This research examines the behavior of a weir-gate hydraulic structure when subjected to the presence or absence of a side obstacle. The study focuses on two scenarios, free flow and submerged flow, and employs a side obstacle with dimensions of 15 cm in height, 30 cm in length, and 1 cm in thickness, located 30 cm downstream from the weir-gate structure. A correlation matrix is derived using measured and calculated data to establish the relationship between various variables, both dimensional and non-dimensional. The results show a clear correlation between Froude number and discharge coefficient, while a random trend is observed between discharge coefficient and Reynolds number. A proper trend is found between Reynolds number and Froude number at the downstream location. The discharge coefficient is significantly influenced by factors such as gate water depth, weir water head, and vertical distance between the weir and gate. Additionally, the study investigates the effect of average water depth at the downstream on discharge and flow velocity. Two models, linear regression and artificial neural network, are developed and tested to estimate the coefficient of discharge (Cd). The ANN model proves to be the best-fit model for all scenarios, with an MSE of 0.00013141 and an R-squared value of 0.99. The machine learning algorithms used in this study demonstrate an increase in prediction accuracy for Cd.