| There are many kinds of disturbances in practical control systems.They degrade the performance of the systems.How to reject disturbances is one of the key issues considered in the design of control systems and is a hot topic of research in the control field.As an active disturbance rejection method,the equivalent-input-disturbance(EID)approach has been widely used due to its simple structure,easy physical implementation,and no need to know the prior information of disturbances.However,the current research lacks an analysis of the disturbance-rejection mechanism of the EID approach,which makes it difficult to further improve the disturbance-rejection performance.This dissertation focuses on improving the disturbance-rejection performance of the control system and investigates the improvement measures of the equivalent-inputdisturbance approach to achieve the high-precision control of the system.The main contents and innovations are as follows:1)It develops a high-order-filter-based EID(HEID)approach that improves the disturbance-rejection performance without amplifying the measurement noise.First,a high-order EID estimator is designed using a high-order filter instead of a first-order low-pass filter.Then,the disturbance-rejection mechanism of the HIED approach is analyzed using the transfer function from disturbance to output.After that,the ideal form of this transfer function is designed by frequency domain indicators,and the design method of the high-order filter is derived.Finally,the effectiveness of the HIED approach is verified by numerical simulation and experimental system.2)It develops an {1}-EID approach using the {1} inverse of the input matrix that improves the disturbance-rejection performance without increasing the configuration complexity.First,a new EID estimator is designed using the {1} inverse of the input matrix instead of its Moore-Penrose inverse.Then,the stability and disturbance-rejection performance of the {1}-EID-based control system are analyzed,and the chosen range of the freedom in {1} inverse of the input matrix is given.Finally,the effectiveness of the {1}-EID approach is verified using a Stewart platform attitude control system.3)It develops a sliding-mode-control-based EID(SMC-EID)approach that eliminates the disturbance-rejection error caused by the filter in the EID estimator.First,the specific form of the disturbance-rejection error caused by the filter is analyzed.Then,a sliding mode controller is added to the {1}-EID approach,in which the above error is designed to be the sliding-mode surface.Next,the system design ensures the asymptotic convergence of the sliding-mode surface.Finally,the effectiveness of the SMC-EID approach is verified using a Stewart platform attitude control system.4)It develops a repetitive-control-based EID(RC-EID)approach that completely eliminates the effect of periodic disturbances.First,the inner mode of the periodic signal is put into the disturbance-rejection loop by replacing the conventional EID estimator with a repetitive controller.Then,the RC-EID-based control system is designed using the linear matrix inequality(LMI)method.Then,the selection of the LMI regulation parameters is optimized by a particle swarm algorithm(PSO).Finally,the effectiveness of the RC-EID approach is verified using a servo system. |
