Performance assessment of adaptive controller for switched systems based on tensor approach

This paper mainly concentrates on the robust multiple model adaptive controller performance assessment for switched systems via using the tensor approach. The multiple model adaptive control scheme is employed for the switched control systems performance assessment. The non-negative tensor factorization model is proposed to analyze the tensor data, which stems from uniqueness of low-rank decomposition of higher-order tensor. The data-driven algorithms based on tensor space approach are derived for the calculation of performance measures by applying non-negative tensor factorization. It is shown that the controller performance can be obtained by the data processing method of nonnegative tensor factorization. Under some sufficient conditions, such as a strong finite time switching, or a finite number of the dynamical subprocess, the closed-loop subprocess controller performance can be improved obviously for multi-variate switched systems. Finally, a simulation example is presented to demonstrate the effectiveness of the proposed tensor approach by comparing the adaptive switching controller with other adaptive schemes or the single controller.