Asymptotically convergent switching differentiator

AbstractSummaryA novel switching differentiator with a considerably simplified form is proposed. Under the assumption that the time‐derivative of a time‐varying signal has a Lipschitz constant, it is shown that estimation error is asymptotically convergent to zero. The estimated derivative shows neither chattering nor peaking phenomenon. A differentiator that estimates the first‐order derivative of a signal is first proposed, and, by connecting this differentiator in series, higher‐order derivatives also become available. Simulation results show that the proposed differentiator shows extreme performance improvements compared to the previously established widely used differentiators such as high‐gain observer or higher‐order sliding mode differentiator.