Pusat Pengajian Kejuruteraan Aeroangkasa - Tesis

Browse

Recent Submissions

Now showing 1 - 5 of 64
  • Thumbnail Image
    Publication
    Performance analysis of yolo and ssd-based deep learning models for detection of oil palm trees in drone images
    (2025-07-01)
    Istiyak Mudassir Shaikh
    This study explores the use of advanced deep learning models for detecting and counting oil palm plants in precision agriculture using drone-based high-resolution images. The motivation stems from the limitations of manual monitoring methods, which are time-consuming, error-prone, and not feasible for large-scale plantations. Given Malaysia’s significant role in global palm oil production, efficient and automated detection systems are essential to support sustainable plantation management. The primary challenge is to accurately identifying oil palm trees in complex conditions, such as overlapping canopies, dense vegetation, varying lighting, and similar surrounding plants. These factors limit traditional image processing techniques, prompting the use of robust deep learning frameworks. This study evaluates four state-of-the-art object detection models: YOLOv5x, YOLOv7, YOLOv8, and SSDv2FPN, selected for their real-time detection capabilities and accuracy in agricultural environments. Two datasets were used: a smaller set of 10 drone images containing 79 annotated palm trees, and a larger dataset of 482 images with 5,233 trees. Evaluation metrics included True Positives, False Positives, False Negatives, Precision, Recall, F1-Score, and Detection Time. SSDv2FPN achieved perfect precision at 100% with an F1-Score of 89.49%, but required 83 seconds per image, which limits its suitability for real-time applications. In contrast, YOLOv5x, YOLOv7x, and YOLOv8x detected palm trees in relatively lower execution time of 16, 12, and 14 seconds respectively, with YOLOv5x achieving an F1-Score of 97.36%. These results demonstrate the clear advantage of YOLO models with regard to high speed execution. On the larger dataset, YOLOv8 models outperformed other frameworks, thereby achieving F1-Scores between 97.36% and 99.31%, precision values ranging from 99.27% to 99.70%, and recall rates between 95.89% and 99.36%. Among the YOLOv8 variants, YOLOv8s and YOLOv8n demonstrated the fastest detection times of 28 and 33 seconds, respectively, effectively balancing rapid inference and detection performance. This makes them ideal for deployment in practical agricultural monitoring systems.
  • Thumbnail Image
    Publication
    Fatigue failure analysis of cfrp composite laminates using modified stiffness degradation method
    (2025-05-01)
    Nabilah Binti Azinan
    Carbon fibre reinforced polymer (CFRP) composite laminates have been extensively utilised in various industrial applications due to their outstanding mechanical properties. In this study, the stiffness degradation behaviour of CFRP laminates subjected to cyclic loading conditions was analysed using two analytical models: the Modified Stage I model and the modified Stage I to Stage III model. Both models were developed based on the original model proposed by Lurie and Minhat, and subsequently enhanced through the introduction of several specific parameters to improve the predictive accuracy between analytical results and experimental data. The application of a linear elastic model facilitated the determination of the mechanical properties of the CFRP materials, which served as input parameters for the modified models. The modified Stage I model was employed to evaluate stiffness degradation during the initial damage phase, particularly matrix cracking. The resulting analytical stiffness degradation curves were compared with the experimental curves. Subsequently, the modified Stage I to Stage III model was applied to assess the full spectrum of stiffness degradation, encompassing damage evolution from the initial stage to the final stage, including fibre breakage. To evaluate the accuracy of both models, the percentage difference between the analytical model and experimental data was calculated for all three CFRP laminate configurations investigated in this study, namely [0,±45]s, [0,902]s, and [0,90,±45]s. Overall, the findings indicate that the Modified Stage I model offers the highest predictive accuracy when directly compared with experimental data, particularly for early-stage damage. The simplicity of its formulation and its strong agreement with empirical results render it highly effective and reliable for predicting fatigue-induced damage in the initial service life of the structure. However, to provide a more comprehensive representation of the material’s stiffness degradation behaviour throughout its service life, the modified Stage I to Stage III model presents a balanced compromise between predictive accuracy and the ability to capture the complete damage progression. Therefore, the selection of the most appropriate model should be guided by the intended application. For short-term assessments or early-stage damage monitoring, the Stage I model is the most suitable. Conversely, for long-term durability predictions involving comprehensive structural integrity evaluation, the Stage I to Stage III model is more appropriate, as it demonstrates a high level of accuracy and inclusiveness that closely aligns with experimental observations.
  • No Thumbnail Available
    Publication
    System design optimisation of dual thrust solid rocket motor using pso algorithm
    (2023-04-01)
    Ahmed Mahjub M. Alhaj
