Journal Articles - Engineering Technology - 2022

Permanent URI for this collection

http://repository.vlu.edu.vn:443/handle/123456789/300

Browse

Browsing Journal Articles - Engineering Technology - 2022 by Author "H. Nguyen-Xuan"
  • Publication
    A semi-empirical approach and uncertainty analysis to pipes under hydrogen embrittlement degradation
    ( 2022)
    Hieu Chi Phan
    ;
    Luan Le-Thanh
    ;
    H. Nguyen-Xuan
    The Hydrogen Embrittlement (HE) appearance is the main reason for the desire to evaluate the burst pressure of pipes. Several existing models predict burst pressure of hydrogen-induced pre-cracking pipes with the fracture mechanics criterion (fracture toughness, Jcr). Still, none of them has taken into account the model uncertainty. In this paper, we propose a surrogated model associated with molecular models. The discrete values of the dimensionless influence functions presented for elastic and plastic components of J are connected by machine learning (Random Forest Regression, RFR) and third-order polynomial functions with optimized factors to avoid the inconvenience of using the lookup table. The molecular empirical models are obtained by the up-to-date Balancing Composite Motion Optimization (BCMO) algorithm. Because samples of elastic and plastic databases are limited and the final output, Pb, is not directly derived from molecular models, all databases are used for training without the conventional data splitting to train and test sets. Consequently, the proposed approaches (RFR model and Empirical model) are validated based on the experiments collected globally. The final models are accounted for the residual distribution as the unavoidable component. The efficiency of the Random Forest and empirical models is validated by experiments or simulations from the literature when the evaluation metrics (i.e., R-square, Mean Absolute Error) are (0.9666, 0.8695 MPa) and (0.9701, 0.7996 MPa), respectively. The drawbacks of the proposed models, heavily dependent on databases, are also illustrated and discussed for further development. Strict boundaries of input and output, especially the fracture toughness, which is commonly degraded due to the effect of HE, from the test set combined with the uncertainty of the models based on the analysis of model residuals are also proposed for validation.
  • Publication
    Machine learning-based real-time daylight analysis in buildings
    ( 2022)
    Luan Le-Thanh
    ;
    Ha Nguyen-Thi-Viet
    ;
    Jaehong Lee
    ;
    H. Nguyen-Xuan
    Daylight analysis is essential in building design to ensure indoor environment quality, including health and thermal comfort vis-à-vis energy. It is a repeating and time-consuming process of design options. Several studies conducted machine learning models to accurately predict daylight performance in particular design situations. Therefore, developing an AI-based real-time daylight analysis platform becomes more promising. However, buildings can be designed with arbitrary shapes, creating a real challenge for the AI to recognize any building layout. From that perspective, the idea of finding the design variables that characterize all the building layouts becomes the key solution. To unlock this challenge, we promote a novel method of creating design variables and building a machine learning model that can efficiently forecast daylight performance with different building layouts. The daylight metric was Useful Daylight Illuminance with four ranges, and the case studies were assumed medium-sized buildings located in Ho Chi Minh City, Vietnam. All the data for training and predicting were created by the simulation DIVA tool. Obtained results showed the excellent performance of the proposed approach, which brings more promising in developing a data-driven machine learning platform for real-time daylight validation. Moreover, the present framework can adapt to any specific machine learning model or daylight simulation tool and daylight metrics.