Before doing a plate solve, be sure that you have tack sharp focus. Use a Bahtinov mask. ASIAIR does not recognize blurry blobs as stars.
It is best not to plate solve with a very bright star in the plate. To be able to collect suficient number of other dimmer stars, that bright star would need the be very over exposed, and saturated stars are also not recognized -- the tails of the point spread function makes the way-over-saturated star very bloated, even if it is in focus. Make sure you have lots of non-saturated stars (like 20 to 50) that is useful for the database search.
Too many saturated stars is bad for plate solving (not enough non-saturated stars left in the database), so is too few stars above the noise floor (not enough stars for matching the database, period). So, change the gain/exposure around a bit. If there is no especially bright star in the plate, just adjust the gain/exposure so that the max ADU is just under 65535 (i.e., maximum number of stars above the noise floor, with non of them saturating).
Be sure tracking is turned on. Star trails can also render a star unrecogizable.
Do not try to plate solve when there are tree branches or chimneys sticking into the frame. They will add extraneous star-like objects in addition to obstructing the real stars, and that will throw off any asterism search. Clouds also confuses the asterism search.
Also make sure that you are at least 30º in declination away from the poles (farther if your FOV is large). Other plate solve systems include the database of star asterisms near the pole, but the ASIAIR does not.
If all else fails, try the web plate solver at astrometry.net . Just take a plate (JPEG, PNG, TIFF...) and send it to astrometry.net. It can take a while, but it will solve almost anything, and also tells you how far from rectilinear (geometrically distorted) your plate is.
Chen