Dynamic Multi-Energy Pricing for Building-Integrated Energy Systems under Extreme Weather with Vehicle-to-Building Flexibility: A Stackelberg-Evolutionary Game

Authors:

Abstract
Building-integrated energy systems couple weather-sensitive thermal loads with photovoltaics, storage, and electric vehicles, but existing pricing models rarely combine multi-energy demand response, bounded-rational vehicle behavior, and physical dispatch. This study develops a Stackelberg-evolutionary framework in which an integrated energy system operator optimizes electricity, heating, cooling, and vehicle-to-building (V2B) incentive prices. Building users respond through a comfort-constrained mixed-integer linear program, EV owners evolve among charging, V2B, and idle strategies through an individual-level Monte Carlo game, and follower outputs feed marginal-cost heat-pump dispatch, battery model predictive control, and grid/virtual power plant exchange. The 288-dimensional price vector is solved by sep-CMA-ES and compared with CMA-ES and particle swarm optimization under Winter, Summer, Extreme Cold, and Extreme Heat. Sep-CMA-ES improves the regularized objective by 1.0~5.1% over CMA-ES and 20.5-108.5% over particle swarm optimization. Extreme Cold raises grid purchases from 1816.6 to 2852.9 kWh and eliminates virtual power plant export, whereas Extreme Heat increases photovoltaic generation by 38.8%, export from 1231.6 to 1931.5 kWh, and operator profit from CNY 6735.0 to CNY 8892.5. Ablation results show that the full model raises operator profit by 10.2-15.9% and social welfare across all scenarios, but can increase building bills and cause peak-purchase rebound in cooling conditions. The framework therefore reveals complementary economic and flexibility benefits from dynamic pricing and evolutionary EV response together with weather-dependent grid and storage trade-offs.

This publication has 15 references indexed in Scilit: