Abstract: To improve the accuracy of clearing price estimation in electricity markets with multiple power sources under source-load uncertainty, this paper presents a day-ahead market clearing price interval prediction model. This model considers the uncertainty of generation-side outputs characterized by multidimensional stochastic distributions and the impact of load-side demand response, aimed at minimizing the total generation cost with the participation of hydropower, thermal, wind, and solar power, based on the physical clearing mechanism of electricity markets and the IEEE 30-bus power system. Its solutions give predictions of the interval of locational marginal price and generation commitments, taking source-load uncertainty into account. Case studies show its improvement on bidding efficiency and generation revenue; compared with traditional point forecasting methods, our new model achieves a 4.09% increase in overall economic efficiency and a 0.47% increase in expected revenue. The results verify its feasibility and effectiveness, and provide reliable decision-making support for generators in formulating stepwise bidding and output strategies in the day-ahead market.

Key words: source-load uncertainty, day-ahead market, clearing price, interval prediction model, IEEE 30-node power system