Ayeni

Ayeni

Machine Learning enthusiast

Posts by Collection

portfolio

publications

Dynamic Modeling and Simulation of Fluid Catalytic Cracking Unit Riser

Published in Ife Journal of Technology, 2019

This paper deep dived into the dynamic effect of coke deposition on FCCU catalyst on the dynamics of the cracking process using mathematical model to realize the dynamics of the system and numerically simulating the model using java programming language

Recommended citation: Osunleke, A., Ayeni, O., Bamimore, A., & Alli, K. (2019). "Dynamic Modeling and Simulation of Fluid Catalytic Cracking Unit Riser." Ife Journal of Technology. vol(26)1,23-28. http://ijt.oauife.edu.ng/index.php/ijt/article/view/143

Data Driven Control Engineering Model of Calcium Regulation

Published in Jupyter Notebook, 2021

This work deep dived into the homeostasis of calcium in the body and attempt to model this with data gotten from the numeric simulation of mathematical model representing the system. A system of Diffrential Algebraic eqaution were sloved to generate data used to train a Long Short Term Memory(LSTM) neural network.

talks

Evaluating mass attenuation and mass energy-absorbtion coefficients of body tissues/ organ using computer software.

Published:

This study formulated an appropriate fitting model; carried out segmented least-square multifit on the attenuation data; derived optimum parameters of the fitting model for each of the energy segment and wrote a Java code. This was with a view to produce a computer program that will evaluate the mass attenuation and mass energy-absorption coefficients of body tissues/organs or any phantom material with excellent accuracy. Linear least-square curve-fitting method was employed to fit the existing attenuation data to a justified, fifth-order, inverse polynomial model by segmented multifit technique. The derived best-fit parameters, the polynomial model and the mixture additive rules were incorporated into Java codes which were packaged into a computer software application called UCOM. The results showed that the uncertainty of UCOM is 0.01% on the average. Comparative results also showed that UCOM gives better speed optimization (~150 milliseconds) and less memory utilization (~1 megabyte). Further comparisons show that UCOM offers more compatibility with computer Operating Systems such as WINDOW, UNIX and LINUX. UCOM gives some degree of freedom to user in the areas of energy grid input and input of number of elements composed in a tissue or substitute. Additionally, the software is user-friendly with a robust educative package.
This study concluded that the polynomial fitting and model used in this software provided a better and easier parameterization scheme for attenuation data than the arduous log-log and linear cubic spline fitting methods adopted in other similar parameterization studies.

teaching

System Modelling (CHE 411)

, Obafemi Awolowo University, Department of Chemical Engineering, 2020

Taught undergraduate tutorial on mathematical modeling of simple processes such as tank flow system and CSTR using first principle material, component and energy balances.

Control and Numerical Simulation of Simple Process

, Obafemi Awolowo University, Process Control and Simulation Lab, 2020

I Guided undergraduate through the act of simulation and control of simple chemical processes using tools such as Matlab(simulink), Java. I also implemented simple Model predictive control scheme using mathematical models of a two tank system. I also assisted in the supervision and grading of undergraduate lab work involving these