LAUSR

Abstract Bed temperature control, being of great significance for the safety, durability and NOx reduction of circulating fluidized bed (CFB) boiler, is nowadays becoming increasingly challenging due to the participation of CFB boiler in load regulation and the existence of various uncertainties such as the feed fuel quality variation and other unknown disturbances. To this end, this paper develops an engineering-friendly multi-model predictive sliding mode control (MM-PSMC) strategy for bed temperature regulation to attain a better adaption to wide load variation. A second-order plus time-delay (SOPDT) model based predictive sliding mode controller (PSMC) is designed to deal with the time-delay, model parameter uncertainty and external disturbances, and taken as local controller in MM-PSMC. An improved hysteresis switching logic is introduced to select a controller that is most suitable for the current operating condition and avoid system chattering caused by fast and unnecessary switching of controllers among local models. Simulation results on a 220t/h CFB boiler demonstrate the merits of the proposed MM-PSMC strategy over a wide operating range, depicting a promising prospect in industrial applications.