chenhg99/NODEs_digital_twin: Memristive NODE

Digital twins, pivotal to Industry 4.0, replicate real-world entities through computer models, revolutionising fields like manufacturing management and industrial automation. While machine learning offers data-driven methods for developing digital twins, its reliance on discrete-time data and finite-depth models fails to capture continuous dynamics and complex system behaviors. Furthermore, the von Neumann architecture's separation of storage and processing leads to high time and energy costs due to frequent data transfers and analogue-to-digital conversions. Here, we present a memristive neural ordinary differential equation (ODE) solver for digital twins that models continuous-time dynamics and complex systems using infinite-depth models. By integrating storage and computation within analogue memristor arrays, we overcome the von Neumann bottleneck, achieving significant improvements in speed and energy efficiency.