Memory Augmented Adaptive Identification and Control of Switched Euler–Lagrange Systems

This letter proposes an online adaptive identification and control method for a switched Euler-Lagrange systems ensuring global exponential stability of the tracking error, filter tracking error and parameter estimation errors to the equilibrium point. Parameter estimators are designed for every subsystem to help with parameter learning during both the active and inactive phases of the subsystem. Parameter convergence to true values is analyzed without the persistence of excitation (PE) condition on the regressor with the dual layer low pass filter architecture. Inspired by the author’s previous work on switched linear systems, this letter extends the idea of memory augmentation for parameter learning in the inactive phase of the subsystem and intermittent initial excitation (IIE) condition to relax the PE condition to switched EL systems. The combination of memory augmentation and IIE conditions helps achieve the global exponential stability (GES) of the overall error of switched EL systems. A numerical simulation is presented to demonstrate the efficiency of the proposed algorithm.