>> There still isn't correction in asiair database.
I wonder if the ASIAIR is not using the current precession Epoch (JNOW), and instead sticking with the year 2019.5 that many databases start with before epoch conversion. In the example shown by Mr. Jiang, he had compared the ASIAIR coordinates with the 2019.5 info in SkySafari.
However, in SkySafari, there is a way for the user to ask the app to convert to whatever epoch (current epoch checkbox == JNOW). We don't know what epoch ASIAIR uses.
I just checked SkySafari with Kochab. With the reference 2019.5 epoch, the SkySafari object database shows a declination for Kochab of 74 deg 4' 39". When I slewed to Kochab using J2000, SkySafari showed 74 deg 9'. When I chose JNOW (current epoch), SkySafari showed 74 deg. 5'. There is significant differences in 20 years.
In addition to precession, there is also the proper motion of the stars (also shown in the SkySafari object database). But those are relatively small compared to precession changes. So I don't think that is the error that you found with ASIAIR.
On top of that, there is atmospheric refraction (which affect stars closer to the horizon), but since I was earlier told that the ASIAIR corrects for it in the polar alignment process, ZWO is probably also accounting for it in GOTO.
Jiang xian sheng: could you please check with your team what Epoch is used when converting the original star database to the true star position? If the current Epoch (JNOW) is not used, that can also affect the accuracy of calculating the NCP during polar alignment, .
Clear skies,
Chen