Opinion dynamics utilize mathematical and physical models and agent-based computational modeling tools to describe the interactive features and rules among individuals, reduction and explain the opinion formation and evolution in the real life. Based on the analysis of the basic ideas and methods of the modeling of opinion dynamics, the main contents, characteristics and achievements of classical model and extended model of opinion dynamics are reviewed, and discussed the current research status. Finally, the main problems and challenges of opinion dynamics are analyzed, correspondingly, shed light on its future directions.