The CTRV Model

Motion Models and Kalman Filters

In the extended kalman filter lesson, we used a constant velocity model (CV). A constant velocity model is one of the most basic motion models used with object tracking.

But there are many other models including:

constant turn rate and velocity magnitude model (CTRV)
constant turn rate and acceleration (CTRA)
constant steering angle and velocity (CSAV)
constant curvature and acceleration (CCA)

Each model makes different assumptions about an object's motion. In this lesson, you will work with the CTRV model.

Keep in mind that you can use any of these motion models with either the extended Kalman filter or the unscented Kalman filter, but we wanted to expose you to more than one motion model.

Robot Motion and Trigonometry

Motion model development relies on some essential concepts of trigonometry. As a trigonometry refresher in the context of robot motion, we have created this optional content .

Limitations of the Constant Velocity (CV) Model

CTRV

Assume a vehicle drives straight at first and then goes into a turn. If we apply a Kalman Filter to track the vehicle (using the process model from the last lesson, which assumes constant velocity), what do you expect to happen with our estimation result for the vehicle position?

The position estimation would tend to result outside of the actually driven circle.

The position estimation would tend to result inside of the actually driven circle

The position estimation would result correctly on the actually driven circle

It is impossible to say what would happen in this case.

SOLUTION:

The position estimation would tend to result outside of the actually driven circle.