HOLISTIC STRATEGIES FOR REDUCING ENERGY COSTS AND MANAGING BATTERY SOC IN ELECTRIC VEHICLE INTEGRATION

Abstract

Background & Scope: The transition to electric vehicles (EVs) represents a significant shift addressing environmental concerns from fossil fuel consumption. This transition brings about increased demand for EV capacity and growth opportunities. This study focuses on achieving an optimal architectural configuration for minimizing energy costs, meeting residential and EV load requirements, and maintaining battery SOC. Three scenarios are examined: Grid to Home and EV, EV to Home and Grid to EV, and Grid to Home and Home to EV. Linear and genetic algorithms are employed for optimization, implemented through PYTHON programming.

Methodology: The study employs linear and genetic algorithms to optimize energy costs and SOC maintenance. Implementation is carried out in PYTHON programming, followed by simulations to assess algorithm performance. Three scenarios are explored, and statistical analysis is used to evaluate the models' predictive effectiveness.

Results: Optimized solutions outperform unoptimized systems in terms of energy costs across all scenarios. Linear algorithm cases show energy costs of $20.65, $20.67, and $20.62, while the genetic algorithm achieves lower costs at $19.16 for all scenarios. Unoptimized systems result in significantly higher energy costs ($24.67, $24.62, and $24.67). The optimization algorithms maintain SOC levels precisely around the target value of 0.9. Genetic algorithm consistently achieves higher average SOC values: 0.8966, 0.8967, and 0.8972 for respective cases.

Conclusion: Statistical analysis highlights the models' predictive efficacy despite negative R-squared values, indicating the potential for improvement. The models' low error values underscore promising predictive performance, with Case 3 consistently showing the smallest errors. Despite limitations, such as negative R-squared values, model improvement and extension to more advanced battery models are suggested. This study demonstrates that intelligent optimization strategies effectively reduce energy costs while maintaining desired SOC levels. Further exploration of advanced battery models, algorithm refinement, and real-world data integration could enhance predictive accuracy and system performance. This research contributes to energy-efficient residential and EV systems and the broader goal of sustainable energy management.