The large-scale integration of renewable energy sources into deregulated power systems introduces significant stochasticity in generation, network congestion, and market operations, leading to pronounced volatility and financial risk in day-ahead electricity prices. Accurate and risk-aware price forecasting is therefore essential for market participants to optimize bidding strategies, manage exposure to price uncertainty, and support sustainable electricity market operation. This paper proposes a multi-objective stochastic market-oriented optimal power flow (MO-SMOOPF) framework for day-ahead electricity price forecasting, explicitly embedding economic efficiency, price formation, congestion effects, and risk management within the physical constraints of the power flow problem.
The proposed framework formulates electricity market clearing as a many-objective optimization problem, simultaneously considering social welfare maximization, producer profit, consumer payment minimization, congestion rent allocation, reserve procurement cost, locational marginal price (LMP) volatility reduction, and conditional value-at-risk (CVaR)-based financial risk mitigation. These objectives are optimized subject to AC power flow equations, transmission capacity limits, generator operating constraints, reserve adequacy requirements, and market settlement balance conditions, ensuring both physical feasibility and financial consistency.
To solve the resulting high-dimensional and nonlinear optimization problem, an improved non-dominated sorting genetic algorithm II (NSGA-II) is integrated with a radial basis function (RBF) neural network. The enhanced NSGA-II incorporates adaptive crossover and mutation strategies, dynamic crowding distance evaluation, and elite retention mechanisms to effectively balance convergence, diversity, and robustness. The RBF neural network is employed to capture the nonlinear and stochastic dependencies between market prices and system states, enabling accurate day-ahead price forecasting under uncertainty.
The proposed approach is validated on the IEEE-118 bus system under stochastic renewable and load scenarios. Numerical results demonstrate superior forecasting accuracy, improved risk-adjusted market outcomes, and enhanced price stability compared with conventional methods. The developed framework provides a robust decision-support tool for financial risk management and sustainable operation of future electricity markets.
