Ricerc@Sapienza

Electric drives are pervasive in many industries, from small domestic appliances to multi-MW compressors, wherever motion control is required. Closed loop speed control requires speed feedback; on the other hand, position feedback is required for field orientation in AC drives, for commutation in BLDC drives and for any application requiring motion control. Due to cost, reliability and space constraints, the only motion sensor used in electric drives is the position sensor: speed has to be estimated from the position measurement. Regardless of the specific technology used for the sensor, the quality of position measurement is fully assessed by three independent metrics: resolution, absolute accuracy and differential accuracy. Resolution, in particular, has a strong impact on both position and speed control, however a complete understanding of its effects on electric drive operation has still to be obtained. One fundamental, yet open issue is: how does a specific resolution impact the performance of a speed-controlled drive under periodic torque disturbances? Such knowledge would be of great help when choosing the position sensor in the drive design stage, in order to minimize the cost and achieve the required disturbance rejection at the same time. To this end, analytical formulations of both quantized position and speed, which are valid under generic periodic torque disturbances, were derived in a recent publication by the PI.
In this project it is intended to experimentally verify the above formulations on an experimental drive test-bench available in the Electrical Machines Lab, which was purchased in part with a 2014 university research grant.