Artificial Neural Networks Based Modeling of the ReverseOsmosis Process Performance

Majid Rasheed, Asseel

	Artificial Neural Networks Based Modeling of the ReverseOsmosis Process Performance
journal of kerbala university
Article 1, Volume 6, Issue 1, 2010, Pages 158-176 PDF (0 K)
Author
Asseel Majid Rasheed
Abstract
This investigation presents a methodology and practical guidelines for developing predictive models for reverse osmosis plants by a data-based approach using neural networks based on the back-propagation algorithm. This study utilizes actual operating data from reverse osmosis (RO) desalination plants. Our resulting neural network model is capable of accurately predicting the actual operating data from RO desalination plants, but the accuracy of a neural network model depends on both the proper selection of input variables and the broad range of data with which the network is trained. A neural network model can handle noisy data more effectively than statistical regression and performs better in predicting the performance variables of RO desalination plants. Permeate flux and salt passage are the key performance parameters. They are mainly influenced by variable parameters such as pressure, temperature, salt concentration, feed flow rate and pH of feed water. When the temperature of feed water is increased, the permeate flux and salt passage increase and permeate flux increases with decreasing pH and concentration of feed water and increasing pressure. The salt passage decreases with increasing pressure, when the concentration of feed and pH decrease too. When increasing the feed flow rate, the permeate flux and salt passage would be increased. A good agreement prediction is obtained using the ANN predictions and the experimental data with a deviation not more than 2% for most of the cases considered. The ANN interpolative levels (which were not represented in the training phase) is shown to be of lesser quality.

Statistics Article View: 191 PDF Download: 149