BP artificial neural network based on improved methods of surface soil in Jilin City Environmental Quality Evaluation

The application of improved BP artificial neural network method for heavy metals in surface soil in Jilin City Environmental Quality evaluation. First, according to the environmental quality of urban soil classification standards for the use of function interpolation Rand randomly generated 400 pairs of training samples and 200 pairs of test samples, and then apply the “trial and error” to determine the number of hidden layer nodes, the eventual establishment of the structure of 7-8-1 BPANN soil environmental quality assessment model, we can see the final adoption of model checking the model has a high evaluation of accuracy, allowing full assessment of environmental quality can be applied to the soil. Evaluation results showed that the soil in line with national standards of an area of about 22%, in line with national standards of two of soil is about 64% of the region.