Abstract: The decision models for risk-averse V2G reserve considering stochastic demand and revenue sharing under the CVaR measurement criterion are proposed,and the analytical solutions of channel members' optimal decision behavior in the integrated and decentralized decision are derived. On this basis,the equilibrium strategies when the random demand variable follows the uniform distribution are comparatively analyzed. The research shows that the optimal V2G reserve factor under integrated decision is positively correlated with the channel overall risk aversion,while the relevance between equilibrium selling price of V2G and channel overall risk aversion is uncertain,and this uncertainty is affected by the distribution function of the stochastic demand variable. The optimal V2G reserve factor under decentralized decision is only related to the grid company' risk aversion,while the equilibrium selling price of V2G is affected by the common influence of risk aversion,purchasing price of grid company,and electric vehicle user' revenue sharing coefficient. The optimal V2G reserve revenue sharing coefficient of electric vehicle user is positively correlated with his risk aversion,but negatively related to the risk aversion of grid company. The results of numerical simulation indicate that revenue-sharing contract in the vast majority of cases does not perfectly coordinate decentralized decision behavior in V2G reserve channel.

Key words: V2G reserve, stochastic demand, risk aversion, CVaR, revenue sharing