Integration Learning Vector Quantization-Perceptron in the Mechanism Battery Charging Decision Model in AC-DC Hybrid Electrical Installations at Plant Factory

Abstract

This study aims to optimize plant growth and development through the integration of learning vector quantization (LVQ) and a single-layer perceptron (SLP) for battery charging in the plant factory to optimize plant growth and development. During this analysis, the plant factory transmitted the concept of smart energy by adopting the relationship between the production and consumption of electrical energy of the Solar Power Plant (SPP) based on direct current (DC) and the service line (SL) following Alternating Current (AC). This combination was for charging the battery in the plant factory, and SPP had a limited time for electrical energy production due to the duration of sunlight. The scenario of sunlight duration and battery charge status (SOC_ES) in the three LVQ classifications, both SPP and battery, was input to the SLP for battery charging decisions. Charging scheduling was determined based on the time-duration relationship scenario between the SPP supply classification (SPP_Supl) and the state of charge (SOC_ES) of the battery. Perceptron decided the SOC on the battery and used the SPP and/or AC network. The results of models showed that the on-grid electrical energy supply process occurred when the state of consumption (0) was detected and applied to all battery clusters or SPP_Supl > SOC_MAX values. In addition, the LVQ-SLP integration model produced accurate simulation results.