Research Of Model Predictive Control On Combined Cycle Power System

Compared with other power generating equipment,Gas turbine based combined cycle power plant can cascade utilize of energy and has the advantages of high power and high efficiencies.Combined cycle power plant is becoming more and more important in the electric power industry.To control this complex nonlinear multivariable system and ensure the producing process securely and efficiently,we need to induce advanced control strategies.Model predictive control method has no derivation of complex precise model but a simplified plan of designation for control system with system robustness.So it is suitable for combined cycle power plant.In this paper,we propose the research of control system on an E class gas turbine based combined cycle power plant.The control methods we apply include conventional PID control method,Dynamic Matrix Control method(DMC)and Generalized Predictive Control(GPC).Then we focus on researching the dynamic performance of these systems in the stable condition.After the simulation experiment of the conventional control system about static performance and dynamic performance,we find out the drawback of temperature control mode in system and propose a modified version which provides the better temperature responses.To solve temperature control problem in the conventional control system,we build a DMC controller as outer loop which cascades with original part in a series connection control system.Using the DMC controller we can implement the optimal control on temperature.Furthermore,the control performance and the system robustness are promoted after adjusting the DMC controller parameters.To research on the Generalized Predictive control method,we build the Linear Parameter Varying model(LPV)of combined cycle power system and find out if LPV model is useful in the GPC method as the predictive model by testing.We design the integral GPC controller for combined cycle power system according to the GPC theory.The GPC system is compared with other systems build before by analyzing the critical frequency character,the frequency responses and the computational complexity from which we can learn the control performance and system robustness.In the end we research in the secondary frequency modulation performance of combined cycle power plant by testing the system dynamic performance in an extremely abnormal load demand disturbance situation.From simulation results we can know that systems using predictive control methods display a remarkable responses performance and powerful stabilization capacity compared to those using conventional control methods and GPC system has the best control performance and system robustness.