Accepted at IEEE IV 2026

In this paper, based on a two-branch model, an online internal voltages estimation method of a supercapacitor is proposed. The goal is to provide a fast and accurate online estimation of such voltages as they are useful for estimating its State Of Health (SOH) and/or State Of Charge (SOC). An hybrid nonlinear observer is designed to estimate voltages using sampled measurements. This kind of observer uses a correction term in the prediction step with the aim of performing with large sampling period, then it is suitable for efficient online estimation. Simulation results highlight good performances of the observer in online parameters estimation in the case of fast dynamics inputs under NEDC and WLTP cycles.

In this paper, an inter-sampled observer-predictor is designed for a class of system which is characterized by sampled data measurements. The main drawback of numerous approaches lead to small or limited sampled time period, which is an important point in industrial processes. To overcome this difficulty, we propose a modified inter-sampled observer-predictor which ensures an exponential convergence and gives the ability to push the sampled time period to its limit. Comparing to other observer structures, our design is char-acterized by its simplicity because we only need a tuning gain. Our main contribution is to show that our modified observer allows to widen the sampling period as much as possible, while guaranteeing the exponential convergence of our observer. The effectiveness of our proposed approach is shown through numerical simulations which have been performed on a well-known bioreactor example.