To address the behavioral conflicts between human-machine co-driving vehicles and human driving vehicles in mixed traffic scenarios, and to achieve the reasonable sharing of driving authorities and the effective coupling of driving behaviors in the human-machine co-driving mode, a double-layer game-based driving behavior decision-making model based on human-machine interaction and vehicle-to-vehicle interaction was proposed. At the human-machine interaction level, a fuzzy game strategy was adopted, and a risk potential field model and a vehicle stability model were established, which were used as the input of the driving authority allocation system; using fuzzy logic methods and fuzzy benefit analysis, the internal logical relationship between the input and the human-machine driving authority allocation was analyzed, and finally, the Nash equilibrium was solved to determine the human-machine driving authority allocation scheme. At the vehicle-to-vehicle interaction level, a perfect information dynamic game strategy was used to construct the human-machine co-driving vehicle behavior decision-making model. For the potential conflict patterns existing in unsignalized intersection, the driver and the autonomous driving system"s preference for benefits were coupled through human-machine driving authority allocation, and the optimal driving strategy was solved. To verify the effectiveness of the proposed model, a simulation environment for unsignalized intersection was built on the Matlab platform, and numerical simulation verification was carried out. The simulation results show that, compared with the intelligent driver model, the proposed double-layer game model could effectively reduce the collision rate in three driving style scenarios, especially in the aggressive driving style scenario, it could reduce the collision conflicts by 47.02%; in the three driving style scenarios, the average traffic speed of the proposed model is higher than that of the intelligent driver model, and still has a high traffic efficiency in the conservative driving style scenario.