Maximizing performance in PHEVs: Intelligent energy management using chaotic monarch butterfly optimization algorithm

Plug-in Hybrid Electric Vehicles (PHEVs) integrate electric propulsion with an internal combustion engine (ICE), providing drivers with the flexibility to operate in electric, gasoline, or hybrid modes. Effective energy management in PHEVs seeks to optimize the utilization of available energy resources to development vehicle performance, efficiency, and driving range while reducing fuel consumption and emissions. This procedure typically includes determining when to engage the ICE for propulsion or battery charging and when to operate the vehicle solely in electric mode utilizing stored battery energy. Advanced control approaches and optimization algorithms are employed to accomplish optimal energy management by considering factors such as driving conditions, vehicle speed, traffic patterns, user preferences, and battery state of charge (SOC). Addressing these complex and nonlinear dynamics often needs mathematical modeling combined with soft computing methods. In this context, this article proposes a new soft computing–based intelligent energy management scheme (SCA-IEM) for PHEVs. The proposed SCA-IEM framework incorporates a Chaotic Monarch Butterfly Optimization Algorithm (CMBOA) with adaptive parameters for the optimal allocation of energy resources and charging stations (CSs). Moreover, the method integrates Model Predictive Control (MPC), supported by a Generative Adversarial Network (GAN), to predict near-future vehicle velocities. The combined utilize of CMBOA for optimal charging station planning and GAN-based velocity prediction constitutes the key innovation of this study. Extensive simulation studies conducted under diverse driving scenarios found that the proposed CMBOA-based energy management method provocatively outperforms conventional rule-based and metaheuristic methods in terms of fuel efficiency, robustness, and overall system performance, achieving an accuracy of 96.98%. These results highlight the efficiency and practical applicability of the proposed technique for next-generation intelligent PHEV energy management systems.