School of Engineering, Computer and Mathematical Sciences - Te Kura Mātai Pūhanga, Rorohiko, Pāngarau

Permanent link for this collection

https://hdl.handle.net/10292/553
AUT is home to a number of renowned research institutes in engineering, and computer and mathematical sciences. The School of Engineering, Computer and Mathematical Sciences strong industry partnerships and the unique combination of engineering, computer and mathematical sciences within one school stimulates interdisciplinary research beyond traditional boundaries. Current research interests include:
  • Artificial Intelligence; Astronomy and Space Research;
  • Biomedical Technologies;
  • Computer Engineering; Computer Vision; Construction Management;
  • Data Science;
  • Health Informatics and eHealth;
  • Industrial Optimisation, Modelling & Control;
  • Information Security;
  • Mathematical Sciences Research; Materials & Manufacturing Technologies;
  • Networking, Instrumentation and Telecommunications;
  • Parallel and Distributed Systems; Power and Energy Engineering;
  • Software Engineering; Signal Processing; STEM Education;
  • Wireless Engineering;

Browse

Recent Submissions

Now showing 1 - 5 of 1441
  • Item
    Lightweight and Efficient Deep Learning Models for Fruit Detection in Orchards
    (Springer Science and Business Media LLC, 2024-10-30) Yang, Xiaoyao; Zhao, Wenyang; Wang, Yong; Yan, Wei Qi; Li, Yanqiang
    The accurate recognition of apples in complex orchard environments is a fundamental aspect of the operation of automated picking equipment. This paper aims to investigate the influence of dense targets, occlusion, and the natural environment in practical application scenarios. To this end, it constructs a fruit dataset containing different scenarios and proposes a real-time lightweight detection network, ELD(Efficient Lightweight object Detector). The EGSS(Efficient Ghost-shuffle Slim module) module and MCAttention(Mix channel Attention) are proposed as innovative solutions to the problems of feature extraction and classification. The attention mechanism is employed to construct a novel feature extraction network, which effectively utilizes the low-latitude feature information, significantly enhances the fine-grained feature information and gradient flow of the model, and improves the model’s anti-interference ability. Eliminate redundant channels with SlimPAN to further compress the network and optimise functionality. The network as a whole employs the Shape-IOU loss function, which considers the influence of the bounding box itself, thereby enhancing the robustness of the model. Finally, the target detection accuracy is enhanced through the transfer of knowledge from the teacher’s network through knowledge distillation, while ensuring that the overall network is sufficiently lightweight. The experimental results demonstrate that the ELD network, designed for fruit detection, achieves an accuracy of 87.4%. It has a relatively low number of parameters (4.3 x 10⁵), a GLOPs of only 1.7, and a high FPS of 156. This network can achieve high accuracy while consuming fewer computational resources and performing better than other networks.
  • Item
    Dust Impact on Photovoltaic Modules: Global Data, Predictive Models, Emphasis on Chemical Composition
    (Elsevier BV, 2024-10) Almukhtar, Hussam; Tjing Lie, Tek; Al-Shohani, Wisam
    This study explores the influence of dust on optical properties such as transmittance, absorptance, and emissivity of photovoltaic (PV) modules using over 300 experimental readings from various dust types. These readings were collected during regional storms and ground sources, data encompass different weight levels. Incorporating 690 global datasets and leveraging Artificial Neural Networks (ANN) and Multiple Linear Regression (MLR) in MATLAB, the study integrates key dust chemical components (Si, Fe, Ca, Al) and weight to predict the PV optical properties. This approach enhances models’ predictive accuracy across diverse environmental settings, which in turn enables more accurate forecasting of PV power output and thermal behavior under varying dust conditions, as these optical properties govern the module equations. Additionally, comparative analysis with existing literature shows superior accuracy, achieving Mean Squared Errors (MSEs) of 1.8 and 8.44, surpassing previous benchmarks. Results underscore the global efficacy of our methodologies in revealing dust’s impact on PV module thermal behaviour and efficiency.
  • Item
    Effect of Build Orientation and Heat Treatment on the Microstructure, Mechanical and Corrosion Performance of Super Duplex Stainless Steels Fabricated via Laser Powder Bed Fusion
    (Royal Society of Chemistry (RSC), 2024-09-11) Davidson, KP; Liu, R; Zhu, C; Cagiciri, M; Tan, LP; Alagesan, A; Singamneni, S
