The performance of adaptive optical system of a ground-based large telescope when observing spatial objectives degraded comparing with observing astronomical objectives . To adjust the parameters of adaptive optical system purposively, and get the best correction result, it was necessary to predict the performance of adaptive optical system for spatial objectives. Presence of the best observation elevation range for spatial objectives having different orbits was validated by analysis of orbits based on the residual wavefront errors after correction. The corresponding calculation method was also concluded. The effects of the variation of atmospheric coherence length on the spatial fitting errors, Greenwood frequency on the temporal fitting errors, signal intensity on the wavefront measurement errors were analyzed based on the 1.23m telescope made by CIOMP. The result showed that the temporal fitting errors were primary compared to the spatial fitting errors, the wavefront measurement errors could be neglected when signal intensity of the spatial objectives varying, and the best observation elevations were different for different spatial objectives, orbits and telescope systems. An experiment system was constructed to test an adaptive optical system for different spatial objective by changing the aperture diameter of the turbulence phase plate, velocity and the light source intensity to simulate the variation of the atmospheric coherence length, Greenwood frequency and the signal intensity individually.