Reproduce “An Integrated Architecture for Data-Efficient Real-Time Neural Imitation of Nonlinear Time-Optimal Model Predictive Control” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Maarten De Baecke, Flavia Sofia Acerbo, Andrea Giusti, Wilm Decré

The authors introduce SIFS, a method for training neural model predictive controllers for time-optimal control. It uses a network’s predicted time-to-go, or time remaining, to flag rollout deviations and guide expert labeling. In a Furuta-pendulum simulation, SIFS reached comparable failure-rate performance about three times faster with less than half the expert data of DAgger and MC-dropout in that example.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.