Abstract
AbstractCoagulation process is complex and nonlinear and its control plays a crucial role in a water treatment plant (WTP). Traditionally, aluminum sulphate (alum) is used as a coagulant in the coagulation process and the optimum coagulant dose is determined using a jar test. The jar test is quite time-consuming and expensive too. Jar tests are conducted periodically, which means they are reactive rather than proactive. The development of predictive models for coagulant dose in a WTP is needed. The aim of this study was to use an artificial neural network (ANN) to predict coagulant dose. For ANN modelling, the plant laboratory provided data for 48 months of daily water monitoring in terms of inlet and outlet water turbidity and coagulant dosage. By applying various training functions and evaluating the coefficient of regression (R) and mean square error (MSE), the appropriate architecture of the cascade feed forward neural network (CFNN) and feed forward neural network (FFNN) coagulant models were developed. Additionally, the best performing Levenberg–Marquardt (LM) and Bayesian regularization (BR) training functions among resilient back propagation (RBP), one step secant (OSS), conjugate gradient (CG), and variants of gradient descent (GD) were used to build four ANN models of FFNN and CFNN for predicting coagulant dose at WTP. With 40 hidden nodes, a cascade feed forward neural network with BR training function produced very accurate estimates for coagulant dose with architecture (2-40-1) with the highest values of R = 0.952 during training, R = 0.922 during testing, and overall R = 0.947 and MSE = 99.28 mg/lit. As a result, plant operators and decision-makers may benefit from using ANN as a performance assessment tool, as well as an important diagnosing tool for understanding the nonlinear behaviour of the coagulation process.KeywordsArtificial neural networkWater treatment plantCoagulant dose
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.