On the design of nonlinear controllers for Euler-Lagrange systems (Q2755705)

The authors consider a control governed by an Euler-Lagrange equation NEWLINE\[NEWLINE{\partial{\mathcal L}\over\partial q} (q,\dot q)-{d\over dt} \Biggl[{\partial{\mathcal L}\over\partial\dot q} (q,\dot q)\Biggr]- {\partial{\mathcal F}\over\partial\dot q}(\dot q)+ Mu= 0,NEWLINE\]NEWLINE where \(q\) is the vector of generalized coordinates, \({\mathcal L}(q,\dot q)={\mathcal T}(q,\dot q)-{\mathcal V}(q)\) is the Lagrangian and \({\mathcal F}\) is Rayleigh's dissipation function. The kinetic energy \({\mathcal T}\) is quadratic. The problem studied in the paper is the construction of a dynamic controller NEWLINE\[NEWLINE\dot z= \varphi(z, w),\qquad u= \psi(z,w)NEWLINE\]NEWLINE guaranteeing that every bounded solution converges to an equilibrium point of the closed loop system. The result is applied in particular to systems with several nonlinear amplifiers.