Adaptive field-oriented control of current-fed induction motors: a discrete-time implementation

Abstract

Indirect field oriented control (IFOC) is the industry standard for high performance current-fed induction motors. It is well-known that this controller preserves stability in the face of large rotor resistance variations – which are inevitable in practical scenarios – but its performance is significantly degraded. To overcome this drawback many adaptation schemes have been reported, both in drives and control-oriented publications, with the overwhelming majority of these developments done in continuous-time. This, in spite of the fact that the IFOC, as well as their proposed adaptive versions, involve the solution of highly nonlinear differential equations, whose discretization is far from obvious. In this paper we try to reverse this trend and present a discrete-time estimator of the motor parameters and an adaptive IFOC based on it. It is shown that the proposed estimator and adaptive controller are globally convergent under very weak prior knowledge and excitation assumptions. Simulations, that validate the theoretical claims and illustrate the robustness of the schemes, are also presented.

Keywords:

• Nonlinear models with uncertainies
• nonlinear control
• adaptive and robust control
• parameters estimation
• control applications
• induction motors

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 By exact discretization it is meant that the trajectories of the continuous-time model (1(1) $\begin{array}{rl}\dot{\lambda }& =-\frac{R}{L}\lambda +\frac{R}{L}u\\ \tau & =\frac{n_p}{L}{u}^{\mathrm{\top }}\mathrm{J\lambda }\\ D\dot{\omega }& =\tau -{\tau }^L\end{array}$(1) ) $x(t)$ exactly coincide with the ones of the discrete-time model (2a(2a) $\begin{array}{rl}{\lambda }_{k+1}& =a{\lambda }_k+(1-a)u_k\end{array}$(2a) ) at the discrete time instants, that is $x(t){|}_{t=\mathrm{kT}}=x_k$.

2 Interestingly enough, this coupling reflects the well known fact that to avoid magnetic saturation the flux norm must be increased when the machine is requested to deliver large torques, i.e., the flux weakening policy.

3 ${\mathbf{q}}^{\pm n}[x(k)]:=x(k\pm n)$ for some sequexnce $x(k)$ and integer number n.

Additional information

Funding

The work was written with the financial support of the Ministry of Science and Higher Education of the Russian Federation, agreement 075-11-2023-015, 10.02.2023. This work was also supported by 111 project, No. D17019.