Competitive Electricity Pricing with Forecast Accuracy Control in Renewable Microgrids: A Stackelberg Game Approach

Document Type : Research Paper

Authors

1 Department of Industrial Engineering, NT.C., Islamic Azad University, Tehran, Iran.

2 ¹ Department of Industrial Engineering, NT.C., Islamic Azad University, Tehran, Iran.

10.22054/jiee.2026.95099.2215
Abstract
With the growing penetration of renewable energy in microgrids, accurately managing demand uncertainty and pricing decisions has become essential. This study develops a three-layer framework for a single-channel electricity supply chain in which the microgrid acts as the leader and the distribution network operator acts as the follower. In the first layer, baseline demand is forecast using multi-layer perceptron, long short-term memory, random forest, and gradient boosting models. In the second layer, forecasting accuracy and error stability are evaluated through performance metrics (including R-squared of 0.98, MAPE of 0.2%, RMSE of 1.2, and MAE of 0.8) and individual-moving range and cause-selecting control charts. The results indicate that the cause-selecting control chart achieves a faster shift detection with an average run length of 1.8 at a shift of 3. In the third layer, the selected forecast is incorporated into a Stackelberg game to determine optimal wholesale and retail prices, yielding equilibrium profits of 22620.89 for the microgrid and 4416.91 for the distribution network operator. The results identify long short-term memory as the superior model, providing the lowest forecasting error, more stable control limits, and reduced pricing-related decision risk.

Keywords

Subjects


Articles in Press, Accepted Manuscript
Available Online from 18 November 2026

  • Receive Date 25 August 2026
  • Revise Date 06 November 2026
  • Accept Date 14 November 2026
  • First Publish Date 18 November 2026
  • Publish Date 18 November 2026