Adaline and Recursive Least Square Error Based Techniques for Submodule Voltage Monitoring for the Cascaded High Frequency AC Link System

Abstract

This paper presents two techniques to monitor the voltage of the submodules of the cascaded high-frequency AC link (CHFAL) system. The first proposed technique is using the adaptive linear neuron (ADALINE) algorithm, and the second proposed technique is using the recursive least square error (RLSE) technique. Both of the two submodule voltage monitoring techniques need only one voltage sensor per phase to measure the output phase voltage. The proposed techniques eliminate the need for submodule voltage monitoring sensors and the associated communication links with the central controller, which makes the proposed techniques appealing to CHFACL systems with large number of submodules. The results are validated through Matlab/Simulink simulation software.