LAUSR

Abstract Control of physiological systems, like tumor growth dynamics, can contribute to modern medicine by designing optimized therapies or automating treatments. Typical challenges in physiological control are the positivity of the input and the interpatient variability of model parameters. We use positive dynamics extension to ensure the positivity of the input, and design a robust controller for a nominal model acquired after exact linearization and stabilizing feedback. During the controller design, we minimize the effect of the model perturbation in the worst-case sense and minimize the energy of the performance criteria. The results show enhanced performance compared to our similar robust control approach where only the performance was minimized in the worst-case sense with similar characteristics in the input signals.