A Robust Integrated Estimation of Unknown Disturbances, Parameters and States for Linear Time-Varying Systems

In the industrial fields, widespread disturbances often negatively impact system control performance. Accurately estimating the unknown system dynamics and disturbances is the key for improving the system control performance. Traditional estimation methods typically require prior knowledge of model parameters and disturbance dynamics. However, due to practical constraints, accurate prior information is difficult to obtain. Facing such a practical demand, in this article we propose a novel approach for multiinput multioutput time-varying systems to estimate the system states, parameters, and disturbances simultaneously. More specifically, a robust integrated estimation approach is designed to achieve this goal by cooperatively regulating all estimation errors. The only requirement for the design of the proposed integrated estimation approach is the measurable data from system operations. Besides, the solid theoretical results, experiments on a numerical example and application to a quadrotor under payload disturbance are carried out to validate the effectiveness of the proposed approach.