Title: Application of deep learning algorithm in hydrometry

Authors: Mohammad Zakwan; Shaik A. Qadeer; Mohammed Yousuf Khan

Addresses: Department of Civil Engineering, Polytechnic Hyderabad, Manuu, Hyderabad-50032, India ' Department of Electrical Engineering, Muffakham Jah College of Engineering and Technology, Hyderabad-50032, India ' Maulana Azad National Urdu University, Hyderabad – 500032, India

Abstract: Estimation of discharge in a river is an integral part of water resource engineering. In this regard, various artificial intelligence (AI) techniques have been employed to model the discharge ratings. The present work compares the performance of two neural networks namely, back propagation neural network (BPNN) and radial basis neural network (RBNN), to model the discharge rating. The estimated discharge was also compared with the discharge estimated using conventional method. Published data of two gauging station was used for the comparative analysis. It was observed that application of neural networks greatly improved the estimates of discharge as compared to conventional method. Application of artificial neural network (ANN) reduced the sum of square of error (SSE) by about 90% on an average. Maximum absolute error was reduced from 51.36 and 141.21 to 5.04 and 7.68 respectively for the two stations for RBNN when compared to conventional method during validation. Calibration results reveal that among the BPNN and RBNN, RBNN could model the ratings at both the stations, better than BPNN.

Keywords: hydrometry; discharge; river; ratings; neural networks; back propagation neural network; BPNN; radial basis neural network; RBNN; artificial neural network; ANN; sum of square of error; SSE.

DOI: 10.1504/IJHST.2024.135135

International Journal of Hydrology Science and Technology, 2024 Vol.17 No.1, pp.91 - 100

Received: 08 Jun 2021
Accepted: 16 Oct 2021
Published online: 01 Dec 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article