Fuzzy modeling for electrical market price forecasting

In the paper, market clearing price (MCP) short-term forecasting implementation using advanced fuzzy modeling technique is presented. The described approach uses the well-known Takagi-Sugeno-Kang (TSK) fuzzy model with scatter partition structure. A modified structure identification algorithm for the TSK model is introduced in detail. Particularly, since the presented algorithm is computationally simple and partitively accurate, we can set the good initial parameters and build the MCP forecasting fuzzy models rapidly. Retrospective MCP real-world data was used for modeling and testing the TSR model. The results presented in the paper confirm considerable value of the fuzzy modeling based approach in forecasting the MCP