The forecasting of market cleaning price based on TS multi-model

The price forecasting model is built based on TS multi-model theory,and fuzzy C-means (FCM) clustering is introduced to pre-process the data due to the paroxysmal characteristic of MCP.After the sub-model is built in every interval,the problems about selection of data in each sub-model and determination of membership function are addressed.At last,the final output is normalized from all sub-models.The practical price data from California electricity market is simulated and analyzed,and the results show the feasibility of proposed method.