Wiener modeling and identification of a reverse osmosis desalination process using least square support vector machine

Reverse osmosis (RO) desalination is the most common method for purifying brackish water. Due to its sensitivity to quality of the feed and plant operating conditions, RO desalination process needs an efficient and accurate control system to maintain operation at optimum conditions that ensures the least energy utilization and prevent scaling and fouling. Nonlinear systems identification techniques have been used widely to model many chemical processes. Recently, support vector machines (SVMs) and least squares support vector machines(LS-SVMs) have demonstrated powerful abilities in approximating linear and nonlinear functions. In this Paper, an algorithm to identify the Wiener models using least square support vector machine regression is developed and used to identify a Hollow Fiber B-10 Permasep Permeator reverse osmosis (RO) desalination process. The obtained results showed 96 % matching of the model output and actual system output variances.