KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Settings

Access type: UnknownSubject: search.filters.subject.Artificial neural networks (ANNs)

Search Results

Now showing 1 - 3 of 3
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Development of a Global Climate Model for Atmospheric Temperature Using Machine Learning
    (2026-01-01)
    Okoh, Daniel
    ;
    Awuor, Adero
    ;
    Ochieng, George
    ;
    Baki, Paul
    ;
    Habarulema, John Bosco
    This article presents a novel three-dimensional global model of atmospheric temperature developed using Artificial Neural Networks (ANNs) trained on radio occultation (RO) data from the COSMIC I and COSMIC II satellite missions. Over 14.7 million quality-controlled profiles were used, providing approximately 9.5 billion data points that capture temperature variability across latitude, longitude, altitude (0-60 km), and time (2006-2025). The global domain was divided into 1296 spatial grid cells (10° × 5°) to enable localized ANN training and ensure efficient handling of regional atmospheric dynamics. Model performance was evaluated through cross-validation and independent testing against radiosonde measurements from 684 stations worldwide. Results show mean absolute errors of 1.5 °C-4.5 °C and root-mean-square errors of 2.5 °C-6.5 °C, with best performance in the tropical troposphere and increasing errors toward high latitudes. The model successfully reproduces key climatological structures (including the tropospheric lapse rate, stratospheric inversion, and seasonal hemispheric asymmetries), and accurately captures diurnal and annual thermal cycles. Long-term simulations (2006-2025) reveal a distinct tropospheric warming trend (∼+0.07 °C per year at 11 km) and a corresponding stratospheric cooling (∼-0.03 °C per year near 32 km), consistent with established satellite and reanalysis records. These results demonstrate that ANN-based frameworks can effectively model global atmospheric thermal structure and evolution, providing a scalable approach for future climate monitoring and forecasting applications.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    PEA distribution reliability (SAIFI, SAIDI) determination using artificial neural networks
    (2005-12-01)
    Kaewmanee, Keattisak
    ;
    Jiriwibhakorn, Somchat
    This paper purposes the methodology of SAIFI and SAIDI determination of Provincial Electricity Authority (PEA) distribution network in Thailand using Artificial Neural Networks (ANNs). Data used in this study was obtained from the Reliability Program[1]. The data of feeders 2 and 7 of Pattananikom substation was used as the examples in this research from January to July 2004. The three inputs of ANNs for reliability indices (SAIFI, SAIDI) consist of the number of customers behind the protective equipment, interruption frequency of protective equipment per month and total time interruption per month. Referring to the results obtained after inputting ANNs and SAIFI to be trained in neural networks, the mean absolute percentage error (mape) of feeders 2 and 7 are 0.0042%, 0.0168% respectively, while inputting SAIDI into the network, the mean absolute percentage error (mape) of feeders 2 and 7 are 0.6377%, 3.2942% respectively.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Network observability determination using artificial neural networks
    (2005-12-01)
    Tanprasert, Pornthep
    ;
    Jiriwibhakorn, Somchat
    This paper purposes a method for the determination of network observability of the Provincial Electricity Authority (PEA)'s power system in Thailand using the artificial neural networks (ANNs). The network observability problem related to the power system configuration or network topology, called the topological observability, is studied to solve the topological observability problem. The artificial neural networks (ANNs) based on back propagation learning are used as a tool to solve this problem.