In addition, the approximate voltage function expressed in terms of voltage angle is used as the denominator in the duty ratio calculation. means of solving the problem

[0034] The DTC control method of a PMSM using a PWM method proposed in the present invention is,

[0035] In order to control the PWM inverter linked to the above PMSM,

[0036] Calculate the dq-axis voltage angle using the rotor position information of the above PMSM, calculate the dq-axis approximate voltage value using the level of the torque error and the level of the magnetic flux error of the above PMSM, calculate a predetermined duty cycle using the dq-axis approximate voltage value, the torque error, and the magnetic flux error, and then control the duty cycle of the voltage vector applied to the PWM inverter,

[0037] The above dq axis voltage angle within a predetermined sector is

[0038]

[0039] The above dq-axis approximate voltage value

[0040]

[0041]

[0042] The above predetermined duty is is,

[0043] Here,

[0044]

[0045] And,

[0046] The above torque error refers to the difference between the torque command value and the torque estimate, and

[0047] The above magnetic flux error refers to the difference between the magnetic flux command value and the magnetic flux estimate, and

[0048] The level of the above torque error represents the quantitative level of the above torque error, and

[0049] The level of the above magnetic flux error is characterized by representing a quantitative level of the above magnetic flux error.

[0050] In the present invention, the level of the torque error is divided into multiple levels according to the magnitude of a predetermined set torque error, and the level of the magnetic flux error is divided into multiple levels according to the magnitude of a predetermined set magnetic flux error.

[0051] In the present invention, the voltage vector according to the position of the rotor applied to the PWM inverter is determined according to the level of the torque error and the level of the magnetic flux error.

[0052] In the present invention, the voltage vector according to the position of the rotor applied to the PWM inverter is characterized as being a voltage vector divided into 12 sectors. Effects of the invention

[0053] When using the control method of the present invention, it was confirmed that controlling the motor using the DTC-PWM method of a PMSM based on an approximate voltage function significantly reduces torque ripple compared to the conventional DTC method.

[0054] That is, in the method proposed in the present invention, the approximate voltage function is approximated as a quadratic function for each dq-axis and used as the denominator for calculating the duty ratio. The numerator for calculating the duty ratio is derived from torque, magnetic flux error, and speed information, and, unlike conventional DTC-PWM, is used to obtain an appropriate duty ratio by referring to the motor speed. The effectiveness of the proposed control method was verified through actual experiments using a commercial PMSM, and in the experiments, it was confirmed that the proposed control method had superior control performance compared to the conventional method. Brief explanation of the drawing

[0055] Fig. 1. A diagram illustrating the block diagram and voltage vector of a conventional DTC. Figure 2 is a diagram showing the analyzed dq axis voltage according to the voltage angle and sector. Fig. 3 is a diagram showing 12 sector areas and 12 voltage vectors. Fig. 4 is a diagram illustrating the proposed torque error level sector. Fig. 5. This is a diagram illustrating the voltage analyzed on the dq side of the selected voltage vector according to the magnitude of the torque error and magnetic flux error in the 12-sector method. Fig. 6 is a diagram illustrating the severance voltage and approximate voltage on the dq axis. = PB and = PS ) Fig. 7. This is a block diagram of the proposed DTC-PWM method. Fig. 8 is a graph showing the results of a comparative experiment. Specific details for implementing the invention

[0056] Hereinafter, a DTC control method for a PMSM using a PWM method proposed in the present invention will be explained with reference to drawings and tables.

[0058] B. Voltage Vector Analysis of 12-Sector DTC

[0059] To reduce torque and magnetic flux ripple, it can be used as a 12-sector area. FIG. 3 shows a 12-sector area and possible 12 voltage vectors. Among the 12 voltage vectors, six voltage vectors, , , , , and is the effective voltage vector of VSI (voltage-type inverter). And, the added voltage vector , , , , and is a combinational voltage vector. As shown in Fig. 3, a combinational voltage vector can be supplied by the transition of two adjacent voltage vectors. Thus, is a voltage vector class It can be created by a combination of. The power switch of phase B is turned on for half the time of the switching cycle to combine the two voltage vectors. As shown in Fig. 1, the notations (1,1), (1, -1), (-1, 1), and (-1, -1) represent torque and magnetic flux error, respectively. and It is a region divided according to. As shown in Fig. 3, three voltage vectors can be selected in the same torque flux region. For example, voltage vector , and is the rotor position It can generate positive torque and positive magnetic flux. And, voltage vector , and It can generate positive torque and negative magnetic flux. However, the three voltage vectors have different characteristics. Rotor position In, voltage vector It generates positive low torque and positive high magnetic flux. However, the voltage vector It generates positive low magnetic flux while having positive high torque. Voltage vector It can generate positive intermediate magnetic flux at positive intermediate torque.

