Nonlinear model identification and state-observer design for water distribution systems

This paper discusses the nonlinear model identification and the state-observer design for water distribution systems. A nonlinear model of the flows dynamics is put forward, on the basis of the Saint-Venant equations. Since some of the model parameters are not known or may be slowly time-varying, an identification procedure is needed. The author´s approach is based on output-error minimization, by use of nonlinear programming techniques. On the basis of the identified model, the problem of state-observer design is then stated.