Author(s): B. Kumar; Vinayakam Jothiprakash

Linked Author(s): V. Jothiprakash

Keywords: PKW; Discharge; ANN; Empirical-Equations; Accuracy

Abstract: A Piano Key Weir (PKW) considered as a modern free flowing non-labyrinth weir. Commonly such weirs are fitted in low head dams to increase the discharging capacity. So far, no unique design guidelines of a PKW are presented for the discharge coefficient. For an accurate design of PKW, it is needed to calculate the coefficient of discharge precisely. Several works are found on the assessment of discharge coefficient but very few studies investigated the application of Artificial Neural Network (ANN) for assessing the coefficient of discharge considering large amount of experimental data reported. In this current research, ANN model has been developed to predict the coefficient of discharge based on the available data points in the literature. Out of the several input -output data collected from the literature (set of 395 data points) randomly 70% of the data set has been exercised to train the ANN model, 15% of the data length is used for testing and rest 15% was used for validation purposes. On considering the performance metrics, it is noticed that the ANN model has greater accuracy than the non-linear regression methods. The efficiency of the ANN model is found to be very high with a correlation coefficient of 0.996, and Root Mean Square Error is less (0.025) with Mean Absolute Percentage Error around 1.065. The present study concludes that ANN model can be considered for the PKW in assessment of coefficient of discharge precisely, within the trained limits.

DOI: https://doi.org/10.3929/ethz-b-000675921

Year: 2024