On-Line Identification and Nonlinear Control of pH Processes

Abstract

The control of pH is widely recognized as a difficult problem. A novel method for representing the nonlinearity of pH processes is developed. A system consisting of a combination of known and unknown chemical species is viewed as a combination of impulses (for the known species) and a piecewise continuous distribution (for the unknown species). This results in on-line identification of a small number of parameters that can be used to affect the process nonlinearity over a wider pH range than if only impulses were used. Computer simulations and experimental results using the proposed method for large step changes in concentrations of a simple chemical system demonstrate the performance of this approach. In particular, the simulations and experiments provide comparisons between the control of systems with and without identification and investigate the tuning of the controller and identification algorithm.

Hybrid Fuzzy Logic and Pid Controller for PH Neutralization Pilot Plant

“…A comparison of linear and non-linear control for neutralization plant is presented in [20]. In [21,22], author presents the investigation of the controller performance, such as tracking of the lime flow-rate set point, investigation of different conditions of normal process operation and operation without ignition. A new approach to control which utilizes an identification reactor to incorporate the nonlinearities of the neutralization process is described in [23].…”

Adaptive Fuzzy Control of a pH Process

“…According to the sensitivity of pH to the amount of acid or base, the whole dynamic curve can be roughly divided into a high-gain region, roughly pH∈ [5,9], and two low-gain regions, pH> 9 and pH< 5, as shown in Figure 2. In the highgain region, a small change of the base flow rate would result in a large change in pH value, while in the low-gain regions, considerable changes have to be made to make an appreciable pH change.…”

“…In the modeling aspect, rigorous models from first principles involving the material balance and equilibrium equations were established in [4] and later extended in [5]- [7] through the concept of the reaction invariant, and more complicated situations were considered in [8]. Besides these physicochemical models, parameter models based on identifications were developed and particularly for control-oriented [9]- [11]. The control of pH processes is beyond the conventional proportional-integral-derivative (PID) feedback mechanism.…”

State Estimation in Nonlinear Processes. Application to pH Process Control

“…In this case, the control law q B calculated through eq 10 is with Figure shows the measured pH, which was obtained by adding to the real deterministic system output a white noise signal previously filtered by a first-order transfer function. Although many authors have considered the measured pH only corrupted with white noise, herein we contemplate the more realistic situation where the sensor dynamics is considered. This is an important assumption in order to test the proposed observer/controller structure under more disadvantageous operation conditions.…”