基于时空差异化充电响应半动态混合车流模型的路-电耦合网络协同优化策略OA

With the rapidly increasing penetration of electric vehicles(EVs)in China's transportation system,urban road networks are increasingly characterized by a mixed traffic pattern consisting of both electric and gasoline vehicles(GVs).The dynamic interaction between these heterogeneous traffic flows significantly affects the coupling between transportation and power distribution networks.However,most existing studies focus solely on EV-based modeling or static traffic assumptions,overlooking the influence of GV flows and the spatiotemporal variability of EV charging demand.To address these limitations,this study proposes a collaborative optimization strategy for traffic–electrical coupling networks based on a semi-dynamic mixed traffic flow model incorporating spatio-temporal differentiated charging responses. Firstly,a semi-dynamic mixed traffic flow model is established to capture the temporal evolution of traffic distribution across multiple time intervals.This model retains the computational simplicity of static models while integrating residual flow transfer to reflect dynamic traffic states.The proposed model accounts for the heterogeneity in travel behaviors and charging requirements of EV and GV users.Cumulative prospect theory(CPT)is applied to construct travel utility functions under bounded rationality,capturing individual preferences regarding time cost,congestion levels,and EV battery state-of-charge(SOC).Secondly,a spatio-temporal differentiated EV charging response model is developed.This model introduces the price elasticity coefficient of charging demand to quantify EV users' responsiveness to dynamic electricity prices across different regions and times.The EV traffic flow response is further mapped to the distribution network using a charging load conversion coefficient,which translates charging-related vehicle flow changes into electrical load variations at distribution nodes.To guide EV route and charging behavior,a differentiated pricing mechanism is embedded in the model.EV users are assumed to select routes that include at least one charging station while minimizing generalized cost,incorporating travel time,charging price,and queuing delay.The resultant variation in charging decisions reshapes the spatial and temporal distribution of EV traffic and affects load profiles in the distribution grid. Based on the developed traffic and charging models,a collaborative optimization model is formulated to minimize the total cost of the traffic-power coupled system.The objective includes both traffic-related travel time costs and power system operational costs,such as generation costs,peak-valley penalties,and electricity purchasing costs.The constraints encompass power balance,generator output limits,road capacity limits,and charging station load capacities.To solve the nonlinear optimization problem,the model is transformed into a variational inequality formulation,and an improved method of successive weighted averages(MSWA)is adopted for equilibrium flow assignment.Additionally,a Cross-Entropy radar scanning differential evolution algorithm is used to solve the system-level optimization problem.Simulation studies are conducted on a case system combining a 22-node traffic network and an IEEE 33-node distribution network.Results demonstrate that the proposed strategy significantly improves the spatial-temporal distribution of traffic flows,alleviates road congestion,and reduces EV charging concentration at specific locations.Charging demand is effectively shifted from peak to off-peak periods,enhancing the load balancing of the power grid.Compared with static models,the proposed semi-dynamic approach yields a 5.33%reduction in total traffic cost. In conclusion,this study presents an integrated optimization strategy that combines behavioral modeling,differentiated pricing,and hybrid traffic flow simulation to coordinate the operation of coupled transportation and electrical systems.The framework provides new insights for guiding EV users'behavior and enhancing system-wide efficiency.Future research will explore stochastic extensions of the model to address uncertainties in EV travel patterns,renewable energy generation,and user participation in demand response programs.