Abstract: To obtain precise orbit information of space debris from Two-Line Element (TLE) has always been a research focus. Due to a lack of understanding of the thermosphere atmosphere, atmospheric drag is the largest source of error in orbit determination, with errors in atmospheric density models and trajectory coefficients being time-dependent. However, TLE-based orbit determination algorithms have not taken into account the impact of time-dependent errors, resulting in the inability to further improve orbit prediction accuracy. Therefore, this article introduces a time variable and uses the simplex optimization search algorithm to solve it together with other orbit parameters, weakening the impact of time-related errors on orbit accuracy. This article conducted experiments using TLE data and CPF (Consolidated Prediction Format) precise ephemeris data from 8 satellites, and the relative accuracy of orbit prediction considering time variables was improved by 0.11%～78.60%. Therefore, introducing time variables to weaken errors related to time variables can help improve orbit determination accuracy, and research results are expected to be applied in fields, such as atmospheric re-entry forecasting, risk assessment, collision warning, etc.