Optimal Controller for Steam Boiler Drum Water Level Control Using Neural Network Estimator and LQR

No Thumbnail Available

Date

2018-10

Journal Title

Journal ISSN

Volume Title

Publisher

Addis Ababa University

Abstract

The water level of boiler drum is one of the crucial control parameters for any process industries, which reflects the control of mainly boiler load and feed water indirectly. The increasing demand for rapid changes in industries leads to more uncompromising requirements on the control systems for the processes. A very common control problem elsewhere in steam generation is that of controlling the water level in a boiler drum. Effectively controlling the drum water level in an industrial boiler helps to maximize overall energy efficiency. The problem behind various approaches used to control the drum water level is lack of practical aspects such as online tuning and ease of implementation within a real plant distributed control system and also bad parameter tuning leading to poor level performance. An optimization technique can solve those problems stated by simplifying the design process and allowing easily usage with physical system while satisfying the optimality condition such as speed and accuracy of the response should be within specified limits. In this thesis an optimal controller using a multivariable feedback technique using a neural network state estimator based linear-quadratic regulator control is introduced. The purpose of NN based LQR is to keep the level of the steam boiler drum water at the specified zero reference value which avoids damage of boiler occurs by either overflow of water on the top of drum which affects steam quality or by shortage of water in the drum causes the drum burnt. Controller is designed and simulated on MATLAB/SIMULINK environment and the designed optimal controller is compared with conventional PID controller. Simulation result indicates that PID controller reaches its steady state value of pressure of 5.45 bar and water level oscillates about the set with slightly increasing amplitude. The NN based LQR controller achieves steady state pressure and level value of 5.455 bar and 0.00322 mm respectively. Comparing the two controllers, the proposed neural network based linear quadratic regulator achieved good performance in both steady state response tracking and speed of response.

Description

Keywords

Drum level control, optimal control, linear quadratic regulator, neural network estimator

Citation