scholarly journals Understanding chemical production processes by using PLS path model parameters as soft sensors

2020 ◽  
Vol 139 ◽  
pp. 106841 ◽  
Author(s):  
Geert H. van Kollenburg ◽  
Jacoline van Es ◽  
Jan Gerretzen ◽  
Heleen Lanters ◽  
Roel Bouman ◽  
...  
atp magazin ◽  
2016 ◽  
Vol 58 ◽  
pp. 64 ◽  
Author(s):  
Marc Kalliski ◽  
Benedikt Beisheim ◽  
Daniel Krahè ◽  
Udo Enste ◽  
Stefan Krämer ◽  
...  

Operational decisions in the day-to-day business of chemical production processes can have a significant impact on the energy and material efficiency. We propose to use real-time resource efficiency indicators (REI) to accurately monitor the energy and material efficiency in real-time and subsequently to use these in decision support for the operating staff. To guide industrial users during the development of REIs, a Namur ad-hoc working group (AK) Resource Efficiency Indicators for the operational use was set up to prepare a Namur Recommendation.


2012 ◽  
Vol 49 (11) ◽  
pp. 1309-1318 ◽  
Author(s):  
Katherine S. Johnston ◽  
Bruce Jamieson ◽  
Alan Jones

Extreme snow avalanche runout is typically estimated using a combination of historical and vegetation records as well as statistical and dynamic models. The two classes of statistical models (α–β and runout ratio) are based on estimating runout distance past the β-point, which is typically defined as the point where the avalanche slope incline first decreases to 10°. The parameters for these models vary from mountain range to mountain range. In Canada, α–β and runout ratio parameters have been published for the combined Rocky and Purcell Mountains and for the British Columbia Coast Mountains. Despite active development, no suitable tall avalanche path model parameters have been published for the Columbia Mountains or for the Lizard Range area around Fernie, B.C. Using a dataset of 65 avalanche paths, statistical model parameters have been derived for these regions.


2006 ◽  
Vol 53 (8) ◽  
pp. 129-137 ◽  
Author(s):  
B.R.H. Peiris ◽  
P.G. Rathnasiri ◽  
J.E. Johansen ◽  
A. Kuhn ◽  
R. Bakke

Hydrogen can be produced by fermentation of organic wastes as a renewable CO2 emission free fuel. The production potential as a function of feed composition is investigated using the ADM1 and experimental data from the literature. Lactate and ethanol are included in the model as intermediates to simulate the bio-hydrogen production processes more closely. Simulated effects of carbohydrate to protein ratio in the feed on pH, H2, biomass and fatty acid production using standard model parameters compare quite well with experimental results. The overall hydrogen and biomass production corresponds well with measurements for some feeds and less for others. The maximum theoretical yield is significantly higher than the simulated and measured values and is highest when the feed consists of only carbohydrates. The analysis suggests that the modified ADM1 is capable of simulating the main mechanisms involved in biological hydrogen production processes, implying that the model can be used to identify, and find strategies to influence limiting factors in bio-hydrogen production processes. Model weaknesses regarding the acidogenesis processes are observed and areas for further improvements discussed.


1991 ◽  
Vol 02 (01) ◽  
pp. 276-283 ◽  
Author(s):  
GÖTZ H.R. BOKELMANN ◽  
PAUL G. SILVER

A scheme for extracting multiple phase body wave traveltimes is presented which is most applicable to teleseismic broadband and short period data from permanent and portable instruments. The method specifically allows for the nonlinear dependence of waveforms on travel time by performing a nonlinear search over a reduced set of “projected model parameters”. This reduced set is found by splitting the method into two parts, the nonlinear dependence of waveforms on traveltime perturbations (particularly strong for broadband and short period data) and the linearizable dependence of travel time perturbations on variations in seismic velocity. By use of generalized inverse, the traveltime perturbations can be adequately characterized by a reduced number of linear combinations of model parameters. Consequently, the nonlinear search is performed over an optimally compact model space. The technique can consider simultaneously a large number of body wave phases creating a systematic methodology for extracting large numbers of traveltimes from single source-receiver pairs. The resulting path model may or may not be of physical significance; the primary goal is the extraction of travel times that may be subsequently used for a more comprehensive travel time inversion.


2014 ◽  
Vol 49 (9) ◽  
pp. 1289-1297 ◽  
Author(s):  
Fan Zhang ◽  
Mary E. Rezac ◽  
Sudipto Majumdar ◽  
Praveen Kosaraju ◽  
Stuart Nemser

Sign in / Sign up

Export Citation Format

Share Document