Forecasting for complex uncertainty systems has always been a very difficult problem, how to make accurate forecasting for complex uncertainty systems has also become a focus of the area fore many researchers. Focusing on the properties of complex uncertainty system, by related knowledge discovery theory and methods with prediction heory and methods. The paper presented a model of integrated forecasting model based on knowledge (IFMK) of multipletargets and multiple-factors, the model utilizes suitable knowledge discovery theory and data mining methods for forecasting. Forecasting method based on IFMK can provide a stable and relatively accurate forecasting results. And analysis of exmaples indicates that the model is valid and practical by forecasting petroleum reservoir, production and demand.