Surge avoidance using speed and valve methodologies controlled by PID, fuzzy and neural networks approaches

Surge is a global nonlinear instability that affects compression systems. It limits the operating range and degrades the system performance. This paper introduces two common surge avoidance techniques to extend the operation range of the compressor. These two methods are controlling the compressor flow and controlling the compressor speed. Air bleeding is utilized by two bleeding valves one of them with high gain and the second with low gain. Fuzzy logic, PID, and neural networks were used to control such two methods. A comparison between the three control techniques is performed to demonstrate the most effective of them using simulations based on SIMULINK. The results show that the two bleeding valves produce more stable behavior than using one bleeding valve. Furthermore, neural network controller shows quick arrival for the different set points and also is more stable than other techniques in avoiding the surge.