Neural Evidence Integration Model and Its Application

Abstract

The oilfield remaining oil distribution forecast is called world-level difficult problems by oil domain specialists in the world. The source of low forecast correctness are only consider objective evidences or subjective evidence, so the forecast results still exist limitation, it result in low accuracy, reliability and so on to identify the classification characteristics and to compute quantitative parameters. So, how to fuse all objective evidences and subjective evidences is a key problem to research remaining oil distribution. A new model is proposed, it integrated BP neural networks combination models and two-level D-S evidence reasoning models, the exact classification results are implemented about many remaining oil distribution characteristics. The classification output reliability of each BP network and the reasoning result reliability of each domain fuzzy expert system are regarded as basic probability assignment of input evidence in D-S evidence reasoning model. The model has applied successfully in Daqing Oilfield of China.