Multi-source information perception-driven dynamic energy management and charging decision framework for regional integrated energy system with cold chain logistics

In the context of the global electrification and low-carbon transformation of cold chain logistics, the high dynamics of vehicle refrigeration demand, the random charging and discharging behavior of electric vehicles, and the volatility of distributed photovoltaic systems present significant real-time coordination and scheduling challenges for regional energy systems. To address these issues, this study proposes a multi-source information perception-driven dynamic energy management and charging decision framework that incorporates a three-tier vehicle-park-network collaborative structure for cold chain logistics. First, energy consumption at the source is reduced through photovoltaic direct supply and multi-temperature zone collaborative control. Second, a real-time dynamic electricity pricing mechanism for V2V, based on bilateral auctions and iterative compromise with a forced transaction rule, is designed to establish a decentralized internal fleet electricity market, providing real-time price signals that reflect localized supply-demand conditions. Building on these signals, a multi-source information matrix charging decision model is developed, which couples both objective and subjective factors to achieve the co-optimization of system economics and user satisfaction. Additionally, an improved multi-objective dhole optimization algorithm is proposed, introducing scent-marking directed initialization, divided encirclement hunting, and a fatigue factor adaptive switching mechanism, significantly enhancing the efficiency and accuracy of solving high-dimensional nonlinear optimization problems. Simulation-based case studies indicate the potential effectiveness of the proposed framework and algorithm: the photovoltaic direct supply and V2V collaborative mechanism reduces the total system charging cost by 27.0% and increases user satisfaction by 20.4%; the multi-source information decision model further improves the average user satisfaction to 0.7732, with a 25.5% improvement in service targeting. These results provide a useful reference for future practical deployment and field validation of cold chain logistics energy management systems.