| Description |
-
Abstract: This thesis
study demonstrates one of the applications of Design of Dynamic Experiments (DoDE) and
Dynamic Response Surface Model (DRSM) methodologies on the development of hybrid model
based optimization. The optimization method seeks to improve existing process
performance by simply running experiments, collecting data, and proposing a new set of
experiments that optimize the ... read moreprocess. We compare the purely data-driven method, which
uses only DoDE, we call it "Classic Approach", with the newly suggested hybrid approach.
The Hybrid Approach uses our knowledge of the energy balance around the reactor to
predict adiabatic temperature rise. Hence, the hybrid approach is superior to the
classic approach in terms of the probability of avoiding process safety violations. It
was possible to reduce the process safety violations by approximately 75%. Moreover, the
final result is within 2.00% of the theoretical best batch time found using model-based
optimization.
Thesis (M.S.)--Tufts University,
2018.
Submitted to the Dept. of Chemical and
Biological Engineering.
Advisor: Christos
Georgakis.
Committee: Matthew Panzer, and Caleb
McGruder.
Keywords: Chemical engineering, Applied
mathematics, and Systems
science.read less
|
|---|
This object is in collection
