Efficient time stepping for numerical integration using reinforcement learning

Abstract

Many problems in science and engineering require an efficient numerical approximation of integrals or solutions to differential equations. For systems with rapidly changing dynamics, an equidistant discretization is often inadvisable as it either results in prohibitively large errors or computational effort. To this end, adaptive schemes, such as solvers based on Runge--Kutta pairs, have been developed which adapt the step size based on local error estimations at each step. While the classical schemes apply very generally and are highly efficient on regular systems, they can behave sub-optimal when an inefficient step rejection mechanism is triggered by structurally complex systems such as chaotic systems. To overcome these issues, we propose a method to tailor numerical schemes to the problem class at hand. This is achieved by combining simple, classical quadrature rules or ODE solvers with data-driven time-stepping controllers. Compared with learning solution operators to ODEs directly, it generalises better to unseen initial data as our approach employs classical numerical schemes as base methods. At the same time it can make use of identified structures of a problem class and, therefore, outperforms state-of-the-art adaptive schemes. Several examples demonstrate superior efficiency. Source code is available at https://github.com/lueckem/quadrature-ML.

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Scikit-learn: Machine Learning in Python
• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Continuous control with deep reinforcement learning
• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• DGM: A deep learning algorithm for solving partial differential equations
• Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
• Data-driven approximation of the Koopman generator: Model reduction, system identification, and control
• Solving differential equations with neural networks: Applications to the calculation of cosmological phase transitions
• A Theoretical Analysis of Deep Q-Learning
• On the Treatment of Optimization Problems with L1 Penalty Terms via Multiobjective Continuation

Cited by in corpus (2)

• Reinforcement learning-based adaptive time-integration for nonsmooth dynamics
• Error estimation and step size control with minimal subsystem interfaces