An expert system for predicting the velocity field in narrow open channel flows using self-adaptive extreme learning machines. (February 2020)

• Record Type: Journal Article
• Authors: Bonakdari, Hossein; Qasem, Sultan Noman; Ebtehaj, Isa; Zaji, Amir Hossein; Gharabaghi, Bahram; Moazamnia, Marjan
• Is Part Of: Measurement. Volume 151(2020)
• Publication Date: 2020-02
• Subjects: Discharge -- Field data -- Flow velocity profile -- Open channel -- Sensitivity analysis -- Velocity; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2019.107202
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: The velocity distribution in narrow open channels was studied. A new method based on self-adaptive extreme learning machines was proposed. An explicit expression for practical engineering was provided. The proposed equation was found to be better than the existing equations. Abstract: This work investigates the ability of a new model based on powerful self-adaptive extreme learning machines to predict the velocity field in an open channel. A total of 363 velocity field data obtained in 8 different hydraulic conditions of a narrow open channel are used to develop the proposed model. The performance of the proposed model in predicting the velocity field is analysed for 3 different discharge rates that have no role in model development. According to the model prediction accuracy comparisons, the proposed model is more accurate than existing equations and can be employed successfully in velocity field predicting. Furthermore, the new model can more accurately predict the negative gradient of velocity near the free surface, which is the most significant/complex feature of a velocity distribution in narrow open channels. Moreover, a sensitivity analysis is done to surrey the effect of the proposed model on each input variable.