Multi-time period operation analysis of coupled transportation and power distribution networks considering self-driving and human-driving behaviors

The rapid development of electric vehicles (EVs) prompts the interaction between transportation networks (TNs) and power distribution networks (PDNs). Meanwhile, the self-driving technology make the interaction process more intelligent. Thus, the human-driving and self-driving behaviors would coexist in a long time and simultaneously influence the operation of the coupled transportation-power distribution networks (TPNs). To depict the TPN operation states, this paper proposes a multi-time period traffic-power flow analysis method, in which a centralized mode and a decentralized mode are established to represent the impacts of self-driving and human-driving behaviors, respectively. For the centralized mode, an upper-level model is built to realize the social optimum of TPNs considering the travel cost of self-driving EVs and the operation cost of PDNs. For the decentralized mode, a lower-level model under stochastic user equilibrium (SUE) principle is proposed to derive the TPN operation state considering the cognitive biases of drivers. The traffic flows and electricity prices are transferred between the two models and then a bi-level traffic-power flow analysis model is formed. A semi-dynamic traffic assignment method is adopted in the bi-level model to depict the dynamic characteristics of TPNs in multiple time periods. Then, an adaptive iterative algorithm is proposed to solve the bi-level model and derive the multi-time period traffic-power flows. Finally, a practical TPN in Shanghai, China is used to verify the effectiveness of the proposed model and the impacts of various parameters on the TPN operation are discussed.