Juan Augusto Paredes Salazar

Juan Augusto Paredes Salazar

I'm a Research Fellow at the University of Michigan with research interests in Guidance, Navigation and Control of UAVs and aerospace systems.

Adaptive Model Predictive Control

Source of Funding: Office of Naval Research (ONR) under grant N00014-18-1-2211 and Air Force Office of Scientific Research (AFOSR) under grant FA9550-20-1-0028.

The aim of the project is to develop an adaptive model predictive control (MPC) algorithm, in which the model is identified online by performing closed-loop linear model identification, for aerospace applications and develop diagnostics to determine the effect of hyperparameters on stabilization performance. In particular, Predictive Cost Adaptive Control (PCAC) is developed, implemented and evaluated as part of this project.

The main tasks performed in this project were the following:

  • Evaluated absolute stability properties of PCAC applied to Lur’e systems (systems with nonlinear feedback).
  • Applied PCAC to a Rijke-tube experiment to test stabilization performance over a range of operating conditions.
  • Applied PCAC to nonlinear longitudinal aircraft dynamics in a simulation environment to test trajectory-following performance.

Papers

  • Absolute-Stability-Based Closed-Loop Stability Analysis of Adaptive Model Predictive Control for Self-Excited Lur’e Systems (More details)
  • Experimental Application of Predictive Cost Adaptive Control to Thermoacoustic Oscillations in a Rijke Tube with Unknown Input Delay (More details)
  • Predictive Cost Adaptive Control of Fixed-Wing Aircraft Without Prior Modeling (More details)