All news
Digital Twin

Parallel branch MPC on GPUs

US & West ·

A GPU-based solver for trajectory planning problems using branch model predictive control is presented. The solver is based on iterative LQR, multiple-shooting formulation, and augmented Lagrangian method for constraint handling. Numerical experiments show superiority over a CPU-based solver on large-scale problems.

Why it matters

Accelerating MPC computations on GPU enables real-time planning for large systems, which is critical for autonomous vehicles and robotics.

Relevant to DT products

Original headline
Parallel Branch Model Predictive Control on GPUs
Read the original: arXiv — Systems and Control (eess.SY)

Translation and summary are machine-generated from the source. The full article is not republished; its rights belong to the source.

Related

ModellingUS & West

Adaptive Nonlinear MPC without Prior Training

A nonlinear adaptive predictive control method (NPCAC) is proposed that identifies a pseudo-linear model online from input-output data without prior training. It combines recursive least squares with information forgetting and iterative MPC, tested with polynomial, Fourier, and spline basis functions. The approach handles systems with high uncertainty.

arXiv — Systems and Control (eess.SY) ↗
Food & agriUS & West

Deep Learning for Fast Climate Risk Assessment of Crops

A deep-learning framework SECSF is presented that emulates the process-based ECroPS model for maize and barley using only daily temperatures and precipitation. Trained on ERA5 data, it reproduces crop growth dynamics and harvest timing while reducing computational cost by ~10^4 times, enabling probabilistic risk assessment for large seasonal and climate ensembles.

DTPDT.ForecastarXiv — Computational Engineering, Finance, and Science (cs.CE) ↗