An approximate high gain observer for speed-sensorless estimation of induction motors

Adaptive non-linear high gain observer based sensorless speed estimation of an induction motor

“…The structure of (46), based on the first approximation is governed by the consideration of reducing complexity without requiring the observer convergence. The motivation for second approximation is safety computational considerations without requiring observer efficiency [20].…”

“…Another important challenge of this technique is its instability under regenerating conditions [17]. Furthermore, it fails to offer satisfactory compromise between the robustness to measurement noise and estimation bandwidth [20].…”

“…Significant effort is spent in [8,10,20] to approximate the standard high gain observer with fixed gains for sensor-less speed estimation of an IM by exploiting a suitable procedure to tackle the singularities occurring whenever the observability is lost at zero and very low speeds. However, as shown in [8,10,20], the approximate observer still suffers from the so-called peaking phenomenon, highfrequency oscillation and large uncertainties especially in the unobservable regions. Indeed, as mentioned in [20,36,37,39], such drawbacks are due to the very large value of the observer gains which remain unchanged even in the presence of noise.…”

“…However, as shown in [8,10,20], the approximate observer still suffers from the so-called peaking phenomenon, highfrequency oscillation and large uncertainties especially in the unobservable regions. Indeed, as mentioned in [20,36,37,39], such drawbacks are due to the very large value of the observer gains which remain unchanged even in the presence of noise. Moreover, authors in [8,10] use a limited type of high gain observer which is based on the system structure assuming a complete triangular observable form in which the dynamics of non-linear part of the system have triangular state dependence.…”

“…Moreover, the authors do not include any detailed study of the Jacobian matrix inverse to tackle the singularities occurring during the operation of the IM, and the analysis of state estimation convergence of the proposed observer is not included. o in the framework of an industrial benchmark stated in [13], the trajectories have been used to test and evaluate the performance of the proposed observer when the IM remains in the unobservable conditions such as in the zero to very low speed range; o the validity of the proposed observer is confirmed also by its comparison with the standard high gain observer which has been performed by using a constant value for the gain parameter [8,10,20] .…”