Intelligent Energy Management for Full-Active Hybrid Energy Storage Systems in Electric Vehicles Using Teaching–Learning-Based Optimization in Fuzzy Logic Algorithms

Electric vehicles (EVs) are a compelling alternative for mitigating CO2-equivalent emissions. In the context of EVs, the architecture and operational efficiency of a hybrid energy storage system (HESS) are pivotal. The present study focuses on a HESS model based on a parallel full-active configuration that integrates a lithium-ion (Li-ion) battery with an ultracapacitor facilitated by two direct current-to-direct current converters. The energy management strategy governing the HESS emerges as a critical element in the overall performance of EVs. Conventionally, fuzzy control strategies have been extensively utilised at the supervisory and high-control levels; these are largely dependent on expert experience. This paper introduces an innovative fuzzy control system that employs teaching–learning-based optimisation (TLBO) in the energy management strategy domain of a HESS in EVs. This intelligent control mechanism aims to optimise various performance metrics – such as energy consumption, lithium battery output current, and peak power. The optimisation focuses on fine-tuning the parameters constituting the fuzzy rule base and membership functions. Simulation results substantiate the adaptability of the TLBO-based fuzzy energy management system in efficiently allocating power across various driving conditions. Comparative analyses are conducted with respect to the power-sharing capabilities of the proposed TLBO-fuzzy (TLBO-F), particle swarm optimisation-fuzzy (PSO-F), and non-optimisation-fuzzy (NO-F) strategies under two distinct driving conditions: the urban dynamometer driving schedule and the European extra-urban driving cycle. The main objective of this research is to extend battery lifetime by minimising both the lithium battery output power and overall energy consumption. In this way, the study aims not only to prolong the lifespan of Li-ion batteries but also to mitigate the associated energy costs for battery charging. Ultimately, the proposed method enhances the power sharing of the ultracapacitor during a span of 1400 seconds. It also confirms the battery ratio power,
${R.P}_{b}^{\Box } $
, is decreased for the proposed method (TLBO-F) to 6.4% compared to the PSO-F strategy and is decreased 35% relative to the NO-F algorithm. The battery’s state of charge is increased to 14% by implementing the TLBO-F method relative to the PSO-F and increased to 30% compared to the NO-F algorithm. This diminishes the charging and discharging rate of the battery during the driving cycle while extending the lifespan of the Li-ion battery and adheres to the ultracapacitor’s state-of-charge constraints.