Estimation of White Noise and Random Walk Innovation Variances \(R\) and \(Q\)

Source: R/VaccaroZaki.R

This function provides an estimate based on the Allan variance for the white noise and the random walk innovation variances \(R\) and \(Q\). The implementation reflects Algorithm 1 in Richard J. Vaccaro and Ahmed S. Zaki, "Reduced-Drift Virtual Gyro from an Array of Low-Cost Gyros".

algorithm_1(X, remove_last_scale = FALSE)

Arguments

X

A matrix of dimension T by p, where T is the length of the time series and p is the number of processes.

remove_last_scale

wether the last scale of the Allan variance should be removed

Value

A list with the following structure:

  • R: A vector of size p of the estimated white noise variances.

  • Q: A vector of size p of the estimated random walk innovation variances.

Author

Davide Antonio Cucci