Linear Active Disturbance Rejection Control Of Circulating Fluidized Bed Boiler Combustion System

As a necessary part of modern power supply chain, boiler has been not only required, but also essential, its working efficiency and influence on environment have drawn more and more attention in recent years. Circulating fluidized bed boiler is taking advantage over traditional boiler combustion rate and coal cost. However, this complex process, large lag and inertia have created such a budern for control system at the same time. For active disturbance rejection controller, more parameters have limited its application in engineering practice. The problem was solved with linear active disturbance rejection controller by an analytic method in complex frequency field, however, neither of them can offer a higher estimation accuracy and non-overshoot response in circulating fluidized bed boiler combustion system. To cure the above problems, this paper presents some innovative designs on the basis of traditional linear active disturbance rejection controller, more detailes is as follows:Firstly, because there exist so many disturbances in the circulating fluidized bed boiler conbustion system, especially the non-differentiable disturbance, so an itegral linear active disturbance rejection controller is proposed combining PID with linear active disturbance rejection control in this paper. By adding an integrator into linear active disturbance rejection controller, a tunable parameter is introduced into this system to control the amplitude of disturbances, so that its observer range would be extended as far as possible and the static and dynamic performance would be improved too. Finally, the mathematical analysis and simulations are presented for verifying the effectiveness of this new model.Secondly, in order to solve this problem that active disturbance rejection controller can not satisfy the requirements of observation accuracy, a secondary observer based on the disturbance data modelling of linear active disturbance rejection extended state observer is proposed in this paper. This secondary observer model is constructed while the system is sufferring variable disturbances so as to ensure its validity in the complex and chaotic environments. After adopting the discrete data sampled with least squares vector machine, it is used to predict the disturbance as an secondary observer. The simulation results showed that the observeration precision and dynamic response of linear active disturbance rejection control are improved by adding into the secondary observer.Finally, as a typical thermal object, there exists time-delay in circulating fluidized bed boiler combustion system, its dyanmic process is extended and may appear overshoot. Linear active disturbance rejection control does not depend on the accurate system model, under the condition that the stability of circulating fluidized bed boiler system is satisfied, its excellent compensation for the pure lag makes the impact brought by pure-lag be minimized, so a system with a shorter dynamic response time and non-overshoot could be achieved.