Multi-Objective Physics-Guided Recurrent Neural Networks for Identifying Non-Autonomous Dynamical Systems.
Oliver SchönRicarda-Samantha GötteJulia TimmermannPublished in: CoRR (2022)
Keyphrases
- dynamical systems
- recurrent neural networks
- multi objective
- evolutionary algorithm
- differential equations
- dynamic systems
- objective function
- neural network
- feed forward
- echo state networks
- reservoir computing
- nonlinear dynamical systems
- particle swarm optimization
- genetic algorithm
- control theory
- artificial neural networks
- linear systems
- neural model
- recurrent networks
- state space
- artificial intelligence
- qualitative simulation
- phase space
- agent environment
- nonlinear dynamic systems
- cascade correlation
- dynamical behavior
- dynamic programming
- partially observable markov decision processes
- control law
- predictive state representations
- fuzzy logic