AutoREC: A Reinforcement Learning Platform for Automated Equivalent Circuit Model Generation
ORIGINAL / AutoREC: A reinforcement learning platform for equivalent circuit model generation
This study introduces AutoREC, an open-source Python platform that uses reinforcement learning (RL) to automatically generate equivalent circuit models (ECMs) from electrochemical impedance spectroscopy (EIS) data, replacing manual trial-and-error, and suitable for autonomous experimental pipelines such as self-driving laboratories.
01 ABSTRACT
This paper presents the architecture and features of the AutoREC platform, and demonstrates a DDQN-based agent trained on synthetic data and applied to experimental spectra from battery, corrosion, oxygen evolution reaction, and CO2 reduction systems. The authors believe AutoREC provides an extensible foundation, but the case studies also reveal challenges related to experimental complexity and limited training-data coverage.
02 KEY FINDINGS
- AutoREC formulates ECM generation as a Markov decision process, where an RL agent sequentially modifies circuit topology based on current state and feedback.
- The platform supports an end-to-end workflow including EIS preprocessing, agent setup and training, ECM generation for new measurements, and visualization-based evaluation.
- It implements a configurable Double Deep Q-Network (DDQN) agent with prioritized experience replay and a dead-loop mitigation strategy.
- A representative agent trained on synthetic EIS datasets was successfully applied to experimental spectra from battery, corrosion, oxygen evolution reaction, and CO2 reduction systems.
- The case studies demonstrate the platform's capabilities while revealing challenges such as experimental complexity and limited training-data coverage.
AI GENERATED SUMMARY / DISCOVERED BY ARXIV CS.LG