A Review of Mathematical Models for Trading Decision-Making in Electricity Markets

As electricity market reforms progress worldwide, power producers, storage operators and retailers face complicated trading challenges driven by massive renewable penetration, multi-time-scale market coupling and frequent extreme weather. Traditional deterministic optimization cannot handle complex uncertainty. Following PRISMA 2020 protocol, this paper systematically screens 129 high-quality Chinese and English studies to summarize mainstream mathematical models for trading decision-making, including stochastic, robust, and distributionally robust optimization, decision-dependent uncertainty (DDU), risk-based models, evolutionary algorithms, and reinforcement learning (RL). Different from existing reviews focused on foreign markets or single algorithms, this paper targets Shandong provincial spot market with locational marginal pricing, dual-settlement deviation rules and over 40% wind–solar installed capacity, quantitatively compares the practicability of seven modeling frameworks for local bidding rules, and elaborates the emerging DDU optimization paradigm. Numerical cases show DDU-DRO improves aggregator profit by 6.8% versus exogenous uncertainty models. We analyze the practical defects of each model under Shandong’s unique trading constraints, summarize unresolved bottlenecks and future research directions, and provide localized theoretical references for domestic provincial spot market intelligent bidding.