In conversions/imu_si_errors.m at line 95 and 101 while calculating Dynamic Bias PSD calculation units are not matched with equation.
imu_si.ab_psd = imu_si.ab_dyn ./ sqrt(imu.ab_corr); % m/s^2/root(Hz)
--- Actual units will be % m/s^2/root(time) ....
--- correct Equation will be imu_si.ab_psd = imu_si.ab_dyn .* sqrt(imu.ab_corr); % m/s^2/root(Hz)
Please check .
same for gyro bias PSD.
In kf_update.m why below equation is being used instead of Joseph stabilized version which is correct as per many references. Is there any advantage with it.?
% Step 5, update the a posteriori covariance matrix, Pp
kf.Pp = kf.Pi - kf.K kf.S kf.K'; % Eq. 3.10 from Särkkä
Hi,
In conversions/imu_si_errors.m at line 95 and 101 while calculating Dynamic Bias PSD calculation units are not matched with equation. imu_si.ab_psd = imu_si.ab_dyn ./ sqrt(imu.ab_corr); % m/s^2/root(Hz)
--- Actual units will be % m/s^2/root(time) .... --- correct Equation will be imu_si.ab_psd = imu_si.ab_dyn .* sqrt(imu.ab_corr); % m/s^2/root(Hz)
Please check . same for gyro bias PSD.
In kf_update.m why below equation is being used instead of Joseph stabilized version which is correct as per many references. Is there any advantage with it.?
% Step 5, update the a posteriori covariance matrix, Pp kf.Pp = kf.Pi - kf.K kf.S kf.K'; % Eq. 3.10 from Särkkä
Thank you, K.Venkatarao.