[0061] In the present invention, unlike conventional DTC, class It has torque and magnetic flux error levels as shown in Fig. 4, PB (+3), PM (+2), PS (+1), NS (-1), NM (-2), and NB (-3). The error levels are classified as follows. Based on the sign and magnitude of the error, they are classified into PB (large positive), PM (medium positive), PS (small positive), NS (small negative), NM (medium negative), and NB (large negative). Table II shows the voltage vectors according to the torque and error levels and rotor position.

[0063] (a) Voltage vectors selected in the proposed DTC (S1 ~ S6)

[0064] Rotor position (sector number) S1 S2 S3 S4 S5 S6 PB PS V12 V2 V23 V3 V34 V4 NS V1 V12 V2 V23 V3 V34 PM PM V2 V23 V3 V34 V4 V45 NM V61 V1 V12 V2 V23 V3 PS PB V23 V3 V34 V4 V45 V5 NB V6 V61 V1 V12 V2 V23 NS PB V3 V34 V4 V45 V5 V56 NB V56 V6 V61 V1 V12 V2 NM PM V34 V4 V45 V5 V56 V6 NM V5 V56 V6 V61 V1 V12 NB PS V4 V45 V5 V56 V6 V61 NS V45 V5 V56 V6 V61 V1

[0065] (b) Voltage vectors selected from the proposed DTC (S7 ~ S12)

[0066] Rotor position (sector number) S7 S8 S9 S10 S11 S12 PB PS V45 V5 V56 V6 V61 V1 NS V4 V45 V5 V56 V6 V61 PM PM V5 V56 V6 V61 V1 V12 NM V34 V4 V45 V5 V56 V6 PS PB V56 V6 V61 V1 V12 V2 NB V3 V34 V4 V45 V5 V56 NS PB V6 V61 V1 V12 V2 V23 NB V23 V3 V34 V4 V45 V5 NM PM V61 V1 V12 V2 V23 V3 NM V2 V23 V3 V34 V4 V56 NB PS V1 V12 V2 V23 V3 V34 NS V12 V2 V23 V3 V34 V4

[0067] III. Proposed DTC-PWM Scheme A. Analysis of Voltage Vector and Approximate Voltage Function

[0068] As shown in FIG. 5, the present invention analyzes that the output voltage within a voltage vector is not constant. To solve this problem, an approximate voltage function is proposed. The approximate voltage function was derived as a quadratic equation using the least squares method for the ripple of the output voltage as shown in FIG. 5.

[0069] (3)

[0070] Here, is the position of the rotor It is the sector number.

[0071] is the voltage angle defined within a sector from 0 to Set to between radians.

[0072] The approximate voltage function is as follows: It can be expressed as a quadratic expression with variables.

[0073] (4)

[0074] (5)

[0075] Here, , , and , , is the coefficient of the approximate voltage function according to the torque and flux error levels and the position of the rotor, and these coefficients are as shown in Tables 4 and 5.

[0076] FIG. 6 is an approximate voltage function derived in the present invention ( , ) and the voltage output from a voltage vector analyzed in Fig. 5 ( , You can see that it matches the curve of ).

[0078] [Coefficients of the DQ axis approximation voltage function]

[0079] (a) Coefficients of the D-axis approximate voltage function

[0080] S1 / S3 / S5 / S7 / S9 / S11 S2 / S4 / S6 / S8 / S10 / S12 PB PS -0.3390 0.3771 0.6058 -0.3915 0.4285 0.6995 NS -0.3933 0.0098 0.8166 -0.3406 0.0085 0.7072 PM PM -0.2848 0.7325 0.3950 -0.2466 0.6343 0.3421 NM -0.2509 -0.3565 0.6191 -0.2898 -0.4116 0.7149 PS PB -0.0881 0.7276 -0.0134 -0.1017 0.8401 -0.0154 NB -0.1086 -0.7227 0.4216 -0.0940 -0.6259 0.3652 NS PB -0.1086 -0.7227 0.4216 -0.0940 -0.6259 0.3652 NB -0.0881 0.7276 -0.0134 -0.1017 0.8401 -0.0154 NM PM -0.2509 -0.3565 0.6191 -0.2898 -0.4116 0.7149 NM -0.2848 0.7325 0.3950 -0.2466 0.6343 0.3421 NB PS -0.3933 0.0098 0.8166 -0.3406 0.0085 0.7072 NS -0.3390 0.3771 0.6058 -0.3915 0.4285 0.6995

