Multi-model-based LQR Control For Super-heated Steam Temperature In Power Plant Boiler

In recent years, the linear quadratic regulator(LQR) has been widely studied and applied by many domestic and foreign scholars. In the control of thermal power plant unit, the super-heated steam temperature is the maximum temperature of the entire thermal unit, super-heated steam temperature control performance directly affects the operational safety and productivity of the entire power plant. Currently, the most commonly used thermal power super-heated steam temperature control method is based on conventional PID cascade control system, but thermal power plant super-heated steam temperature system with a large hysteresis characteristics and dynamic characteristics of large inertia over changes in the unit load and other uncertainties so that this greatly increased the difficulty of the control of superheated steam temperature system. Conventional PID cascade control method is difficult to achieve the satisfactory control effect.This paper introduces the significance and importance of super-heated steam temperature control system and the current research status of super-heated steam temperature control system. Then introduces the control principles and design methods of the linear quadratic regulator and LQR control performance analysis. Analysis and research for the role of LQR controller weighting matrix and the genetic algorithm is used to global optimize the weighting matrices Q and R parameters so as to obtain better LQR controller parameters. Based on the static characteristics exhibited by the super-heated steam temperature with the load changes, we establish a typical load conditions super-heated steam system model and then introduce the multi-model control system on this basis, and introduce multi-model control methods and principles. Using multiple model LQR control strategy, to design and simulate about the super-heated steam temperature of the dynamic characteristics of the mathematical model in the process of liter load. Simulation results show that: multi-model linear quadratic regulator control method for controlling the temperature of the super-heated steam achieved very good control effect, improves the dynamic characteristics of super-heated steam temperature and has better robustness.