acm-header
Sign In

Communications of the ACM

ACM TechNews

Controlling Complex Systems with AI


The artificial intelligence system AI Pontryagin automatically learns quasi-​optimal control signals for complex dynamic systems.

Credit: Shutterstock

Scientists at the Swiss Federal Institute of Technology, Zurich (ETH Zurich) and Germany's Frankfurt School have developed AI Pontryagin, an artificial intelligence (AI) system that can solve complex control problems.

AI Pontryagin combines numerical and analytical approaches to automatically learn quasi-optimal control signals for complex dynamic systems, even when not previously informed of ideal solutions.

Users must first provide the AI with data on the target system's dynamics, in order to ascertain which areas need dynamic optimization.

AI Pontryagin also must receive data on the target system's initial status and its desired status.

Potential areas of application include smart grids, traffic control systems, and supply-chain and production-process optimization.

From ETH Zurich (Switzerland)
View Full Article

 

Abstracts Copyright © 2022 SmithBucklin, Washington, DC, USA


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account