Abstract: In order to adapt to the complex and changeable battlefield environment, the combination of modern radar technology and artificial intelligence is often used to improve the anti-jamming decision-making efficiency of intelligent radar. At present, the anti-interference decision-making algorithm based on reinforcement learning has not carried out in-depth research on the path selection process in the combined anti-interference matrix. Based on this, the paper proposes a method of intelligent pathfinding in the anti-jamming matrix combining the radar system and anti-jamming style when the radar is interfered with. First, Firstly, the anti-interference matrix model is established through the performance evaluation analysis. Secondly, the internal path selection process of anti-interference matrix is analyzed based on the reinforcement learning principle. Finally, the experimental process is simulated, and the influence of Q-learning algorithm and Dyna-Q algorithm on the pathfinding effect is compared, which shows that the proposed algorithm can complete the path selection in less time for interaction training with external interference, and find the best anti-interference method in the combined matrix, which can quickly and effectively deal with the external complex interference environment.