https://ir.unisa.ac.za/handle/10500/2921 Thu, 14 May 2026 08:53:48 GMT 2026-05-14T08:53:48Z https://ir.unisa.ac.za/handle/10500/31193 An image catalogue of micron- and sub-micron-sized constrictions fabricated with the femtosecond laser and created with the AFM Umenne, Patrice In this article a series of constrictions in the micron and sub-micron range fabricated with the femtosecond laser on superconducting YBa2Cu3O7 thin films is scanned with an atomic force microscope (AFM). The AFM images of the resulting constrictions are depicted. The width of the constrictions fabricated range from 1.24 um micron dimension to 812 nm in the sub-micron range. The laser ablation spot size or diameter is set at 15.8 lm for the femtosecond laser and the separation distance between the centers of the laser ablation spots is varied between (16–17) lm as a result generating the variable constriction widths for the different superconductive bridges. The images of the fabricated micron and sub-micron constrictions are shown as a catalogue in the article. The AFM images give the surface morphology, dimensions, and physical continuity of the fabricated constriction qubits. The I–V characteristics of the constrictions are measured and validate the electrical continuity and superconductivity of the fabricated constrictions. Mon, 12 Jun 2023 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/31193 2023-06-12T00:00:00Z https://ir.unisa.ac.za/handle/10500/31166 Natural language processing (NLP) for the decolonisation of low resourced african languages Sumbwanyambe, Mbuyu Natural Language Processing (NLP) has emerged as a critical component of artificial intelligence (AI) with the potential to support the decolonisation and revitalisation of low-resourced African languages. This article explores the relationship between human intelligence and computational systems, focusing on how bio-neural networks inspire artificial neural networks used in modern AI applications. In the human brain, intelligence arises from interconnected neurons, where repeated stimulation strengthens neural pathways through synaptic connections. This biological principle underpins artificial neural networks, which model learning through weighted inputs, processing layers, and outputs. The paper provides a foundational overview of biological neural networks, their artificial counterparts, and the role of NLP in enabling machines to understand, process, and generate human language. Particular attention is given to how AI technologies can address historical linguistic marginalisation by developing digital tools for African languages that lack sufficient datasets and technological support. The study further outlines current and emerging AI applications, emphasizing that the integration of NLP into education, healthcare, governance, and communication is increasingly unavoidable. Ultimately, the article argues that leveraging AI and NLP responsibly can play a transformative role in preserving linguistic diversity, promoting inclusion, and advancing technological equity in African contexts. Fri, 21 Jul 2023 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/31166 2023-07-21T00:00:00Z https://ir.unisa.ac.za/handle/10500/31148 Modelling the mine production chain for improved performance Mulenga, Francois Wed, 19 Jul 2023 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/31148 2023-07-19T00:00:00Z https://ir.unisa.ac.za/handle/10500/29728 The Impact of the Electric Double-Layer Capacitor (EDLC) in Reducing Stress and Improving Battery Lifespan in a Hybrid Energy Storage System (HESS) System Tshiani , Chrispin ; Umenne, Patrice This paper investigates the effect of the electric double layer capacitor (EDLC) in reducing stress and prolonging the battery lifespan in a hybrid energy storage system (HESS). A 65 F, 16.2 V EDLC supercapacitor was connected in a laboratory experiment to produce its charge/discharge profile at a constant current of 5 and 10 A. The EDLC’s Faranda or “two branch model” mathematical parameters were extracted from the experimental charge/discharge profile. The extracted parameters were used as inputs to design the Python/MATLAB/Simulink (PMS)-hybrid model of the EDLC. The charge/discharge profiles of the simulated PMS model of the EDLC were then compared to the charge/discharge profiles derived from the experimental setup of the EDLC and were found to match. The PMS model of the EDLC was then used as a subcomponent in an HESS system modelled in MATLAB/Simulink. Using constant load conditions, the battery’s voltage, current, power and state of charge (SOC) were analyzed for a battery energy storage system (BESS) without a supercapacitor and then compared to an HESS system with a supercapacitor in an experimental setup. This process was repeated with the simulated PMS model of the EDLC in MATLAB/Simulink for HESS and without the EDLC for BESS. Finally using a variable load in an experimental setup, the battery’s voltage and current were analyzed for a BESS system and compared to an HESS system. All these data show that, in an HESS system with a supercapacitor, there is less stress on the battery with a load applied. This is indicated by the voltage and current values in an HESS system being consistently more stable with respect to time as compared to the BESS system. As a result, in an HESS system, the battery will have a longer lifespan. Fri, 18 Nov 2022 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/29728 2022-11-18T00:00:00Z