Nonlinear system identification and control of a pH process using Local Linear Model Networks strategy

Control of pH is a common problem faced in a range of industrial systems, such as chemical, biochemical industries and waste water treatment, due to its high nonlinearity. The main objective in this paper is to identify a pH model from real data by using an application of Local Linear Model Networks (LLMN) and subsequently develop a nonlinear control system for the process. The identified LLMN model of the pH process is tested as an independent model and used for the control simulations. The conventional feedback Proportional Integral (PI) controller parameters for different operation points, identified by the LLMN model, are designed by using a direct design method. These local controllers are then combined into a LLMN controller. Results show the ability of LLMNs for nonlinear system identification and the designed nonlinear control structure results in a good performance.