Experimental System Identification and Signal Processing of V94.2 Gas Turbine Fuel System

The fuel system is one of the significant subsystems of governor which controls output power of power plants. Governing of gas turbine is the procedure of monitoring and controlling the flow rate of fuel into the turbine with the objective of maintaining its speed of rotation at 3000 rpm in V94.2 gas turbines. The flow rate of fuel is controlled by interposing valves between the fuel tank and the turbine. In V94.2 gas turbine these valves are servo electro hydraulic type. At once governor receives command from grid power to rise or decrease the output power, begins processing it in its PLC and then sends a current signal of 4-20mA to coils on servo valve. Subsequently this current can apply a torque on servo flapper which causes a movement in actuator and so it can change the opening of control valve [1]. Of course this response must be at a short time [2]. To measure the dynamic response parameters such as delay time, rise time, settling time there are two ways, 1-Fast sampling of output signal, 2-Using a dynamic model between input and output signals. In this article both of these methods will be used. As a result of first method we will directly acquire the dynamic response parameters and from second method be able to simulate and analysis the system response and then optimizing the controller parameters if it is needed such as Kp, Td, Ti, zeroes and poles positions. Fast sampling will be done by a data acquisition card and then collected data will be used in “system identification” toolbox of MATLAB to obtain a dynamic model through Black box method. This paper is going to describe the way to obtain the fuel system dynamic model through the practical tests based on the "Black Box system identification” method.