Multi-Objective Hierarchical Optimization Framework for Vehicle-to-Vehicle Trading Integrating Hybrid Deep Learning and Dynamic Greedy Matching

Accelerated electric vehicle (EV) adoption imposes complex requirements on grid integration and energy dispatch. Current Vehicle-to-Vehicle (V2V) trading research frequently utilizes monolithic forecasting architectures that fail to account for the stochastic nature of mobility data. Furthermore, traditional optimization strategies often prioritize financial yields at the expense of user-centric utilities, hindering global system optimality. To resolve these limitations, this paper proposes a hierarchical optimization framework, designed to reconcile the interests of stakeholders. The approach first employs a hybrid deep learning architecture, integrating long short-term memory (LSTM), gated recurrent unit (GRU), and Transformer architectures, dynamically weight predictions and refine available dwell time estimations. Then, a multi-objective optimization model is formulated to identify Pareto-optimal solutions that balance economic efficiency with user convenience. Finally, a dynamic greedy matching algorithm is introduced to facilitate rapid transaction pairing for large-scale, real-time V2V requests under multiple constraints. Simulation results demonstrate that this hierarchical framework improves trading success rates, optimizes resource distribution, and enhances overall user satisfaction.