Some new research in Control System
Intelligent control.
This is via a long way the most interesting topic to me, for my part.
Control structures is involved with designing techniques for modeling and constructing
computerized systems, and controlling them.
While for simpler goals and constraints this can be achieved through
rigorous, explicit mathematical instructions (think about a robot, for example), for more
complex tasks conventional control just doesn’t work. You can’t explicitly program a car to
navigate itself through a road while obeying traffic rules in any feasible amount of time.
Yes, control does have some more sophisticated control strategies than
“mindless,” control like PID, such as Model Predictive Control (MPC), which optimizes system
input based on analyzing predicted system output. But even these are very limited when compared
to the capabilities of artificial intelligence and machine learning methods.
That’s where intelligent control comes in. Integrating new techniques
from artificial intelligence and machine learning to conventional control system design.
Fascinating stuff.
Hybrid Control
Prior to hybrid manipulate, manage turned into strictly implemented to either discrete or
non-stop dynamical systems. But this posed a trouble, due to the fact maximum present day
manipulate systems contain discrete AND non-stop elements, no longer just one of the .
For example, reflect onconsideration on a pc that is the use of some algorithm to optimize
manage enter for a car. The automobile’s dynamics are continuous, but the computer’s actions
occur discretely.
Since most modern systems contain both discrete dynamical components and continuous dynamical
components, hybrid control was developed, which is capable of modeling hybrid control systems
(control for hybrid dynamical systems, where “hybrid dynamical,” means both continuous and
discretely evolving parts).