Kalman Smoothing One Step

Problem

Perform one Rauch-Tung-Striebel smoothing step: smoothed = filtered + C * (next_smoothed - next_predicted), where C = p_filtered / p_predicted.

Return the filtered value when p_predicted is 0.

Hint

The smoother gain shrinks with the uncertainty of the predicted next state.

case 1

in: [10,1,12,11,2]

expected: 10.5

case 2

in: [0,2,4,0,4]

expected: 2

case 3

in: [5,0,7,6,3]

expected: 5

case 4

in: [3,1,3,3,0]

expected: 3

Starter codePython

def kalman_smoothing_step(filtered, p_filtered, next_smoothed, next_predicted, p_predicted):
    # Your code here
    pass

4 test cases · run your code against them in the editor