Question in generating the gyro about g_err_b

[chengjunlong0524]

g_err_b(i,:) = ( omega_in_b * gyro_b (i,:)' )' or gyro_err_b(i,:) = ( omega_in_b )';

[rodralez]

omega_in_b is the projection of both Earth and transport rates in the body frame.

It is the inverse operation of:

wb = (wb - DCMbn' * (om_ie_n + om_en_n)); % Titterton, Eq. 3.29, p. 32

in att_update().

[chengjunlong0524]

邮件已收到！！！

[chengjunlong0524]

Outputs from three ideal orthogonal gyroscopes are expressed as:

omega_in_b is the projection of both Earth and transport rates in the body frame, ie

What does the "g_err_b" means in code? The finall output is

wb_sim = gyro_b + g_err_b + g_wn + gb_sta + gb_dyn + g_rrw

gyro_b is same as omega_in_b? What does the code "g_err_b(i,:) = ( omega_in_b * gyro_b (i,:)' )'" means?

[chengjunlong0524]

Any idea?

[rodralez]

Hello,

I think we've got a bug here. Thank you very much for spot it.

I modified files gyro_gen.m and acc_gen.m in branch develop.

Please, check them and let me know your comments.

[chengjunlong0524]

Thanks for your reply. In noise_rrw.m, it probably should be

rrw_n (j, i) = rrw_n(j-1, i) + rrw(i) * sqrt(dt) .* b_noise(j-1);

See that the unit of rrw may be rad/s/root-Hz, rad/s^2/root-Hz, m/s^2/root-Hz or m/s^3/root-Hz. In another form, it is rad/sqrt(s), rad/s/sqrt(s), m/s/sqrt(s) or m/s^2/sqrt(s).

In addition, angle rate random walks(rad/s^2/root-Hz or rad/s/sqrt(s)) seems also reflects bias instability. Are gb_dyn and gyro_rrw repeated? This question also applys to acc_gen.

[chengjunlong0524]

邮件已收到！！！