[0081] (b) Coefficients of the Q-axis approximate voltage function

[0082] S1 / S3 / S5 / S7 / S9 / S11 S2 / S4 / S6 / S8 / S10 / S12 PB PS -0.3390 0.3771 0.6058 -0.3915 0.4285 0.6995 NS -0.3933 0.0098 0.8166 -0.3406 0.0085 0.7072 PM PM -0.2848 0.7325 0.3950 -0.2466 0.6343 0.3421 NM -0.2509 -0.3565 0.6191 -0.2898 -0.4116 0.7149 PS PB -0.0881 0.7276 -0.0134 -0.1017 0.8401 -0.0154 NB -0.1086 -0.7227 0.4216 -0.0940 -0.6259 0.3652 NS PB -0.1086 -0.7227 0.4216 -0.0940 -0.6259 0.3652 NB -0.0881 0.7276 -0.0134 -0.1017 0.8401 -0.0154 NM PM -0.2509 -0.3565 0.6191 -0.2898 -0.4116 0.7149 NM -0.2848 0.7325 0.3950 -0.2466 0.6343 0.3421 NB PS -0.3933 0.0098 0.8166 -0.3406 0.0085 0.7072 NS -0.3390 0.3771 0.6058 -0.3915 0.4285 0.6995

[0083] B. Proposed Duty Ratio Calculation Calculating the duty ratio of a selected voltage vector is crucial for reducing torque ripple. In previous inventions, the duty ratio was calculated using the error between torque and magnetic flux and a proportionality constant. However, this calculation method fails to account for the back EMF generated during motor operation, and thus cannot sufficiently reduce torque ripple.

[0084] Therefore, in the present invention, a term considering the back EMF of the motor was included in the duty cycle calculation.

[0085] The proposed duty cycle is as follows.

[0086] (8)

[0087] (9)

[0088] (10)

[0089] Here, class and are the duty ratios for the d and q axes, respectively. d is the proposed class The sum is the final DTC-PWM duty ratio. class is the proportional gain value for the magnetic flux and torque of each axis. class is the proportional gain value for the speed of each axis. By using the duty cycle calculated in this way and controlling as shown in Fig. 7, the effect of reducing torque ripple can be observed.

[0090] IV. Experimental Results

[0091] In order to verify the proposed DTC-PWM method in this invention, a prototype test was performed. The PMSM used in the experiment has a rated power of 750W and a rated speed of 3000 [rpm]. The digital controller used was the TMS320F28335 DSP from TI (Texas Instruments).

[0092] Figure 8 shows the results of a comparative experiment on each control technique. Figure 8(a) shows the conventional DTC method. In this experiment, the duty cycle was fixed at 95%, and it can be seen that there is high current ripple in each axis. Figure 8(b) shows the 6-sector DTC-PWM method. Compared to Figure 8(a), it can be seen that the current ripple in each axis has decreased. Figure 8(c) shows the proposed method. Figure 8(c) confirms that the current ripple is significantly reduced compared to other control techniques.

Claims

Claim 1 As a method for controlling DTC of a PMSM using a PWM method, in order to control a PWM inverter linked to the PMSM, the above PMSM's Calculate the dq-axis voltage angle using rotor position information, calculate the dq-axis approximate voltage value using the level of torque error and the level of magnetic flux error of the PMSM, calculate a predetermined duty cycle using the dq-axis approximate voltage value, the torque error, and the magnetic flux error, and then control the duty cycle of the voltage vector applied to the PWM inverter, wherein the dq-axis voltage angle within a predetermined sector is The above dq-axis approximate voltage value The above predetermined duty is and, here, And, the above torque error refers to the difference between the torque command value and the torque estimate value, the above flux error refers to the difference between the flux command value and the flux estimate value, the level of the above torque error indicates the quantitative level of the above torque error, and the level of the above flux error indicates the quantitative level of the above flux error, The above Adb, Aqb, the above Bdb, Bqb, and the above Cdb, Cqb are coefficients of the approximate voltage function according to the torque and magnetic flux error levels and the rotor position, respectively, and class is the proportional gain value for the magnetic flux and torque of each axis, and class A DTC control method for a PMSM using a PWM method characterized by being a proportional gain value for the speed of each axis. Claim 2 A DTC control method for a PMSM using a PWM method, characterized in that, in claim 1, the level of the torque error is divided into multiple levels according to the magnitude of a predetermined set torque error, and the level of the magnetic flux error is divided into multiple levels according to the magnitude of a predetermined set magnetic flux error. Claim 3 A DTC control method for a PMSM using a PWM method, characterized in that, in claim 2, a voltage vector according to the position of the rotor applied to the PWM inverter is determined according to the level of the torque error and the level of the magnetic flux error.

Images