Abstract
•The proposed control topology is joined implementation of CHA and XGBOOST.•Modelling of CMLI is composed to gain the optimal control signal using the established controller.•The proposed CMLI is composed with the decreased number of switches, diodes and sources.•XGBOOST performs and predicts the most optimal control signals of the CMLI in the online way.
This paper introduces a hybrid control topology to cascaded multilevel inverter (CMLI) for grid connected hybrid system. The proposed control topology is joined implementation of color harmony algorithm (CHA) and extreme gradient boosting (XGBOOST). The modelling of CMLI is composed to gain the optimal control signal using the established controller. The proposed CMLI is composed with the decreased number of switches, diodes as well as sources. The aim of proposed control method is to meet the load power demand as well as to preserve the power regulation or maximal energy conversion of wind as well as solar subsystems. The proposed method strictly prohibits the difference present at output voltage of CMLI. Here, CHA is utilized for evaluating the optimal gain parameters in view of the variety of source currents from the normal value and it can be useful for building up optimal control signal dataset in offline way. In light of fulfilled dataset, the XGBOOST performs and predicts the most optimal control signals of the CMLI in the online way. The resultant control signals are utilized to control the insulated gate bi-polar switches (IGBT) of CMLI. Through, established model is implemented at MATLAB/Simulink working stage as well as implementation is considered by previous techniques. The efficiency for various sources using proposed and existing method is also analysed. The efficiency of PV and wind of the proposed technique are 99.3975% and 91.2138%.Overall, the outcomes of the comparison show the superiority of proposed method and confirm its potential to solve the issue.