The integration of automotive radar communication is the future development trend and will be installed on more and more vehicles, which will cause serious interference to radar detection and communication, seriously affecting the detection accuracy of radar and communication reliability. This article analyzes the interference models between radar and radar, radar and communication, and communication and communication. To improve the anti-interference ability of the network, the interference is transformed into a signal-to-noise ratio optimization model constrained by resources and power. Use reinforcement learning algorithms to find the optimal solution to minimize system interference. We use the vehicle position, as well as the signal-to-noise ratio of radar and communication, as the state and action space, and the total throughput increment serves as the reward function for Q-learning, and obtain the optimal joint optimization strategy through training. The evaluation results indicate that compared with existing schemes, the proposed algorithm can more effectively reduce radar and communication interference, thereby improving radar detection accuracy and communication throughput.