Abstract:Significant research is achieved by ensuring output tracking along with accurate identification of uncertain parametric uncertainties. In a novel departure from the existing literature, the need for persistency of excitation condition is eliminated. Recent advances in concurrent learning based adaptive controllers have relaxed the persistency of excitation condition required to achieve exponential tracking and parameter estimation error convergence. This was made possible via the use of additional concurrent learning stacks in the parameter estimation algorithm. However, the proposed concurrent learning components, that is, the history stacks, needed to be filled with "selected" values dependent on the actual system states. Therefore, the previously proposed concurrent learning adaptive controllers required the system to be stable initially for a finite time so that the corresponding history stacks can be filled (finite excitation condition). In this work, motivated to remove the finite excitation condition, a novel desired system state based concurrent learning adaptive controller is proposed. In order to remove the system state dependencies in the controller and estimation algorithms, a filtered version of the dynamics and a novel prediction error formulation have been designed. The overall exponential stability, parameter error convergence and boundedness of the system states during closed loop operations are ensured via Lyapunov based arguments. The main advantages of the proposed method are its dependence on the desired system states and the overall stability results that paved the way in removing the need for finite excitation condition. Numerical studies performed on a two link robotic device are also presented to illustrate the feasibility of the proposed method.

Adaptive Control Based on a Modified Kalman Filter Predictor

Adaptive Extended State Space Predictive Control for a Kind of Nonlinear Systems.

Linear Offset-Free Model Predictive Control: A Minimum-Variance Unbiased Filter Based Approach

An Adaptive Kalman Filtering Algorithm Based on Maximum-Likelihood Criterion

Fuzzy Backstepping Adaptive Controller with a Command Filter

Self-tuning Weighted Measurement Fusion Predictive Control

Subspace-Based Adaptive Predictive Control For A Class Of Nonlinear Systems

An Extended Kalman Filter for Data-enabled Predictive Control

Adaptive state feedback controller with Kalman filter observer for machining process

Adaptive control of A class of nonlinear systems with guaranteed parameter estimation: A concurrent learning based approach

A Finite-Horizon Adaptive Kalman Filter for Linear Systems with Unknown Disturbances

Modular Design of Adaptive Robust Controller for Strict-Feedback Stochastic Nonlinear Systems

Adaptive Fuzzy Tracking Control for Nonlinear State Constrained Pure-Feedback Systems With Input Delay via Dynamic Surface Technique

Tuning Function-Based Adaptive Observer Backstepping Control for Parametric Output Feedback Systems

Modular Design of Adaptive Tracking of Stochastic Nonlinear Systems Based on Generalized Least-Square Algorithm

Adaptive Kalman Filter Based Maneuvering Target Tracking

Adaptive tracking of stochastic nonlinear systems based on generalized least-square algorithm

Optimal adaptive controllers based on least squares algorithms

Adaptive Filtering for Stochastic Systems with Generalized Disturbance Inputs

Multiple Models Direct Adaptive Controller for a Stochastic Non-Minimum Phase System

Detection, Estimation, and Accommodation of Loss of Control Effectiveness