TEOMAPIN

Optimized determination of the position of sensors and actuators based on maximum achievable performance: This objective was addressed theoretically, and its results were applied to two different Active Noise Control problems, namely the motorcycle and acoustic duct cases. One of the most significant outcomes is a set of measures that allow for comparing the quality of sensor and actuator placement based on practical criteria in performance, robustness, and controller complexity.

Identification of multivariable dynamic systems (MIMO), nonlinear and/or time-varying, represented in continuous time, discrete time, or hybrid combinations. Each application has posed restrictions and highlighted numerous gaps between theory and practice. Specifically, theoretical analysis and practical solutions have addressed identifying plants with variable delays in MIMO systems (irrigation channels), identifying nonlinear dynamics of acoustic noise propagation in various spaces (helmets, ducts, and chambers with windows), and finally, handling identification in UAV-type helicopters, with the complexity of a multivariable, coupled, and unstable system. In this last case, a methodology was developed that allows the application of classical techniques.

Validation (and invalidation) of the above systems against experimental data, considering variables such as external noise bounds (additive) and uncertainty (global or structured, time-varying or invariant, linear dynamic or static nonlinear). A notable contribution in this area is a robust identification/invalidating method for Linear Parameter-Varying (LPV) models. This method complements nominal LPV identification algorithms by adding uncertainty and noise bounds that generate a set of models consistent with experimental evidence.

Numerical issues associated with the use of computers or DSP in the analysis, synthesis, and implementation of controllers. The application of active control and UAV control requires real-time implementation and/or significant computational limitations that prevent the use of arbitrarily large or complex controllers. Additionally, these systems tend to have positive real-part zeros and lightly damped oscillations. Therefore, the design of a low-order controller is a requirement that poses a significant challenge to control theory.

Determining the most suitable type of uncertainty to describe a physical system through a family of models, and computing the least conservative bounds possible. Here, robust and nominal stability and performance have been analyzed in models with uncertain delays, for global dynamic and structured parametric uncertainty structures.