Artificial Intelligence–Based Optimization of Reverse Osmosis Systems Operation Performance

Abstract

In recent years, reverse osmosis (RO) systems have been highly utilized in industrial processes. One of the most important operational issues of these systems is membrane fouling, which leads to high operating costs and environmental impacts. The purpose of this research is to optimize RO systems’ operation to reduce fouling, increase membrane life span, and minimize system costs. To achieve this purpose, first, RO system characteristics are simulated using a general regression neural network (GRNN) artificial neural network. Then, the controllable factors affecting the performance of the system are optimized by the application of a single-objective optimization model with the total operating cost minimization as an objective function. The proposed method is applied to an under-operation RO system used in a car manufacturer factory in Iran. Based on the results, the optimal values of the inflow, inlet pressure, and recovery rate were 10.4 m3/h, 7.4 × 105 Pa, and 60%, respectively. Accordingly, the total operational cost of the system will be $1,525.95. Moreover, by an appropriate operation, the system can continue to work for more than 5,000 h without the need for cleaning.