Reinforcement Learning-Based Placement Method for Printed Circuit Board

In the modern electronics industry, the printed circuit board (PCB) serves as a crucial carrier for electronic system, and its placement significantly influences the performance and stability of the entire system. Simulated annealing (SA) is an effective algorithm used in PCB automatic placement. Reinforcement learning (RL) can learn optimal strategies through interactions with the environment. In this work, we propose an RL-assisted PCB placement method based on simulated annealing. RL trains an agent to identify effective placements for SA when it produces stable solutions. Experimental results demonstrate that our method can reduce the running time by over 50% while maintaining the same quality of layout results, thus significantly enhancing the efficiency of the original algorithm.