ℒ2 neuro-adaptive tracking control of uncertain port-controlled Hamiltonian systems

This study presents a practical method of neural network (NN) adaptive tracking control of uncertain port-controlled Hamiltonian (PCH) systems. NN is used to compensate for parametric uncertainties and unlike the previous studies, the dynamics of the NN tuning law is driven by both the position as well as the velocity errors owing to the introduction of the information preserving filtering of the Hamiltonian gradient. In addition, the proposed controller achieves the ℒ₂ disturbance attenuation objectives as well as preserves the PCH structure of the system in closed loop. Simulation examples demonstrate the efficacy of the proposed approach.