Abstract: This paper proposes a novel sensorless vector control method for ultra-low-speed asynchronous motors. This method is based on low-frequency signal injection, injecting a low-frequency stator current signal and using the resulting angle error to estimate the motor speed. This method is unaffected by load changes and does not rely on the non-ideal characteristics of the asynchronous motor. Speed estimation in the ultra-low-speed range can be achieved using only the fundamental wave model, thus it is unaffected by the asynchronous motor structure and has universal applicability. Furthermore, this method exhibits strong robustness to motor parameters, eliminates the need for parameter estimation, and has a simple control structure. Simulation and experimental results demonstrate that the proposed low-frequency signal injection method can effectively achieve sensorless vector control of asynchronous motors in the ultra-low-speed range.
Abstract: This paper presents a new sensorless vector control method for induction motor (IM) drive at very low speeds. The proposed method is based on low-frequency signal injection, where a low frequency stator current signal is injected and the corresponding angle error is detected to estimate the rotor speed. This new method is independent of the non-ideal features of IM. Just the fundamental model of IM is needed. As a result, the proposed method isn't affected by IM structures and it can be applied to different IM. Furthermore, this new method is robust to the motor parameters. Therefore no parameter estimation is needed and the control system is simple. Simulations and experimental results prove the good performances of the proposed method on speed sensorless control of IM at very low speeds.
Keywords: Low-frequency signal injection asynchronous motor sensorless vector control
Keywords: Low-Frequency Signal Injection, IM, Speed Sensorless, Vector Control
1 Introduction
In recent years, sensorless vector control of asynchronous motors has become a research hotspot. Currently, sensorless vector control of asynchronous motors has achieved good control performance in the medium and high speed range, but good control has not yet been achieved in the extremely low speed range (<1Hz). This is because commonly used sensorless vector control methods for asynchronous motors require the use of back electromotive force (EMF), which is very small at extremely low speeds and is greatly affected by sampling accuracy and changes in motor parameters, leading to reduced control performance and making sensorless vector control in the extremely low speed range impossible.
To achieve sensorless control of asynchronous motors in the ultra-low speed range, researchers have proposed various control methods. Among them, the high-frequency signal injection method has been studied extensively, which uses the current response generated by the injected high-frequency stator voltage signal to estimate the rotor position [1]-[5]. These high-frequency signal injection methods all utilize the non-ideal characteristics of asynchronous motors, such as rotor salient poles, cogging effect, and saturation effect. However, these high-frequency signal injection methods have a common drawback, namely, the high-frequency response signal is often mixed with other high-frequency harmonics and is difficult to separate. Complex signal processing methods are required to obtain the required high-frequency response signal, which reduces the system response speed and increases the complexity of the control system. In addition, since the high-frequency signal injection method utilizes the non-ideal characteristics of asynchronous motors, it is greatly affected by the motor structure and lacks certain universality.
To avoid the various problems inherent in the high-frequency signal injection method, this paper proposes a method based on low-frequency signal injection. This method transforms the high-frequency harmonic signal in reference [5] into a low-frequency harmonic signal, injects a low-frequency d-axis stator current signal, and uses the resulting angle error to estimate the motor speed. This method only uses the fundamental wave model of the asynchronous motor and does not depend on various non-ideal characteristics, so it is not affected by the structure of the asynchronous motor and has universal applicability. In addition, the low-frequency response signal required by this method is easy to separate, eliminating the disadvantage of difficult signal separation in the high-frequency signal injection method. Moreover, it has strong robustness to motor parameters and does not require parameter estimation, making the control system structure relatively simple. Simulation and experimental results prove that the method based on low-frequency signal injection proposed in this paper can well realize sensorless vector control of asynchronous motors in the ultra-low speed range.
2. Principle of Low-Frequency Signal Injection Method
From the mathematical model and equations of motion of the asynchronous motor, the electromagnetic torque of the asynchronous motor can be expressed as:
Figure 2. Block diagram of the control principle of the low-frequency signal injection method system.
3. Robustness Analysis of Motor Parameters
As can be seen from the above analysis, the low-frequency signal injection method proposed in this paper is only related to the injected signal and the torque response it causes, and is not related to the stator resistance and rotor resistance of the asynchronous motor. Therefore, it has good robustness to stator and rotor resistance.
Furthermore, since the rotor time constant is used in the slip speed calculation, its variation will affect the accuracy of the estimated speed. However, since the rotor time constant is not used in the synchronous speed estimation, its variation has no effect on the estimation of the rotor flux linkage angle. Therefore, the method presented in this paper also exhibits good robustness to the rotor time constant.
Due to its robustness to asynchronous motor parameters, the proposed method effectively eliminates the impact of parameter variations on extremely low-speed performance. Furthermore, since parameter estimation is unnecessary, both the control algorithm and system architecture are significantly simplified.
