Real-Time Optimization in Complex Stochastic Environments
Abstract
The research in this project aims to enable a systematic on-line use of optimization techniques for real-time applications in complex, time-varying stochastic environments. This is in contrast to off-line optimization where computationally intensive methods may be used. The main outcomes of the project are: (a) event-driven distributed optimization algorithms which exploit the spatial decomposition of complex optimization problems and involve minimal communication, (b) receding horizon algorithms which exploit a temporal (rather than spatial) decomposition, (c) a novel "boosting function" approach which takes advantage of structural information to escape local optima and approach global optimality.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 14, 2018
- Accession Number
- AD1085605
Entities
People
- Christos G. Cassandras
Organizations
- Boston University