Subspace Data-Driven Control for Linear Parameter Varying Systems

Abstract

In this research, a unique subspace data driven control for linear parameter changing system with scheduling parameters is presented. This control paves the way for investigating the nonlinear system based on the results regarding the linear system that are already known. Only the data matrix is utilized to represent the output prediction value in the future various time instants, while the input-output observation data matrix is used to identify Markov parameters in the form of state space forms. The cost function in data-driven control is then adjusted using the output prediction value. The optimal control input value of this quadratic cost function is solved using a parallel distribution technique, and the algorithm's iterative convergence is thoroughly examined. Finally, the DC motor, whose mass distribution factor is considered to be one linear parameter varying system, is controlled using the suggested subspace data driven control approach.