This thesis presents a unified nonlinear optimization based speed and position estimation method in position sensorless control of interior permanent magnet synchronous motor (IPMSM) drives at wide speed range including standstill.
The existing electromotive force (EMF) based sensorless methods are suitable for medium and high speed operation, but they can't be applied at low speed and standstill condition due to the reduced EMF values. The conventional saliency tracking based sensorless methods usually employ the continuous voltage or current injection at low speed including standstill condition. However, these methods degrade at high speed by introducing higher loss and torque ripples caused by the injection. Additionally, the initial rotor position needs to be detected at the machine startup to avoid the reverse rotation and to guarantee the delivery of the expected torque. Therefore, different position estimation techniques need to be combined in the controller at wide speed range, which increases the control system complexity.
Hence, a unified nonlinear optimization based speed and position estimation method is proposed. At startup and standstill conditions, three steps are employed for initial position estimation. Step I employs pulse voltage injection in the stationary reference frame and a cost function which contains the knowledge of initial rotor position. The rotor position can be estimated by minimizing the cost function with injected voltage and induced current. Since the estimation results in Step I have an ambiguity of 180 degree, a generalized approach to magnetic polarity detection which exploits asymmetries in machine specific differential inductance profiles is employed as Step II. In order to improve the estimation accuracy, continuous sinusoidal voltage is injected in estimated rotor reference frame in Step III. A modified cost function is minimized based on the injected voltage and resulting current. At running state, cost functions which employ both speed and position as decision variables are proposed and utilized for estimation. The speed and position estimation can be delivered by minimizing the proposed cost functions based on the measurements of the stator voltage and current. Since only one position estimator exists in the drive system, the speed and position estimation is unified at wide speed range. The feasibility of the proposed estimation algorithms is validated with the prototype 5 KW IPMSM drives test bench.
In order to benchmark the proposed estimation method, the performance of the proposed method was compared with existing sensorless control methods on the same prototype IPMSM drives test bench. Under the same test conditions, the proposed method outperforms with improved transient performance and steady state accuracy. Moreover, the proposed method is capable of delivering estimation with different voltage injection types and involving the nonlinear motor parameters, which makes this method more flexible in practice. Additionally, the capability of estimating speed and position with low sampling frequency also makes the application of the proposed method promising in high power AC motor drive systems. / Thesis / Doctor of Philosophy (PhD)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/20894 |
| Date | January 2017 |
| Creators | Sun, Yingguang |
| Contributors | Emadi, Ali, Sirouspour, Shahin, Electrical and Computer Engineering |
| Source Sets | McMaster University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0021 seconds