A new control framework promises to cut charging costs and extend battery life for electric vehicle fleets used in mobility-on-demand services such as ride-sharing. The system addresses a fundamental challenge: coordinating when and how to charge shared vehicles while contending with unpredictable customer demand and fluctuating electricity prices.
The framework integrates two key functions—deciding when to charge each vehicle and allocating vehicles to customer requests—into a single optimization problem. Rather than treating these decisions separately, the unified approach captures how battery characteristics change during charging and accounts for the nonlinear physics of battery behavior.
To manage computational complexity, researchers used piecewise linear approximations of battery dynamics combined with advanced mixed-integer linear programming. This allowed the system to run efficiently across multiple vehicles simultaneously rather than relying on slow, centralized processing. A logical switching mechanism determines optimal charging sequences based on real-time factors including expected rental patterns and hourly electricity prices.
Testing against two benchmarks—a conventional charging baseline and a laxity-based method—showed the new approach reduced both charging costs and battery capacity degradation. By charging vehicles when electricity is cheapest and adjusting schedules based on anticipated demand, the system minimizes expensive peak-hour charging while protecting battery health through optimal charge rates.
For fleet operators, these improvements translate directly to reduced operating expenses and longer vehicle lifespans. The distributed computational approach ensures the system scales across large fleets, making it practical for commercial deployment. As electricity prices become increasingly variable due to renewable integration and time-of-use pricing becomes standard, such demand-responsive charging becomes essential for competitive fleet economics. The work bridges academia and industry practice, offering a pathway for ride-sharing platforms to lower costs while supporting grid-friendly vehicle charging patterns.
