HybridNet: Integrating Model-based and Data-driven Learning to Predict Evolution of Dynamical Systems

TL;DR

HybridNet combines deep learning and model-based methods to accurately predict the evolution of complex dynamical systems with limited prior knowledge, adapting in real-time for improved accuracy.

Contribution

It introduces a hybrid framework integrating ConvLSTM and CeNN for real-time, adaptive prediction of dynamical systems with uncertain parameters.

Findings

Outperforms state-of-the-art deep learning methods in accuracy

Successfully models heat convection-diffusion and fluid dynamics systems

Learns physical parameters in real-time for adaptive predictions

Abstract

The robotic systems continuously interact with complex dynamical systems in the physical world. Reliable predictions of spatiotemporal evolution of these dynamical systems, with limited knowledge of system dynamics, are crucial for autonomous operation. In this paper, we present HybridNet, a framework that integrates data-driven deep learning and model-driven computation to reliably predict spatiotemporal evolution of a dynamical systems even with in-exact knowledge of their parameters. A data-driven deep neural network (DNN) with Convolutional LSTM (ConvLSTM) as the backbone is employed to predict the time-varying evolution of the external forces/perturbations. On the other hand, the model-driven computation is performed using Cellular Neural Network (CeNN), a neuro-inspired algorithm to model dynamical systems defined by coupled partial differential equations (PDEs). CeNN converts the intricate numerical computation into a series of convolution operations, enabling a trainable PDE solver. With a feedback control loop, HybridNet can learn the physical parameters governing the system's dynamics in real-time, and accordingly adapt the computation models to enhance prediction accuracy for time-evolving dynamical systems. The experimental results on two dynamical systems, namely, heat convection-diffusion system, and fluid dynamical system, demonstrate that the HybridNet produces higher accuracy than the state-of-the-art deep learning based approach.