Adaptive Horizon Model Predictive Control and Regulation, Short Horizon Estimation
Abstract
In the past year, we have accomplished two of the goals of this project. The first goal was to develop and refine the Adaptive Horizon Model Predictive Control (AHMPC) methodology so that it can be used to stabilize fast processes. Recall AHMPC is a way to verify in real time that the Model Predictive Control (MPC) methodology is actually stabilizing the plant. It does this by adapting the MPC horizon length in real time keeping it as short as possible consistent with stabilization and feasibility. AHMPC does a simple check at each iteration to see if the current horizon is long enough. If the check is true, then the horizon is kept constant or shortened. If the check is false, then the horizon is lengthened.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 10, 2022
- Accession Number
- AD1230385
Entities
People
- Arthur J. Krener
Organizations
- University of California