A Computational Framework for Lithium Ion Cell-Level Model Predictive Control Using a Physics-Based Reduced-Order Model
Autor: | Gregory L. Plett, M. Scott Trimboli, Aloisio Kawakita de Souza, Kiana Karami, Marcelo A. Xavier |
---|---|
Rok vydání: | 2021 |
Předmět: |
Battery (electricity)
Work (thermodynamics) Mathematical optimization Control and Optimization Computer science chemistry.chemical_element 02 engineering and technology Physics based 010402 general chemistry 021001 nanoscience & nanotechnology 01 natural sciences Extensibility 0104 chemical sciences Reduced order Model predictive control chemistry Control and Systems Engineering Lithium 0210 nano-technology Realization (systems) Voltage |
Zdroj: | ACC |
ISSN: | 2475-1456 |
DOI: | 10.1109/lcsys.2020.3038131 |
Popis: | Most state-of-the art battery-control strategies rely on voltage-based design limits to address performance and lifetime concerns. Such approaches are inherently conservative. However, by exploiting internal electrochemical quantities, it is possible to control battery performance right up to true physical bounds. This letter develops an extensible framework that combines model predictive control (MPC) with computationally efficient realization algorithm (xRA)-generated reduced-order electrochemical models for the advanced control of lithium-ion batteries. The approach is demonstrated on the fast-charge problem where hard constraints are imposed on problem variables to avoid lithium plating induced performance degradation. This letter establishes a general mathematical foundation for the incorporation of electrochemically rich reduced-order models directly into an MPC framework. |
Databáze: | OpenAIRE |
Externí odkaz: |