Passivation and control of partially known SISO nonlinear systems via dynamic neural networks (Q5952994)

The passivity theory, which deals with controlled systems whose nonlinear properties are poorly defined, is applied to the single input-single output (SISO) systems. In fact in the paper such a SISO system is represented by a system of two ordinary differential equations (ODEs). Then a simple dynamic neural network (DNN) containing only two neurons and without any hidden-layers, is used to identify the unknown nonlinear system. Alas, a Lyapunov-like analysis is used to formulate a learning law for such singular DNNs of two neurons. Then a passivation of a partially known SISO system via such DNNs is realized. At the end two numerical experiments on a single link manipulator and benchmark passivation are illustrated. As both, the partially known nonlinear system as well as the applied DNN are represented by two ODEs and in both cases stability analysis can be performed even analytically to this end, the results of the paper are in a sense of a trivial nature and could be hardly exploited in more realistic situations.