Demand response scheduling using derivative-based dynamic surrogate models

• The dynamic response of complex process plants can be synthesized by a single function. derivative-based approach provides model general validity. Optimal scheduling problem computational time reduced orders magnitude. Quantitative assessment downstream dynamics impact on demand scheduling. Computational effective surrogate models meet the need DR for daily update. When assessing to solve optimal problems, optimization algorithm needs coupled with in order quantify behavior monitored variable. For negligible transients, first approximation consists applying steady state correlation between input and output variables. On contrary, when show relevant bias respect estimated response, more accurate is required. coupling entire chemical processes algorithm, effort drastically increases. In those cases, should simplified without losing its accuracy. this research work, we propose modeling applied an ethylene oxide production problem. Thanks validity, once derived, it any setpoint trajectory. This allows reduce magnitude, providing results same accuracy as detailed simulation.