Telemetry data reflects the spacecraft payload's operational state and the spacecraft's operating condition, telemetry data-driven fault detection of spacecraft has great practicability. Considering that a large amount of redundant information contained in the spacecraft telemetry data, principal component analysis(PCA) was adopted to deal with high-dimension telemetry data's dimension reduction in this paper and low-dimensional feature combination was extracted from the initial high-dimensional data set combination. Then an algorithm for spacecraft fault detection and loctation was given. Finally, the validity and practicability of the algorithms were showed by simulation.