COMBINING NEURAL NETWORKS AND FIRST PRINCIPLE MODELS FOR BIOPROCESS MODELING

2001 ◽  
pp. 121-148 ◽  
Author(s):  
B. EIKENS ◽  
M. N. KARIM ◽  
L. SIMON
Keyword(s):  
Neural Networks ◽  
First Principle ◽  
Bioprocess Modeling

scholarly journals CO organization at ambient pressure on stepped Pt surfaces: First principle modeling accelerated by neural networks

2021 ◽  
Author(s):  
Philippe Sautet ◽  
Vaidish Sumaria
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Ambient Pressure ◽  
High Dimensional ◽  
First Principle ◽  
Crucial Importance ◽  
Adsorption Structure ◽  
First Principle Calculations

Step and kink sites at Pt surfaces have a crucial importance in catalysis. We employ high dimensional neural network potential (HDNNP) trained using first-principle calculations to determine the adsorption structure...

Cyclic models and recurrent neural networks

2019 ◽  
pp. 183-205
Author(s):  
Thomas P. Trappenberg
Keyword(s):  
Neural Networks ◽  
Dynamical Systems ◽  
Bayesian Networks ◽  
Recurrent Neural Networks ◽  
Directed Graphs ◽  
First Principle ◽  
Temporal Modeling ◽  
Cyclic Graphs ◽  
Cyclic Models

This chapter discusses models with cyclic dependencies. There are two principle architectures that are discussed. The first principle architecture of cyclic graphs comprises directed graphs similar to the Bayesian networks except that they include loops. Formally, such networks represent dynamical systems in the wider context and therefore represent some form of temporal modeling. The second type of models have connections between neurons that are bi-directional. These types of networks will be discussed in the context of stochastic units in the second half of this chapter.

Hybrid Neural Network-First Principles Modeling of Critical Heat Flux in a Thin Annular Channel

2002 ◽  
Author(s):  
T. Xie ◽  
S. M. Ghiaasiaan ◽  
S. Karrila ◽  
T. McDonough
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Heat Flux ◽  
Critical Heat Flux ◽  
First Principle ◽  
Flow State ◽  
State Variables ◽  
Two Phase ◽  
Slip Ratio ◽  
Artificial Neural

The hybrid artificial neural network-first principle modeling (ANN-FPM) is a powerful and flexible methodology that can be particularly useful for complex multi-phase flow processes. In this method the flow state variables are obtained from the solution of conservative equations using first principle-based closure relations whenever possible, and using trained artificial neural networks for poorly-understood rate processes. The ANN-FPM methodology is applied to a set of experimental data representing critical heat flux in a uniformly heated horizontal annular test section cooled with water. The first principle method is based on numerical solution of one-dimensional two-phase equilibrium conservation equations with velocity slip represented by an algebraic slip ratio correlation. The critical heat flux is predicted using trained neural networks that use local hydrodynamic parameters estimated by the first principle method. It is shown that the methodology works well for the studied data.

Neural Networks and First Principle Models for Bioprocesses

1999 ◽  
Vol 32 (2) ◽  
pp. 6974-6979 ◽  
Author(s):  
Bernd Eikens ◽  
M. Nazmul Karim ◽  
Laurent Simon
Keyword(s):  
Neural Networks ◽  
First Principle
Sign in / Sign up

Export Citation Format

Share Document