Comparative performance of on-off, pid and fuzzy logic controllers pH process in boiler application

In steam generator, water quality is part of mandatory aspects that must be complied by industries especially in power generation due to its significant impact on plant operation. The corrosion and scale formation are the major impact which lead to tube damage and can cause a catastrophic failure due to poor heat transfer. The most common methods to control water quality in steam generator are internal water treatment and external water treatment process. This study focuses on the control and a modeling of highly nonlinear internal water treatment process by using various control strategies. The On-Off control, PID control and Fuzzy Logic control systems have been used to perform intelligent control based on the development of empirical modeling equation dilution of ammonia hydroxide solution. These systems are designed to perform an auto tuning process during forward and reverse action for pH correction. The LabVIEW virtual instrumentation (VI) has been used to verify the performance of each control strategy in term of the system accuracy and effectiveness. The simulation results show that these strategies can be applied to perform intelligent control in this dilution process. In non-linear static system, the On-Off controller shows remarkable results in forward and reverse action in maintaining the acceptable pH level. However, the Fuzzy Logic controller shows higher accuracy measurement compared to PID control strategies in both actions. The PID controller is only applicable in forward action. In order to model the nonlinear dynamic system, various tests to model the pH plant for both forward and reverse behavior has been carried-out by using Matlab system identification toolbox. The result shows the linear ARX model reached 91% best-fit performance, while it reached 89% with ARMX Model. Therefore, ibest-fit represented compare than ARMAX and state space models. On the other hand, the nonlinear modeling result shows that NARX model much more fit compare an Hammerstein-Wiener with 98.02%, which indicate the system is act more nonlinearity in process. The PID controller has been used to measure the performance of NARX and Hammerstein-Wiener model in controlling the dynamic pH response. The simulation results show that the PID controller based on NARX model provides better response than Hammerstein-Wiener based model, the reason mostly that the nonlinear function that has been used for mapping the output signal fit better than piecewise functions that been used in Hammerstein-Wiener.

Preview PDF FK 2013 69R.pdf Download (771kB) | Preview

Actions (login required)

View Item