Passivity-Based Output Feedback Control for Systems with Nonlinear Outputs via High-Gain Observers

In this paper, a high-gain observer with nonlinear output is designed. The scaled estimation error system is constructed with a passive function based on the nonlinear output function and a strictly positive real transfer function of boundary-layer system. The ultimate boundedness and exponential stability of the estimation error for the global and regional two cases are demonstrated, as long as high-gain observer's decay rate is fast enough. For the regional case, due to the restriction on the passive function, the estimation error has a region of attraction which is a subset of the intersection of a positively invariant compact set and the strip coming from the restriction. The extended results under passivity of the output function and strictly positive realness of the transfer function are presented. The performance recovery property of the output feedback using high-gain observer with nonlinear output is validated. Some examples are applied in the simulation to illustrate the proposed results in this paper.