Search

We present a method of stabilizing single-input (approximately) feedback linearizable systems in the generic situation where the linearizing feedback and coordinate transformation map are not directly available. In obtaining a stabilizing feedback, we utilize a technique of an approximate feedback linearization which is based on an approximate integrating factor calculated from a recursive algorithm. We construct a feedback controller for the approximately linearized system so that the system is asymptotically stable in a neighborhood of the origin where the one-to-one relation holds with the original coordinate system.

In this paper, a new mathematical modeling for a six-phase interior permanent magnet synchronous motor (IPMSM) is presented. The proposed model utilizes two synchronous reference frames. First, the flux model in the $abcxyz$ frame is mapped into the stationary $dq$ frames and then to two synchronous rotating frames. Then, differentiating the flux models, voltage equations are derived in rotating frames. Through this analysis, the interaction between the $abc$ and $xyz$ subsystems is properly described by a coupling matrix. The torque equation is also derived using the two reference current variables. Flux model was verified through FEM analysis. Experiments were done using a 100 kW six-phase IPMSM in a dynamo system. The validity of the torque equation was checked with some experimental results under a shorted condition on an $xyz$ subsystem. [ABSTRACT FROM AUTHOR]

A current controller having an inverse dynamic model and a proportional-integral (PI) controller has been proposed with the aim of compensating the effects of deadtime. The proposed control scheme, having two degrees of freedom, can be used for compensating the effects of disturbance without degrading tracking performance. To be specific, the disturbance-rejection property is enhanced by increasing the gains of the PI controller, but the PI gain does not affect the transfer function between the current command and its output. By nature, it is robust to the parameter mismatch error between the model and the plant, since the mismatch error can be counted as another disturbance. With the experimental results, we have shown the superiority of the proposed method in suppressing the current distortion caused by deadtime. Index Terms--Coprime factorization, deadtime, disturbance rejection, tracking, two-degrees-of-freedom controller, voltage-source pulsewidth modulation inverter, zero-current-clamping phenomenon.

In this paper, the problem of the robust H ∞ finite-time control is investigated for discrete-time polytopic uncertain switched linear systems. The parameter-dependent switched Lyapunov function approach is used to derive sufficient conditions that guarantee the systems to be robust H ∞ finite-time bounded in terms of linear matrix inequalities (LMIs). A striking feature is the use of scalar parameter that leads to LMI conditions that are less conservative than those of previously published conditions. Consequently, the reduced bound of the state trajectory and disturbance attenuation level can be obtained by searching for the optimal scalar value within a certain bound and solving a convex optimization problem. Furthermore, a new robust H ∞ finite-time state-feedback controller is designed based on the proposed conditions. A numerical example is given to demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

In this note, an improvement on the stability result in the above paper is given based on a modified Lyapunov function. It is shown that our result is much less conservative than that in the paper especially when the size of derivative of the time delay increases. For comparison, we give two examples. Index Terms--Uncertain system, delay, stability.