Control Methods for Long-Primary Synchronous Linear Motors without mechanical velocity- or position-sensors

点击翻译按钮获取中文摘要

Electrical motor drives require the mover position to be a known variable in order to control them. This variable can be determined directly by means of position sensors or indirectly by measuring the stator’s voltages and currents. As position sensors are expensive and reduce the system’s reliability, the indirect methods, also known as sensorless methods, are preferred.Avoiding position sensors is more significant in linear drives than in rotative ones, because unlike rotative drives that require only one sensor unit attached to the rotor shaft, in linear drives the sensor is spread along the whole track, becoming even more complex, more expensive and less reliable. Even when the research in sensorless methods has been usually focused on rotative motors, the same fundamentals can also be used for linear ones. However, in the case of long-primary motors, additional difficulties must be overcame, e.g.: The position-dependent inductance, used to track the position, is non-periodic and very small due to the fact that only a fraction of the stator is covered by the mover. Mutual inductance due to the end-effect of the stator will interfere with the position detection. In motors arranged in multiple segments, the position must be continuously determined on the segment-transitions. It is proposed to research the characteristics, constrains and opportunities of synchronous longprimary linear motors in relation to sensorless position detection. Based on them, a method to detect position down to zero speed will be developed and tested. Even when focusing on the long-primary linear motor, contributions to the sensorless field in general are also expected.