Looking back at the previous content, we know that there is a measurement update and a motion update (prediction).
The motion update is done by full probability or an addition, we have solved the more complicated case.
I solved it and came up with the formula.
And also implemented this step with code.
The movement part doesn't want to go too deep, it's a very simple step, let's write it down. Suppose you live in a world where the center point is your best estimate of where you are, and the Gaussian function is your error.
Suppose you moved a certain distance to the right, and the movement itself contains its own error, then you arrive at a predicted position that adds the amount of movement to the mean, and this also adds an error to the initial error.
If you move this distance to the right (green line), you are expected to get where you want to be, but you will lose information, because your movement would have lost information, which is reflected in the error here.
This is easy to do mathematically, the new mean is equal to the old mean plus the action, often called U.
Suppose you moved 10 meters here would be the green line would be 10 meters, the new sigma squared equals the old sigma squared plus the variance of the Gaussian distribution of the movement.
That's all you need to know, it's just an addition.
test:
