To address single route planning strategies and limited applicability in the decision-making process of autonomous vehicles, this study proposes a route planning strategy that accounts for both route safety and efficiency. Initially, leveraging risk analysis theory and integrating road data from digital maps with historical accident data from data centers, the study formulates a road safety assessment model to evaluate the safety of road sections. Subsequently, a vehicle dynamics model is established for diverse road alignments to determine the safe speeds on various road sections, and an estimation model of time taken to pass a road section is developed to assess traffic efficiency. Finally, utilizing the A* algorithm and integrating the road safety assessment model with the time estimation model, this paper devises a comprehensive cost function. Simulation experiments are conducted in SUMO to validate the proposed strategy. The results reveal that although the route planning time increases by 11.1%, the route risk simultaneously decreases by 56.6% in the case of using the proposed strategy compared with the conventional strategy. This indicates that the strategy investigated in this paper exhibits more comprehensive performance and improved applicability.