Thanks to Udacity.
1 dimensional Kalman Filter
def update(mean1, var1, mean2, var2):
new_mean = (var2 * mean1 + var1 * mean2) / (var1 + var2)
new_var = 1/(1/var1 + 1/var2)
return [new_mean, new_var]
def predict(mean1, var1, mean2, var2):
new_mean = mean1 + mean2
new_var = var1 + var2
return [new_mean, new_var]
for i in range(len(measurements)):
[mu, sig] = update(mu, sig, measurements[i], measurement_sig)
[mu, sig] = predict(mu, sig, motion[i], motion_sig)
No comments:
Post a Comment