Electronic Theses and Dissertations (Masters)

Permanent URI for this collectionhttps://hdl.handle.net/10539/37975

Browse

Filters

2023 - 20242

Settings

Has files: YesSubject: search.filters.subject.SDG-9: Industry, innovation and infrastructure

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Improving grade estimation using machine learning: a comparative study of ordinary kriging against machine learning algorithms
    (University of the Witwatersrand, Johannesburg, 2024) Akpabio, Aniekan
    This study investigated the efficiency of machine learning (ML) methods in the accurate prediction of ore grades, placing them in direct comparison with the established Ordinary Kriging (OK) methodology, a mainstay in geostatistical analysis. Utilising a dataset from a complex platinum group elements (PGE) deposit, the research assessed a suite of ML algorithms—namely, Random Forest (RF), Decision Trees (DT), Support Vector Regression (SVR), and particularly 𝑘- Nearest Neighbours (𝑘NN). The latter is highlighted for its adeptness in assimilating spatial data correlations intrinsically, echoing the insights from Nwalia's analytical explorations. The research engages with detailed swath plot analyses, comparative metric evaluations, and a nuanced understanding of spatial continuity, to illustrate the distinct advantages and operational competencies of the models. 𝑘NN, with its reliance on local data proximities and non-parametric nature, alongside RF, with its ensemble-based approach, emerged as capable in point estimate predictions. These models adeptly delineated local grade variations, demonstrating a high degree of reliability to the observed data and outperforming the OK model in both precision and accuracy. Further, the study examined block estimate predictions, a cornerstone in practical mining and resource estimation, where both 𝑘NN and RF demonstrated a commendable ability to generalise predictions over larger spatial extents. This translates into significant potential for enhancing mineral resource estimation processes, tailoring them to the granular specifics of a given ore body, and refining block model accuracy to inform more strategic mining operations. While the results endorse the ML methodologies as robust alternatives to traditional geostatistical techniques, the research also highlights the nuanced nature of these predictions. Factors such as the ore body's heterogeneity, the appropriateness of the variogram model, and the interplay between prediction scale and algorithmic performance are examined, offering a critical lens through which the suitability of each method is assessed. iv The research suggests that while some models like LR and SVR are bounded by linear assumptions and hyperparameter sensitivities, non-linear models such as DT and RF can innately navigate the complex, multifaceted layers of geological data. The comprehensive evaluation extends to propose a novel set of performance metrics designed to capture the intricacies of grade prediction, thereby aligning closely with the operational demands and decision-making processes in the mining industry.
  • Thumbnail Image
    Item
    Support Design Approach for Crusher Chambers: A Case Study of Palabora Mining Company
    (University of the Witwatersrand, Johannesburg, 2023-01) Masole, Nyeleti Venus; Stacey, T.R.
    This report project aimed to design a support system for crusher chambers at Palabora. The research project focused mainly on the two crusher chambers (12m wide by 25m high and 61m long) planned for the Lift 2 project as part of the ore handling system. The main research questions that the researcher sought to answer were; what are the differences between Lift 1 & Lift 2 in rock mass characterisation, classification and the ground control district?; how suitable is the Lift 1 crusher chamber support system for Lift 2?; what could be support requirements for Lift 2 crusher chambers in terms of empirical, analytical and numerical design methods and what are the recommended support design approaches for Lift 2 crusher chambers? The methodology used to design support for the Lift 2 crusher chambers was based on determining the expected failure mode first and then selecting suitable design methods to cater for the extent of failure. This study combined empirical and analytical methods to determine the failure mode and required support system. The results were then validated using Finite Element Method numerical modelling software called RS2 (Phase 2) from RocScience. Research findings revealed that the ground control district, classification and characterisation of rock masses differ slightly between Lift 1 and Lift 2. Jointing in dolerite dykes (DOL) was slightly dense in Lift 2 compared to Lift 1 and was associated with increased mining depth. Furthermore, the Lift 1 crusher chamber support system was found to be suitable for Lift 2 but must incorporate dynamic support. Unwedge (RocScience) analysis simulated wedge type of failure in the crusher chamber walls. The empirical and analytical design approach proposed cable bolt lengths of between 6m and 9 m and 3-4 m for roof bolts with bolt spacing of 1.4 m and 1.0 m respectively. The simulation results using RS2 confirmed that the cable bolt length and spacing were appropriate. The recommended support system was expected to provide sufficient support to the crusher chamber in terms of controlling rock mass deformation and yielding. The general conclusion was that the design approach selected for crusher chambers must be able to adequately represent rock mass behaviour and the support required to maintain long-term stability.