Abstract⎯ ⎯ ⎯ Computing the optimal values of Proportional Integral derivative (PID) control gains is an important task in the design of PID controller. This paper presents the application of Sugeno fuzzy model for on-line tuning of PID controllers in pH process. The optimal PID controller parameters required to develop the Sugeno fuzzy model are estimated by genetic algorithm. The developed fuzzy controller can give the PID parameters on line for different operating conditions. The suitability of the proposed approach has been demonstrated through computer simulation using MATLAB Simulink. Index Terms⎯ ⎯ ⎯ pH process, PID controller, Genetic Algorithm, Sugeno Fuzzy Logic. I. INTRODUCTION Proportional-Integral-Derivative (PID) controller has been widely used in process industries, as they have a simple structure and are robust against disturbance. The tuning of PID controller parameters is necessary for the satisfactory operation of the plant [1, 2]. Generally, the PID controller parameters are determined using Ziegler-Nichols (ZN) [3, 4] and Cohen Coon methods [5]. In both these methods, tuning is obtained for an operating point where the model can be considered linear. This implies that there is sub- optimal tuning when a process operates outside the validity zone of the model. In general, plant parameters change due to ageing of the plant or changes in the load. Hence, the retuning of the controller parameters is necessary. To solve the above said problems adaptive control approaches have been proposed. Adaptive control techniques provide robustness over changes in the plant dynamics. However, in the case of plant parameters which are unknown or change over large ranges, the adaptive control strategies may become complex, requiring considerable computation time and may lead to instability [6]. Hence, a different approach to the tuning of PID controllers is necessary. Recently, many artificial intelligence (AI) techniques have been employed to improve the controller performance for a wide range of plants while retaining their basic characteristics. In [7] fuzzy logic has been applied for the K. Valarmathi , Department of Electrical and Electronics Engineering, Kalasalingam University, Krishnankoil – 626190, Tamilnadu; e-mail: kvalarmathi@yahoo.com. D.Devaraj, Prof and Head, Department of Electrical and Electronics Engineering, Kalasalingam University, Krishnankoil – 626190, Tamilnadu. T.K.Radhakrishnan, Prof, Department of Chemical Engineering, NIT, Trichy tuning of PID controller parameters. Genetic Algorithm (GA) based methods have also been used in setting the parameters of PID controllers [8, 9]. Mwembeshi et al., [10] proposed GA based Internal Model Control (IMC) controller for pH process. Tan et al., [11] explored the use of GA techniques for designing a wiener model controller for the pH process. But GA approach suffers from computational burden and memory. Hence these techniques are impractical for on-line application. This paper proposes a Sugeno fuzzy model to obtain the optimal PID gains during on-line applications. The optimal PID gains of the nominal operating condition required to develop the fuzzy system are obtained through GA approach. II. MODELING OF PH PROCESS The pH is the measurement of the acidity or alkalinity of a solution. The pH process consists of neutralization of two monoprotic reagents of a weak acid and a strong base as shown in Fig.1. The Continuous Stirred Tank Reactor (CSTR) has two inlet streams: the influent process stream, and the titrating stream, and one effluent stream at the output [12]. The model of the pH neutralization process used in this work is suggested by McAvoy et al. [13] and is given below. Fig.1 Block Diagram representation of pH Process While developing the model of the pH neutralization process, it is assumed that the mixing is perfect. The material balances in the reactor can be given by ( ) a X b F a F a C a F dt a dx V + - = (1) ( ) b X b F a F b C b F dt b dx V + - = (2) Where a F is the flow rate of the influent stream, b F is the flow rate of the titrating stream, a C is the concentration of the influent stream, b C is the concentration of the titrating stream, a x is the concentration of the acid solution, b x is A Combined Genetic Algorithm and Sugeno Fuzzy Logic based approach for on-line Tuning in pH process K. Valarmathi, D.Devaraj and T.K.Radhakrishnan 2008 4th International IEEE Conference "Intelligent Systems" 978-1-4244-1739-1/08/$25.00 © 2008 IEEE 20-2