Hermite neural network-based second-order sliding-mode control of synchronous reluctance motor drive systems

International audienceThis paper proposes a novel Hermite neural network-basedsecond-order sliding-mode (HNN-SOSM) control strategy for thesynchronous reluctance motor (SynRM) drive system. The proposedHNN-SOSM control strategy is a nonlinear vector control strategyconsisting of the speed control loop and the current control loop.The speed control loop adopts a composite speed controller, whichis composed of three components: 1) a standard super-twistingalgorithm-based SOSM (STA-SOSM) controller for achieving the rotorangular speed tracking control, 2) a HNN-based disturbanceestimator (HNN-DE) for compensating the lumped disturbance, whichis composed of external disturbances and parametric uncertainties,and 3) an error compensator for compensating the approximationerror of the HNN-DE. The learning laws for the HNN-DE and the errorcompensator are derived by the Lyapunov synthesis approach. In thecurrent control loop, considering the magnetic saturation effect,two composite current controllers, each of which comprises twostandard STA-SOSM controllers, are designed to make direct andquadrature axes stator current components in the rotor referenceframe track their references, respectively. Comparativehardware-in-the-loop (HIL) tests between the proposed HNN-SOSMcontrol strategy and the conventional STA-SOSM control strategy forthe SynRM drive system are performed. The results of the HIL testsvalidate the feasibility and the superiority of the proposedHNN-SOSM control strategy