Efficient predictive model-based and fuzzy control

Anahita Jamshidnejad,, Technical University of Delft


The main focus of my presentation is on efficient, integrated model-based, predictive, and artificial intelligence-based  control approaches for systems with a highly nonlinear behavior and multiple control sub-objectives (in particular urban traffic networks).

My presentation will include three parts:
- In the first part, I will introduce general approaches for recasting nonsmooth optimization problems into smooth
ones, in order to solve them efficiently by gradient-based optimization methods. 

- In the second part, I will discuss a novel integrated control architecture that consists of several parameterized and
optimization-based controllers in two separate "base" and "parallel" blocks. The main aim of using this architecture is
to effectively control the behavior of complex systems in real time.

- In the third part, a coordinative multi-agent architecture is proposed that integrates predictive control methods
with fuzzy control approaches to develop a two-layer adaptive control system. Potential applications of the proposed
control architecture include search-and-rescue missions, deep space exploration, underwater operations,
and autonomous aircraft.