4 Simulation Results
This paper presents a simulation of the proposed low-frequency signal injection method, and the motor parameters used are shown in Table 1.
Based on the motor parameters, the frequency of the injected low-frequency d-axis stator current signal is set to 25Hz, and the amplitude is 0.6 times the rated d-axis stator current.
Table 1 Asynchronous Motor Parameters
Figure 3. No load, reference speed abruptly changes from 0.3Hz to -0.3Hz.
Figure 4. 60% rated load, reference speed sudden change from 0.1Hz to -0.1Hz.
Figure 3 shows the simulated waveform of an asynchronous motor suddenly changing from forward to reverse when running under no-load at 0.3Hz.
Figure 4 shows the simulated waveform of an asynchronous motor suddenly switching from forward to reverse rotation when running at 0.1Hz with 60% rated load. As can be seen from Figures 3 and 4, regardless of whether there is a load, the system can quickly recover to stability after the sudden change in speed, and the steady-state error is small.
Figure 5. Reference speed 0.1Hz, load abruptly 0->60%Tn
Figure 5 shows the simulated waveform of the asynchronous motor running at 0.1 Hz, transitioning from no-load to 60% rated load. The figure shows that the system quickly recovers to stability after the load change, with a small steady-state error.
This paper also combines the proposed low-frequency signal injection method with the voltage model method. The low-frequency signal injection method is used when the speed is below 1Hz, and the voltage model method is used otherwise. Figure 6 shows the simulation waveforms of the asynchronous motor acceleration and deceleration when switching between the two methods with 60% rated load. As can be seen from the figure, this method has good dynamic and steady-state performance and can realize accurate speed regulation of the asynchronous motor over a wide range.
Figure 6. 60% rated load, reference speed 0.1Hz->10Hz->0.1Hz
5. Experimental Results
This paper experimentally verifies the proposed low-frequency signal injection method. The experiment uses a DSP TMS320C31-based experimental platform, and the motor parameters and injected signals are consistent with those in the simulation.
Figure 7. No load, reference speed 0.3Hz
Figure 7 shows the steady-state waveform of the asynchronous motor under no-load operation at 0.3 Hz. As can be seen from the figure, the steady-state error between the estimated speed and the actual speed is small, with only small pulsations.
Figure 8. No load, reference speed abruptly changes from -0.3Hz to 0.3Hz.
Figure 8 shows the experimental waveform of an asynchronous motor suddenly switching from reverse to forward rotation during very low-speed no-load operation. As can be seen from the figure, after the reference speed abruptly changes, the actual speed can recover to stability relatively quickly, and the steady-state error is small.
6. Conclusion
This paper proposes a sensorless vector control method for ultra-low-speed asynchronous motors based on low-frequency signal injection. Through theoretical analysis, simulation, and experimental verification, the proposed method achieves excellent control performance for asynchronous motors in the ultra-low-speed range. This method utilizes only the fundamental wave model of the asynchronous motor and does not rely on various non-ideal characteristics, thus it is unaffected by the asynchronous motor structure and has universal applicability. Furthermore, the required low-frequency response signal is easily separated and exhibits strong robustness to motor parameters, eliminating the need for complex signal separation and parameter estimation, resulting in a relatively simple control algorithm and system structure.
References
1 PL Jansen and RD Lorenz, “Transducerless field orientation concepts employing saturation-induced saliences in induction machines,” IEEE Trans. Ind. Applic., vol.32, no.6, pp. 1380-1393, Nov./Dec. 1996.
2 N. Teske, GM Asher, M. Sumner and KJ Bradley, “Encoderless position estimation for symmetric cage induction motor under loaded conditions,” IEEE Trans. Ind. Applic., vol. 37, no. 6, pp. 1793-1800, Nov./Dec. 2001.
3 F. Briz, MW Degner, A. Diez, and RD Lorenz, “Measuring, modeling, and decoupling of saturation-induced saliencies in carrier-signal injection-based sensorless AC drives,” IEEE Trans. Ind. Applic., vol. 37, no. 5, pp. 1356-1364, Sept./Oct. 2001.
4 J.-I. Ha, S.-K. Sul, K. Ide, I. Murokita, and K. Sawamura, “Physical understanding of high frequency injection method to sensorless drives of an induction machine,” in Conf. Rec. IEEE-IAS Annu. Meeting, Vol. 3, Rome, Italy, Oct. 2000, pp. 1802-1808.
5 S.-I. Yong, J.-W. Choi, S.-K. Sul, “Sensorless vector control of induction machine using high frequency current injection,” in Conf. Rec. IEEE-IAS Annu. Meeting, Vol. 1, Oct. 1994, pp. 503-508.
