Research on combined POL consumption forecast based on bayes adaptive weighting

In the future informatization battlefield with high technology, POL (petroleum, oil and lubrication) consumption is featured with openness, non-linear, dynamic, uncertainty and self-similarity. Based on Bayes, known probability distribution and deduction of observed data, this paper aims to conduct adaptive weighting for adaptive filtration forecast model; Case-Based-Reasoning (CBR) forecast model, and grey-fractal dimension forecast model. So, to form a combined POL consumption forecast model based on Bayes adaptive weighting, optimize the POL consumption forecast model, and improve the forecast precision of POL consumption.