Observers for Sensorless Synchronous Motor Drives: Framework for Design and Analysis

Abstract

Tuovinen, T. (2018)Abstract-This paper deals with the speed and position estimation for synchronous reluctance motors (SyRMs) and interior permanent-magnet synchronous motors (IPMs). A unified design and analysis framework for a class of back-electromotive-force (back-EMF)-based observers is developed and the links between apparently different estimation methods are brought out. State observers equipped with a speed-adaptation law are shown to be mathematically equivalent to voltage-model-based flux observers equipped with a position-tracking loop. The error signal driving the adaptation law or the tracking loop is presented in a generalized form. Using the framework, a stabilizing gain design is reviewed and detailed design guidelines are given. Selected observer designs are experimentally evaluated using a 6.7-kW SyRM drive and a 2.2-kW IPM drive.

Permanent-Magnet Flux Adaptation for Sensorless Synchronous Motor Drives

“…1(b) shows the internal structure of the observer, operating in estimated rotor coordinates. The flux observer is defined by [4]…”

“…1) Flux Observer: The first condition in (17) is also met, if the observer gain matrix is selected as [4], [16]…”

“…where the coefficients b ≥ 0 and c ≥ 0 are related to the pole locations of the flux observer. If the PM-flux adaptation bandwidth is zero (a = 0), the observer gain reduces to the gain in [4], [16], where no PM-flux adaptation is used.…”

“…In order to keep the observer gain (19) within reasonable limits, the design parameters b and c should be selected such that the closedloop poles of the flux observer remain in the vicinity of the open-loop system poles, which can be solved from det(sI − RL −1 − ω m0 J) = 0. The damping of the open-loop poles decreases as the speed increases [4]. In the closed-loop system, it is favorable to increase the damping of the poles at higher speeds.…”

“…As an example, removing the need for a fragile motion sensor makes it possible to reduce the overall size and to improve reliability in traction drives. Rotor-position estimation for sensorless control is typically realized using a back-electromotive-force (back-EMF)-based observer [1]- [4], augmented with a signalinjection method for operation at lowest speeds.…”