Bifurcations are detected in experiments involving the electrochemical oxidation of copper in phosphoric acid. Ideally, a bifurcation detection protocol requires a model with good qualitative and quantitative prediction capabilities that is also able to reproduce the drift inherent to the electrochemical experiment. A generic protocol for manipulation of nonlinear systems is proposed to achieve this goal. This protocol is based on a good qualitative predictor coupled to a modified Kalman filter that corrects the quantitative errors of the model point by point, adapting to the slowly varying conditions imposed by the electrochemical drift.