基于神经网络的土壤肥力综合评价

Fuzzy comprehensive fertility evaluation based on BP artificial network

基于BP神经网络算法的模糊综合评判法对深州市土壤肥力进行等级评价,结果表明,该方法比主成分分析改进的模糊综合评判法更客观、稳定,适用性强,深州市至少89.1%的耕地肥力处于中下等水平;同时利用GIS的插值技术,生成土壤肥力评价图,为深州市的土壤改良和测土培肥提供一定依据。

Comprehensive evaluation method based on fuzzy BP neural network was used to evaluate soil fertility of Shenzhou.The result showed that the method was more objective and stable and applicability than fuzzy comprehensive evaluation method based on the principal component analysis,89.1% of the arable land in Shenzhou was the lower level of fertility.At the same time,fertility evaluation plans by interpolation of GIS technology was helpful for soil fertility improvement and soil testing.