<p dir="ltr">With advancements in aerospace technology and the widespread deployment of low Earth orbit constellations, challenges to astronomical observation and deep space exploration have intensified. The need for accurate orbital data on space targets, as well as the analysis of satellite positioning, constellation configurations, and deep space satellite behavior, has grown more critical. However, there is a significant gap in publicly available, real-world datasets to support research on space target maneuver prediction and anomaly detection. This paper addresses this gap by collecting and organizing representative maneuver data from the Starlink satellites, which includes precise star-tracking predictions and Two-Line Element (TLE) catalog orbital data. This dataset offers a more realistic and multi-dimensional scenario for modeling space target behavior. It provides valuable insights into practical deployment of maneuver detection and anomaly detection approaches.</p>