Efficient ESD Generator Modeling Using Reinforcement Learning

This paper describes Reinforcement Learning (RL) based technique for estimating the parameters of ESD generators modeled using circuit templates, given the current discharge waveform. The proposed algorithms can be applied to any circuit template to tune the circuit parameters for the desired application or optimization efficiently with minimum prior knowledge. The model can then be used in system-level dynamic ESD simulation with 3D Electromagnetic simulation for optimal ESD protection design.