Comparative Analysis of Machine Learning-Based Estimation of Output Current Ripple in PFC-IBC Used in Electrical Vehicle Battery Chargers concerning LR, RF, and ANN Methods

Electrical Engineering

In this study, an artificial neural network (ANN) model is developed based on the inductance current ripple, switching frequency, and load changes to estimate the output current ripple of a power factor correction (PFC) AC/DC interleaved boost converter (IBC) used in battery chargers of electrical vehicles (EVs). Additionally, the enhanced ANN model is contrasted with a few other machines learning (ML) methods, such as random forest (RF) and linear regression (LR). To estimate the output current ripple, the PSIM simulation programme is used to simulate the PFC-IBC. Consequently, 336 output current ripple values are calculated using various switching frequencies, load variations, and inductance current ripple. Next, to manage the current harmonics obtained from the grid and ensure dependable battery charging, the output current ripple value is approximated by training the input parameters using LR, RF, and ANN machine learning methods (MLTs). It may be observed that the estimation value produced by MLTs is rather consistent with the real value that the simulation produced. Furthermore, the simulation-based study requires several days to yield the estimation findings; in contrast, the estimating process using machine learning techniques can be finished in a matter of minutes. This makes the benefit of MLTs very evident. As a result, this value is highly accurately approximated using MLTs prior to the design of the charging apparatus to keep the output current ripple at a safe level, which is crucial for the charging of batteries in electrical vehicles. Additionally, LR, RF, and created ANN approaches were used in this estimating process are looked at and contrasted independently in the WEKA programme, and it is found that the created ANN model offers superior outcomes than alternative methods.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.