Traffic sign recognition can quickly extract and identify road perception information, making it one of the key technologies in the field of autonomous driving. The reliable recognition of traffic signs is the primary task in achieving safe autonomous driving. However, current machine learning-based traffic sign recognition technologies still suffer from low interpretability and an excessive dependence on the sample space. Therefore, this paper proposes a Data and Knowledge Dual-Driven Traffic Sign Recognition Algorithm that leverages the convenience of data-driven methods and the reliability of knowledge-driven approaches. The algorithm integrates data-driven methods such as color recognition, shape recognition, and convolutional neural networks, along with knowledge-driven methods that involve reasoning based on traffic sign knowledge base, to accomplish the task of traffic sign recognition. The experimental results demonstrate that compared to other methods, the dual-drive approach proposed in this paper can identify traffic signs with greater accuracy and reliability.