Advanced motion control of PMLSM based on a novel scaleless position estimation method

Abstract

The detent force is the most dominant, strongly position dependent and nonlinear disturbance in iron-cored permanent magnet motor systems. In this dissertation, a novel method to identify and compensate not only the detent force but also the entire external disturbances in a lumped form was proposed. First, a feed-forward based position dependent disturbance compensator was designed. And, an initial position estimator was added to create synchronization between the model and real disturbance because general linear motor systems are equipped with incremental encoder. Furthermore, an adaptive strategy was proposed using an extended Kalman filter to improve the model against the variation in the real circumstances. The effectiveness of the proposed scheme was validated by comparison with other existing researches in the overall disturbance shape and position tracking error.

Meanwhile, the optical encoder which is the most widely used position measurement sensor in motor systems, requires scale proportionally to operating distance. However, its reliability and durability against environmental disturbances are poor. As an alternative, many scaleless position estimation methods using hall effect sensors had been suggested using arctangent method with two sine and cosine signals. However, for combining more than two linear hall sensors, the sensor offset, difference in scale and unwanted phase between them and harmonics decrease the performance. To solve this problem, other studies had focused on signal conditioning method to make ideal sinusoidal signals. However, perfect compensation for those impure components are impossible and there exist certain amount of effects of them. In this dissertation, fast Fourier transform and fixed point iteration method were applied to solve those issues without any additional signal conditioning. The effectiveness of the proposed scheme was validated.

Finally, above two proposed methods, i.e., scaleless position estimator and lumped disturbance compensator for permanent magnet linear motor systems were combined and its effectiveness was also validated.