Control is the process of converting intentions into actions. We use control to move the robot with respect to the environment but also to articulate sensor heads, arms, grippers, tools, and implements. Dynamic models are useful in control for purposes of analysis but they are also used explicitly in refined implementations.

Classical Control

The main objective of a control system is to provide the inputs necessary at the hardware level that will generate the desired motions. This section describes the methods used to implement the reactive autonomy layer that was initially described in Chapter 1.

Introduction

There are many motivations for the use of controllers on mobile robots:

• Real hardware ultimately responds to forces, energy, power, etc, whereas we are usually concerned with positions and velocities etc. Controllers map between the two.

• Measurements can be used to measure what the system is doing, thereby reject disturbances, and even alter the dynamics of the system in some favorable manner.

• Models of the system can be used to elaborate a terse description of the desired motion into the details necessary to make it happen.