Time-Constrained MPC Approach for Switched Nonlinear Systems With Applications to Biomedical Problems

Switched systems are important for modeling biomedical control problems, where the control action can be considered to be a switching signal to select the active mode (e.g., a drug therapy or intervention). Practical implementations impose constraints not only on the variable magnitudes, but also on each mode’s active and inactive times (AT and IT). To address this, a model predictive control strategy is proposed using an enlarged model with integer state variables to track past AT/IT for each mode. Two biomedical applications were selected to demonstrate the controller’s effectiveness through simulations. The results highlight that our approach is suitable for biomedical applications with intricate temporal requirements.