Research Article
Modelling and Simulation of Flow and Level Dynamics in a Pharmaceutical Batch Mixer
Augustine Kweku Mbeah*
,
Erwin Normanyo
Issue:
Volume 14, Issue 3, September 2026
Pages:
50-58
Received:
25 August 2026
Accepted:
4 September 2026
Published:
20 September 2026
Abstract: A mathematical model and dynamic performance study of a Supervisory Control and Data Acquisition (SCADA) system and a Programmable Logic Controller (PLC) that were previously utilized to automate the pharmaceutical batch mixer are presented in this research. While the earlier work implemented the discrete, sequential control logic for the mixer using Siemens STEP 7 and validated it with RSLogix 5000, the continuous-time hydraulic behaviour of the three interconnected process vessels (heating, mixing and storage tanks) had not been characterised. In this study, the batch mixing system is partitioned into three sections, each represented by a pump-and-tank arrangement, and first-principles mass-balance, Bernoulli energy-balance and Darcy friction-loss equations are used to derive the transfer functions relating pump flow rate and tank level to valve resistance and tank cross-sectional area. Laplace transformation of the governing equations yields an overall block-diagram model of the three-tank system, which is implemented and simulated in MATLAB/SIMULINK. The simulation results show that the flow response settles at 0.5 m3/s with a settling time of 1.32 x 10^4 seconds, an overshoot of 11.5% and a rise time of 4.03 x 10^3 seconds. The tank-level responses show a successive filling pattern consistent with the batching sequence, with Tank 1 filling within 20 minutes, Tank 2 reaching its maximum level at 65 minutes, and Tank 3 completing filling at 210 minutes from the start of the process. These results confirm that the batch mixer exhibits stable, well-damped dynamic behaviour, providing a quantitative basis for the timer presets and control logic used in the PLC-SCADA automation reported previously, and supporting the suitability of the process for closed-loop SCADA-based control.
Abstract: A mathematical model and dynamic performance study of a Supervisory Control and Data Acquisition (SCADA) system and a Programmable Logic Controller (PLC) that were previously utilized to automate the pharmaceutical batch mixer are presented in this research. While the earlier work implemented the discrete, sequential control logic for the mixer usi...
Show More
Research Article
An Uncertainty-Calibrated Ensemble Framework for Fault Diagnosis in Industrial Hydraulic Power Units: Rethinking Stacked Neural Architectures
Emmanuel Unekuojo Enejor*,
Nelson Oluwole Ibigbami,
Ismail Olusegun Lawal,
Temidayo Adeoti,
Ademola Adesoye,
Samson Oluwasina Ajibesin,
Adekunle Francis Adeniyi,
Kolawole Bamidele Jayeola,
Oludare Michael Shotade
Issue:
Volume 14, Issue 3, September 2026
Pages:
59-72
Received:
29 August 2026
Accepted:
10 September 2026
Published:
20 October 2026
DOI:
10.11648/j.acis.20261403.12
Downloads:
Views:
Abstract: Stacked artificial neural network architectures have been proposed for industrial condition monitoring on the premise that adding sequential learning stages improves predictive accuracy, yet this premise is rarely tested against simpler alternatives under leakage-free, cross-validated evaluation. This study conducts a rigorous, block-aware empirical investigation of predictive fault diagnosis on the benchmark UCI Condition Monitoring of Hydraulic Systems dataset (2,205 operating cycles, 17 multi-rate sensors, four superimposed component-degradation labels). We first demonstrate that naive random cross-validation on this dataset inflates reported accuracy by up to 26 percentage points relative to a block-aware (leakage-free) protocol, because the dataset is organised as long contiguous condition blocks. Under the corrected protocol, we reproduce and stress-test the stacking mechanism used by a representative elaborative stepwise stacked neural network (ESSANN) architecture from the recent literature, implementing it faithfully as a deep-ensemble multilayer perceptron (MLP) with feature-concatenation stacking and a boosting-style hard-example variant, and evaluate all architectures under 4-fold block-aware cross-validation with paired per-fold robustness checks. Both neural stacking variants degrade accuracy relative to a single-stage MLP ensemble consistently across every fold on both primary targets (valve condition, accumulator pressure), revealing that naive depth is not a principled correction mechanism and is vulnerable to training-data starvation. In contrast, a two-stage Random Forest architecture and Gradient Boosting, which stage learners on the complete training set every round, improve on the single-stage baseline consistently across all folds on valve condition (Gradient Boosting: 98.1% ± 1.1%, +11.3 points over Random Forest in every fold), but the same improvement is not sign-consistent across folds for accumulator pressure, where no tested architecture reliably outperforms a plain Random Forest baseline once fold-to-fold variance is accounted for. We further show that predictive entropy from a deep ensemble is a well-calibrated, monotonic indicator of classification risk (100% accuracy in the lowest-entropy quintile versus 41-63% in the highest), but that thresholds calibrated on training-set entropy are unusable in deployment, over-flagging 94-98% of unseen samples as "uncertain"; calibrating instead on a held-out validation block reduces this to a deployable 44-60%. We conclude with a practical, evidence-graded architecture recommendation and a risk-stratified decision framework for confidence-gated industrial control actions.
Abstract: Stacked artificial neural network architectures have been proposed for industrial condition monitoring on the premise that adding sequential learning stages improves predictive accuracy, yet this premise is rarely tested against simpler alternatives under leakage-free, cross-validated evaluation. This study conducts a rigorous, block-aware empirica...
Show More
