Research Papers - Department of Mechanical Engineering

Permanent URI for this collectionhttps://rda.sliit.lk/handle/123456789/604

Browse

Filters

2022 - 20263

Advanced Search

Filter by

Settings

Author: search.filters.author.Perera, MStart date: 2022

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    PublicationOpen Access
    Distributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
    (Elsevier Ltd, 2026-10-01) Perera, M; De Hoog, J; Bandara, K; Weeratunge, H; Halgamuge, S
    Accurate forecasts of distributed solar generation are necessary to maintain grid stability amid the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with attention that focus on important regions of an image have been applied with success in many computer vision applications. However, whether such methods provide meaningful benefits for cloud movement forecasting, and how such improvements propagate through to downstream solar generation forecasting accuracy, remains under-explored. In this study, we conduct a large-scale empirical investigation of the impact of attention-based cloud forecasting on solar generation forecasting, addressing a gap that has been overlooked in the literature. To this end, we develop a pipeline that incorporates an attention-enhanced convolutional long short-term memory network and an existing self-attention-based video prediction method to forecast cloud movement using satellite imagery. The effectiveness of the resulting cloud forecasts is evaluated through their downstream impact on solar forecasting across 50 PV sites in Australia. We further provide insights into the cloud conditions under which attention-based cloud forecasting methods yield the most significant improvements in downstream solar forecasting accuracy. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
  • Thumbnail Image
    PublicationOpen Access
    Aeroacoustic Noise Produced from Novel Wind Turbine Rotor Design for Small-scale Applications in Sri Lanka
    (SLIIT, 2022-02-11) Perera, M; Bandara, U. H
    Growing concerns regarding non-renewable energy sources have driven academic and industrial scholars as well as global superpowers to seek sustainable, greener power generation alternatives. One such prominent renewable substitute is wind power which was initially utilized in harnessing electricity towards the late nineteenth century though archaeological evidence has proved that wind power had been employed for various purposes since predynastic Egypt. Extensive research and development has enabled the efficient operation of multi megawatt wind farms at present though inherent drawbacks still persist, of which aerodynamic noise, also referred to as aeroacoustic noise, is of major concern. This paper details the simulative investigation of the aeroacoustic sound levels produced by an optimized novel wind turbine design intended for the use in small scale applications with medium wind speed conditions in Sri Lanka, using ANSYS Fluent. A transient analysis using the Shear Stress Transport turbulence model was used to obtain the converged pressure fluctuations which subsequently revealed the sound pressure levels via Fast Fourier Transforms at six predetermined locations of interest. The results revealed the presence of acoustic vibrations within the Infrasonic and Low Frequency Noise range with sound pressure levels exceeding one hundred decibels, particularly up to a frequency of twenty five Hertz. Prolonged exposure to elevated levels of low frequency noise has been identified to cause severe discomfort to humans though further conclusive research is required. Finer mesh controls which incorporate minute boundary layer variations during motion and precisely encapsulate the turbine geometry could further improve the accuracy of the results, however this would require adequate computational capacity. The results of this research primarily serve as a basis for identifying possible improvements for the novel rotor design in addition to providing a comparative study for future research, both simulative and empirical, on the aerodynamic noise emissions associated with wind turbines.
  • Thumbnail Image
    PublicationOpen Access
    Development of an underwater robotic arm using multibody dynamics approach
    (2022-02-05) Fernando, S; Perera, M
    Underwater robotic arms are important devices that enables workers to carry out tasks remotely from a safe distance reducing or eliminating the risks that are involved with the task. The primary objective of the robotic manipulator is to perform maintenance and cleaning activities of the hull of a ship. However, the control of these devices underwater is quite complicated due to the numerous factors that make these systems unstable and non-linear. The aim of this study is to develop a multibody dynamic robotic manipulator model, integrated with a control strategy to optimize and obtain stable kinematics solutions. The hydrodynamic forces are integrated to the manipulator model considering buoyancy forces and surface drag forces. A basic algorithm is used to generate the joint angles using 7 geometrical parameters. The control of the manipulator was done to simply follow any path that represents the given coordinates. The P, I and D parameters are tuned individually to optimize the kinematic solution of the manipulator. 3-DOF articulated manipulator is the commonly used manipulator configuration. However, a 6-DOF manipulator configuration was selected in this study to allow for change in orientation using wrist motions.