    So far, chemical rocket propulsion is the only practical means to reach space. Among this rocket propulsion category, Solid Rocket Motor (SRM) placed itself well as the simplest and cheapest propulsion system compared to liquid and hybrid propulsion systems. The design of modern SRMs is a complicated and highly interactive process that involves significant trade-offs between competing objectives such as performance, cost, safety, size, weight, etc. One of these designs is the dual￾thrust single-chamber SRM (DT-SRM) that is used in several new rocket vehicles. In the literature, the design of DT-SRM has only been treated from the propellant grain perspective ignoring the inherent interaction with other motor design parameters.Moreover, most of these design models were numerical which increased the computational burden. Such approach is limited when implemented on a real-world problem, where all design parameters are affecting each other, and subsequently need to be optimized from a system point of view to give a more efficient and reliable design. Consequently, the ultimate goal of this thesis is to propose, validate, and implement a cost-effective and integrated system design optimisationframework to the preliminary design of DT-SRM based on the main motor design disciplines namely, casing, propellant grain, nozzle, and mass. To satisfy the condition of the dual thrust, an analytical mathematical model for the complex 3-D finocyl propellant grain shape is built and integrated to the other motor design models to form a system design model. The particle swarm optimisation (PSO) algorithm was utilized, with slight modification, in a multidisciplinary design optimisation (MDO) framework to search for the optimum design variables that minimize the total mass of the motor under specified performance and geometrical constraints. The validation results against existing test-firing data confirmed the reliability of the proposed system designoptimisation approach. The proposed approach was successfully implemented in thepreliminary design of a DT-SRM test case, where the performance simulation results of the optimized system showed an improved DT-SRM system design. As a complementary study, the design optimisation results were supported by a flight performance analysis using a six-degrees of freedom (6-DOF) simulation, in which the optimized DT-SRM was capable to deliver a certain payload to a certain distance in a stable flight path. The proposed approach represents an affordable and reliable tool to the preliminary design and optimisation of DT-SRMs intended to serve as a propulsion system for medium-size aerospace vehicles.
  • Thumbnail Image
    Publication
    Self-supervised learning frame work and localization using micro air vehicles for water leak detection
    (2024-09-01)
    Mohd Yussof, Nurfarah Anisah
    Real-time detection and localization of water leakage are crucial for effective watermanagement in smart buildings. Traditional detection technologies based on static sensors frequently entail significant costs for both installation and operation. Utilizingsmall mobile robots such as Micro Air Vehicles (MAVs) provides a cost-effective and efficient for detecting objects in confined areas. Nevertheless, due to constrained processing power andlimited payload capacities, MAVs canonlydependonlightweight sensors, such as, tiny thermal sensors, which provide low image resolution and hence reduce detection distances typically within 1 𝑚. Therefore, this research presents a Self-Supervised Learning (SSL) framework, where a computer vision algorithm is developed to directly detect water leakage from thermal images, which then is used as supervised output for the training of a deep learning model by using RGB images as input. A pre-trained YOLOv4-tiny model is fine-tuned using 1080 laboratory images. Training test with 50,000 steps and 340 negative images achieves an optimal balance of accuracy, with a detection time of 0.0617 𝑠 and an average precision of 98.97%. In addition, a control strategy that combines the RGB deeplearning modeland the thermal vision algorithm is shown to allow autonomous MAVs for preliminary predictions ofwater leakage from further distances and accurately localize the leakage areas when they get closer. To validate the proposed concept, static detection tests were conducted, followed by flight tests in indoor environments. In static tests, the SSL-trained model extends the detection range from 1 𝑚 to 3 𝑚. In real-world flight tests, two scenarios are conducted: three experiments with varying initial positions and six experiments targeting different leak locations. Both static and flight tests confirm the effectiveness of the control strategy and detection algorithm in localizing water leaks in indoor environments. This research advances the sensory capabilities of MAVs equipped with RGBand thermal cameras and extends their detection range of water leakage, thereby mitigating potential damage in large or complex indoor environments.
  • Thumbnail Image
    Publication
    Stability and tribological performance of dispersed graphene (gr) and aluminium nitride (aln) nanoparticles in mineral oil
    (2024-08-01)
    Ku Wadzer, Ku Nooryasmin
    Additional nanoparticles often agglomerate and are incompatible with base fluids. With time, the mixed nanoparticles would phase-separate from the fluids, losing the benefits of aggregated nanofluids. Thus, this work examines the stability, thermal conductivity (TC), rheology, and tribology of graphene (GR) and alumnium nitride (AlN) in SUNISO 3GS refrigerant lubricant (compressor oil) and PETRONAS SYNTIUM 500 (engine oil). The study determines the best nanolubricant surfactant, the optimum concentration of GR and AlN nanoparticles in SUNISO 3GS and Engine oil 15W-40 nanolubricant, and their rheological and tribological properties. Nanolubricants' stability has been examined by visual observation and UV-vis spectrum intensity. Thermal conductivity has also been measured by changing volume percentage and surfactant presence. The highest percentage difference for TC is 22.58 in comparison with pure compressor oil. Viscosity readings at various temperatures, the ASTM 2270 standard for viscosity index measurement, and flash point were used to estimate rheological parameters. Tribological characteristics were measured using a pin-on disc tribotester to quantify wear rate according to ASTMG99. This study revealed that CTAB is the best surfactant for GR and SPAN80 is the best surfactant for AlN for both compressor oil (CO) and engine oil (EO). Next, the samples with 0.1 vol% with surfactant for both oils have the best stability and highest thermal conductivity. Furthermore, based on rheology performances, GR(0.05)-EO(CT), AlN(0.1)-EO(SP), GR(0.05)-CO(CT) and GR(0.1)-CO have the best performance. Moreover, addition of surfactant proves that it can improve the tribological performance where GR with 0.05 vol% with CTAB for both CO and EO shows the best tribological performance. Lastly, by using pin-on disc tribotester, it can be observed surfactant in nanolubricants reducing the specific wear rate (SWR) and the highest percentage difference is high as 76.37% for nanolubricant.