Robotics Basics - Control Paradigms

Feedback Control 20210709201550 is a great way to tell a controller how to perform a single behavior and to keep that behavior. However, we need a way to select the best behavior to achieve our current goal, and this is the job of the Control architecture.

A paradigm is a set of assumptions and techniques which are used to characterize a certain problem. An architecture is a set of different architectural component and it’s interaction, and they are used to describe how to apply a paradigm, specifying the concrete elements.

There are different paradigms for robot control, which can have many different types of specific architectures:

These paradigms types differ in three major aspects:

  • Time: how fast has the robot to respond and how quickly it senses
  • Modularity: how the system is divided in pieces and how they interact with each other
  • Representation: how information is stored/encoded 20210711201755

Notes References

20210514183815 INDEX - Robotics

20210711201755 Robotics Basics - Representation

20210709201550 Robotics Feedback Control - Basics

20210711201454 Robotics Basics - Deliberative Paradigm

20210711201545 Robotics Basics - Reactive Paradigm

20210711201605 Robotics Basics - Hybrid Paradigm

References

(Murphy 2000)

Murphy, Robin. 2000. Introduction to AI Robotics. Intelligent Robotics and Autonomous Agents. Cambridge, Mass: MIT Press.