Feedback-based mitigation of torque harmonics in switched reluctance motor drives.

摘要

The Switched Reluctance Machine (SRM) has numerous advantages over alternative machines such as ease of construction, durability in harsh environments, and tolerance to winding faults. Although the advantages can be significant, an issue that often arises in SRM drives is excessive vibration/noise created by harmonics of electromagnetic torque (torque ripple). Over the past decade, a significant body of research has appeared that develops methods for mitigating torque harmonics and has been concentrated on two main areas: open-loop control and estimation-based feedback control. Both require detailed knowledge of machine parameters and are therefore sensitive to parameter variation.; In this research, closed-loop control techniques are developed to eliminate torque harmonics produced by an SRM. Toward this objective, a piezoelectric washer is first designed and incorporated into an SRM drive system to sense harmonics of torque-ripple-induced vibration. The advantages of the sensor include flexibility in terms of placement on a drive, relatively low-cost, and durability. Simulation and laboratory experiments are presented that validate that the sensor output is correlated to torque ripple.; Using the sensor for feedback, two control techniques are developed. In the first, a gradient algorithm is used to establish current harmonics that minimize a cost function based on measured vibration harmonics. In the second, analytic techniques are used to map vibration harmonics to current magnitude and phase angle. The map is used to formulate a proportional plus integral (PI) control that adjusts current amplitude and phase in response to measured vibration.; Hardware and simulation results validate the effectiveness of both controllers. In addition, they are used to demonstrate tradeoffs in computational complexity, convergence rate, and resilience to machine parameter variation.