Multi-objective electric-carbon synergy optimisation for electric vehicle charging: Integrating uncertainty and bounded rational behaviour models

While aggregating electric vehicles (EVs) through public charging stations to participate in the electricity market (EM) offers a sustainable means of achieving orderly charging amidst transport electrification and EM reforms, designing effective charging strategies faces a chain of challenges: “market signal guidance — multi-stakeholder interest alignment — charging behaviour uncertainty and bounded rationality.” This study develops an electric‑carbon synergy multi-objective charging strategy that integrates market electricity prices and dynamic marginal emission factors (MEFs) while balancing stakeholder interests. More significantly, this strategy incorporates a stochastic charging behaviour model that combines K-means (KM), Kernel Density Estimation (KDE), and Monte Carlo (MC) methods based on extensive charging data. Furthermore, by defining utility functions and decision reference points for users during the charging process, the strategy effectively embeds a bounded rationality charging decision model. The research evaluates four charging strategies, assessing their impact on operator profits, user utility, peak-valley difference ratios, and emissions, with NSGA-III and TOPSIS methods used for multi-objective optimisation. The results indicate that: 1) The KM-KDE-MC model successfully identifies nine typical charging behaviour patterns, accurately simulating stochastic charging behaviours (R² = 0.94). 2) The proposed strategy demonstrates optimal performance in multi-objective balancing, significantly reducing the peak-valley difference ratios (−28 %) and user charging costs (−6.88 %) while maintaining emissions at stable levels and effectively managing losses in user utility and operator profits. 3) Further comparative scenario analysis shows that the distributed photovoltaics and energy storage system (PESS) enhances system flexibility, improving multiple evaluation metrics without compromising user utility. However, neglecting bounded rationality may overestimate optimisation potential and significantly reduce the user charging experience, increasing users' perceived utility loss by 73.65 %. 4) Among different behaviour patterns, the NT mode should be prioritised in regulatory incentives, as it plays a pivotal role in balancing peak-valley differentials and reducing carbon emissions. This research underscores the importance of well-designed charging strategies in advancing EV-grid integration amid market-driven reforms.