The Tumbleweed Algorithm (TA) is an excellent newly proposed algorithm that has received a lot of attention in recent years. This proposed algorithm is inspired by the process of tumbleweeds growing and spreading seeds by rolling with the wind. However, the algorithm suffers from poor convergence accuracy and tends to fall into local optimality. Therefore, the Advanced Tumbleweed Algorithm (ATA) is proposed in this paper. The algorithm generates the initial population by Opposition-Based Learning (OBL) and chaotic mapping and introduces the tangent flight operator to improve the global search ability of individuals. Finally, it is evaluated by the CEC2018 function set and compared with several other algorithms. The experimental results show that the advanced tumbleweed algorithm has strong optimization search ability and high accuracy.