Auto Disturbance Rejection Control For Circulating Fluidized Bed Boiler Combustion Process

Characterized by nonlinearity, multivariable coupling, time variation of parameters and large inertia, large time delay, circulating fluidized bed boiler (CFBB) combustion system face a hard work in the control implements. In this thesis, auto disturbance rejection control (ADRC) is proposed to solve the difficulties in this field. ADRC is concerned the advantage of independence on mathematical models, decoupling, adaptive and robust in rejecting. The main task of the thesis is to study and develop parameters tuning and control approaches applied to the characters of circulating fluidized bed boiler combustion system using ADRC. First, deep study of the structure of circulating fluidized bed boiler and technical flow and summarize all kinds of CFBB mathematic models. Predigesting the progress of CFBB combustion system model is thoroughly investigated and a viable control parameter and algorithm of models is developed for practical uses. The theoretical analyses and simulation experiments, set for CFBB nonlinear dynamical systems, guarantee the engineering applications of ADRC.Second, the structure and algorithm of ADRC controller is discussed in detail and modified the structure and parameters tuning of the ADRC applied to the typical large inertia and large time delay characters of first order and second order plants. Time scale was introduced to tune parameters and simplify the ADRC controller of plants in same type. So comparing the new type ADRC to the old one on example of the ball mill system, the result of numerical simulation was presented that new type ADRC is better than the old one applied to the large inertia and large time delay plants with good dynamic performance.Finally, applications of ADRC controller in a practical thermal system are studied in the combustion system of circulating fluidized bed boilers. The control system is based on the CFBB combustion mathematic model and decoupling analysis with related gain matrix, design the structure of ADRC with decoupling matrix and the control parameters of ADRC. Comparing to the traditional PID control and advanced PID neural-network adaptive-tracing control, the study results demonstrate that the ADRC possesses better decoupling control qualities,adaptive adaptive and robust in rejecting in rejecting under time variation of parameters,model uncertainties and strong multivariable couplings.