Stability Analysis of Neuro-Adaptive SMC-Based Current Sharing Technique for Multiphase Converters in Battery Charging Applications

Abstract

In the present scenario, battery plays a vital role in most of the domestic and industrial applications as an uninterrupted power source for sensitive loads, storage systems, electric vehicles, etc. Conventional single port chargers are upgraded to multiport circuits in most of the battery charger applications. However, it faces the following problems; 1. Insufficient control over the voltage regulated and current shared among the multiport converters, 2. Transient perturbation and stability control. This paper presents a new hybrid control strategy for an integrated renewable energy load bus. It comprises of Photovoltaic (PV) and Fuel Cell (FC) based source side converters delivering power to the battery chargers connected on the load side. The hybrid control strategy includes; 1. Neural Network (NN) based self-adapting Proportional-Integral (PI) tuner for adjusting the droop resistance values in multiport network and current sharing among the converters, 2. Sliding Mode Controller (SMC) based PWM control for increased voltage regulation with constant current control as well. The hand in hand process of Neuro Adaptive Droop (NAD) - SMC takes care of the proper load sharing, battery charging and a rapid response control system with wide stable region. The proposed self-adapting and self-tracking technique is evaluated in Matlab/Simulink platform and the results were validated in real time prototypical hardware by implementing the control law using dSPACE-MicrolabBox-1202.