Robust interval observer for substrate state estimation of nonlinear reactive sputter processes

Process monitoring of the substrate surface condition of nonlinear, unstable, uncertain reactive sputter processes is crucial for reproducible film forming. But the substrate state is not easily measurable during deposition. This contribution proposes a new method to design a nonlinear reduced interval observer based on the Berg model including uncertain process parameters. A lower and an upper estimate are generated, and the thereby defined interval is guaranteed to contain the plant state. The input-to-state stability, the framer property and the accuracy of the interval observer are proven. Simulation studies show the applicability of the proposed design method by a narrow estimated interval width.