What is disturbance rejection?
The term "disturbance rejection" was used a number of times in lecture today to describe the purpose of feedback. With feedback, the controller is able to use the output to shape the input of the system. In this way, various disturbances don't affect the system as much and do not create such huge deviations from our desired output. Thus, our system can "reject" the disturbance.
Throughout the course, we will see a number of ways of representing disturbances and dealing with them, but let's look at "Example 2: Speed Control" model from Lecture 1, Slide 10. It shows a block diagram, which represents the disturbance as an additive noise to the system. The disturbance described in this example is the force of a hill.
If Bob were an open-loop control system (which, thankfully, it is not), it could only try to get itself to a particular velocity by using some sort of schedule to accelerate itself. However, in this case, Bob would not know how to adjust for any disturbances (such as the presence of a hill) that would affect the velocity as well.
Thus, by closing the loop on the system, Bob continually measures its velocity to determine how much it should accelerate to achieve its desired velocity. This feedback allows Bob to reject (not be affected by) the disturbance of the hill.