Browsing by Subject "Machine Learning,"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Data-Driven Discovery of Governing Equations(Pulchowk Campus, 2020-07) Shikhrakar, Rojesh ManTheoretical equations are the basis of scientific progress. Many scientific domains still lack the appropriate theoretical model to reason about the phenomena. With the rise of data, there is an increasing need for methodology in data-driven science and engineering for understanding the physical phenomena. This thesis on Data-Driven Discovery of Governing Equations aims to provide a method for model discovery and find governing partial differential equations from data by training physics-informed neural networks. Given data, our method generalizes a neural network to compute a matrix of candidate terms for Partial Differential Equation(PDE). Minimizing the residuals from the candidate matrix allows us to find the coefficients for the governing equation. We present a framework to discover PDE not restricted to first-order time derivative equations.Item Machine Learning Approach for Fault Detection and Diagnosis of PV Modules(I.O.E. Pulchowk Campus, 2023-11) Maharjan, KabinaFault analysis in solar Photovoltaic (PV) arrays is a crucial aspect that helps to increase the PV system’s efficiency, safety, and reliability. Faults and defects, if not detected, it not only compromises the system’s generation but also accelerate system aging and jeopardizes the functionality of the overall system. Currently, the solution involves manual monitoring by system operators, but this approach is time-consuming, prone to inaccuracies, and poses safety risks. Therefore, it is imperative to implement automatic detection and diagnosis methods to detect faults to ensure the PV systems’ safety and reliability. Existing techniques either lack the precision to provide detailed fault information or are overly complex. This research introduces a fault detection and diagnosis method in solar PV systems using a machine learning approach. The research extends to defining normal conditions and four distinct fault categories for the proposed fault detection and classification algorithm. A predictive model is prepared using a machine learning approach to forecast the DC output power. The trained Multilayer Neural Network (MNN) model is found to have the Root Mean Square Error (RMSE) of 6.74 and 6.11 for the training and validation sets. By analyzing the difference between the power predicted by the MNN model and the actual PV system power, the predefined fault types in the PV modules are detected.