Optimization Design And Application Research Of Active Disturbance Rejection Controller

With the development of the electric power industry, the dimension of controlled plant is becoming bigger and bigger and its structure is being more and more complex, then these reason increase the control difficulty of the system. In order to improve system performance, a variety of advanced control algorithms have emerged, such as adaptive control, predictive control, robust control and decentralized control and so on. Though these algorithms improved the system performance well, they also needed extensive and complex calculations, and it is very difficult to come true in practical application. Therefore, we have to consider the active disturbance rejection control (ADRC) because of its simple structure, strong disturbance rejection ability, compared with other control algorithms. ADRC have made much progress since be proposed in last century eighties and so far it’s still the center problem of control theory research. The paper studies the optimization design and application research of active disturbance rejection controller, with the motivation and improves the performance of the control system.Firstly, on the basis of nonlinear ADRC, this paper firstly combines the nonspecific and the specific immune in immunization theory with the genetic algorithm, and forms an improved immune genetic algorithm (IGA) IGA is effective to overcome the disadvantages of the genetic algorithm which is easy to be trapped into local optimal solution and has low convergence accuracy. Then this algorithm is applied to the variable speed constant frequency wind turbines system ADRC parameter optimization, which achieved very good results.Secondly, owing to the strong-coupling of multivariable systems, this paper firstly adopts the invariance principle to design a dynamic decoupling compensator in order to partly decouple the cross-coupling of the circulating fluidized bed and then to simplify the model of decoupling compensator. Thirdly on the basis of dynamic decoupling, the linear active disturbance rejection controllers (LADRC) are designed separately to further reduce the coupling, for the extended state observer estimates the coupling among variables and external disturbances in real time by taking them as a total disturbance, and the linear state error feedback control law is designed to compensate their combined effect. Finally the simulation is designed to compare the control effect between LADRC and PID control for the circulating fluidized bed boiler’combustion system, then the results show that LADRC can provide better decoupling, stronger robust ability and disturbance rejection.Finally, time-delay is well widespread in real process of production, so this paper firstly combines the Smith predictor with active disturbance rejection control, and proposes Smith-ADRC control scheme. This scheme is very easy to be realized and has excellent adaptability to operating regimes. Secondly, this algorithm is applied to the main temperature system, which proved the effectiveness of this control algorithm. Finally this algorithm is applied to the ball mill pulverizing system, which proved the superiority of this control algorithm.