    In this study, the effect of build orientation (0°, 45° and 90° from a build platform) on microstructural response as well as mechanical and corrosion properties was investigated by comparing laser powder bed fusion-produced samples in the as-built and solution-annealed states. By increasing build orientation, Widmanstätten γ-austenite formation was lowered because of faster cooling and shorter melt tracts, whilst retaining similar δ-ferrite/γ-austenite phase fractions. This is correlated with improved corrosion performance in the 90° orientation from chemically homogeneous grain boundary γ-austenite. The prevailing δ-ferrite as-built samples exhibit a strong 〈001〉 δ-ferrite crystallographic texture in the normal direction across all orientations together with greater hardness and mechanical strength in comparison to solution-annealed samples by virtue of less slip systems in the BCC δ-ferrite structure and fine cellular solidification structure. The 45° build orientation exhibits a greater Widmanstätten γ-austenite content and periodic recrystallisation between scan checkers, contributing to improved mechanical strength and ductility. Solution annealing softened structures, from an increase in the γ-austenite content, via intergranular nucleation or through prior grain boundaries and Widmanstätten needles. The underlying δ-ferrite grain structure and crystallographic texture relationship is retained, although weakened from the recrystallisation process. Tensile strength is reduced compared to the as-built structures and worsened in the 90° orientation due to few Widmanstätten needles, although elongation is significantly increased, and pitting corrosion performance is improved by the removal of stresses and the equilibrium microstructure.
  • Item
    Experimental Study on Human Kinetic Energy Harvesting with Wearable Lifejackets to Assist Search and Rescue
    (MDPI AG, 2024-10-15) To, Jeffrey; Huang, Loulin
    This study explores the integration of a human kinetic energy-harvesting mechanism into lifejackets to address the energy needs of aid search and rescue operations in aquatic environments. Due to the limited data on the movement patterns of drowning individuals, a human motion model has been developed to identify optimal design parameters for energy harvesting. This model is developed from computer vision analysis of underwater footage and motion capture laboratory experiments and is used to quantify the potential for power generation. The field testing experiment is conducted underwater, replicating the environment used for footage collection and analysis for the modelling. During the field testing, the participant wears a lifejacket integrated with the energy-harvesting device. Field testing data are then collected to verify the model. The efficacy of this approach is demonstrated with observed power outputs ranging from 0 mW to 754 mW in simulations and experiments. Despite challenges such as the “dead zone” in a drowning person’s motion, the success of the experiments underscores the potential of the proposed energy-harvesting mechanism to efficiently harness the kinetic energy generated by a drowning person’s movements. This study contributes to the development of sustainable, energy-efficient solutions for search and rescue operations, particularly in remote and challenging aquatic environments.
  • Item
    A Geometric Approach to Textual Augmented Data Filtering
    (IOP Publishing, 2024-09-09) Feng, SJH; Lai, EMK; Li, W
    Data augmentation is necessary if the amount of training data is insufficient for supervised learning. For natural language processing tasks, obtaining good quality augmented data is not easy. This paper introduces GATFilter, a novel method for filtering out inappropriate augmented textual data for text classification (TC). Utilizing geometric concepts, more specifically the principle component and convex hull analyses, this method adeptly preserves the semantic integrity of words within augmented texts. GATFilter is versatile and applicable across various types of textual augmentation methods. Experiments using several datasets and augmentation strategies showed that classifiers trained with GATFilter-filtered augmented data sets showed improvements in key performance metrics, including accuracy, precision, recall, and F1 score. The method’s efficacy is notably influenced by the quality of the underlying augmentation techniques, indicating its potential to complement and refine various text augmentation strategies. Furthermore, our analysis showed that GATFilter is particularly able to amplify the effectiveness of methods that generate good quality augmented data. GATFilter is openly available online on Github1, and as a Python package2
Items in these collections are protected by the Copyright Act 1994 (New Zealand). These works may be consulted by you, provided you comply with the provisions of the Act and the following conditions of use:
  • Any use you make of these works must be for research or private study purposes only, and you may not make them available to any other person.
  • Authors control the copyright of their works. You will recognise the author’s right to be identified as the author of the work, and due acknowledgement will be made to the author where appropriate.
  • You will obtain the author’s permission before publishing any material from the work.