Linear quadratic tracking optimum controller model design to optimize high frequency power supply performance /

Abstract

The performance of the High Frequency Power Supply (HFPS) induction heating system is improved by building an optimum controller to achieve optimum closed loop control, using the Linear Quadratic Tracking (LQT) method. The optimum controller is designed to minimize the difference between the HFPS actual system output and the desired reference signal, while keeping the system control input minimized. The utilization of switching devices in HFPS induction heating system results in high power loss, poor line power factor, and harmful harmonics. In this research, first a continuous-time linear system model is developed to simulate the HFPS induction heating system with a series-parallel resonant load. Second, the LQT optimum controller is developed to achieve optimum closed-loop control in both continuous and discrete time domains to optimize the system performance. Third, a desired pure sinusoidal reference signal is constructed with desired magnitude and frequency by programming devices. Fourth, the desired reference signal and HFPS induction heating system feedback as inputs to the LQT optimum controller are used to simulate the ideal and real system performance with the LQT optimum controller. By using the LQT optimum controller, the system output is forced to track the desired reference signal closely, in both continuous and discrete time domains. The simulation results are compared with the industry test data to confirm the theoretical consideration.