Modeling and Classification a Vibration Signal of Steam Turbine Based on Fuzzy System

Abstract

Vibration signal is a complex and nonlinear signal, especially in steam turbine (ST) where vibration in a dynamic system is classified as acceleration, velocity, and displacement. However, these modes either damage or destroy the turbines when they are in severity region. This work presents a modeling of vibration signal based on a FFT signals related to acceleration modes firstly, then the design and implements a two fuzzy system by Sugeno and Mamdani fuzzy methods to produce a linear signal to be used in a control system. The input and output Membership Function (MF) is set based on the ISO 10816-2 and experienced work in a power plant, respectively. The satisfactory results that obtained by Sugeno fuzzy method produces, it is suitable output and reduces the complexity and non-linearity of the vibration signal. The output of this simulation, with more steps of signal processing, can be employed with a control system for prevention vibration in ST.