Reinforcement Learning for the Optimization of Decoupling Capacitors in Power Delivery Networks

This paper proposes an advantage actor-critic (A2C) reinforcement learning (RL)-based method for the optimization of decoupling capacitor (decap) design. Unlike the previous RL-based methods used for the selection of decap types or decap placements, the proposed method enables placement and the simultaneous selection of both decap types and their placements, thereby simplifying the design process. The results show that the proposed method can provide a larger number of optimized decap design solutions compared with previous methods, and can yield decap solutions even for multi-port optimization.