Research on forecast model of heavy metal in sludge based on BP network

Content forecast of heavy metal is emphasized during sludge utilization in agriculture. General forecast algorithms are not competent for the forecast since the relationship between content of heavy metal in sludge and its influences are multidimensional and nonlinear. With influences of both sequence and consequence upon content of heavy metal in sludge considered, a forecast model based on BP network was constructed and trained with sludge data sampled from a municipal wastewater treatment plant. The comparison of forecast performance was made between the constructed and linearity regression algorithms. Result shows that the constructed algorithm has better optimization searching mechanism and is more precise in forecast.