Next-Generation Reservoir Computing for Dynamical Inference
This provides a flexible framework for dynamical inference in fields like physics and engineering, though it is incremental as it builds on existing next-generation reservoir computing variants.
The authors tackled the problem of modeling dynamical systems from partial and noisy time series data by developing a simple, scalable next-generation reservoir computing method that uses pseudorandom nonlinear projections instead of polynomial-based ones. The result was stable models that generalize beyond training data, enabling precise control of system states for applications like surrogate modeling and digital twins.
We present a simple and scalable implementation of next-generation reservoir computing for modeling dynamical systems from time series data. Our approach uses a pseudorandom nonlinear projection of time-delay embedded input, allowing an arbitrary dimension of the feature space, thus providing a flexible alternative to the polynomial-based projections used in previous next-generation reservoir computing variants. We apply the method to benchmark tasks -- including attractor reconstruction and bifurcation diagram estimation -- using only partial and noisy observations. We also include an exploratory example of estimating asymptotic oscillation phases. The models remain stable over long rollouts and generalize beyond training data. This framework enables the precise control of system state and is well suited for surrogate modeling and digital twin applications.