【摘要】 为满足水土工程深入研究及复杂系统建模优化的需要,针对利用高含沙洪水治沙淤地后土壤养分、颗粒组成变化的复杂性,采用人工智能技术、传统多元逐步回归技术,以分析利用高含沙洪水治沙淤地后土壤养分含量与颗粒组成之间的关系,建立了二者间的关系模型,通过对比分析,两种模型均能利用土壤颗粒组成来预测土壤养分含量,但在精度上,人工神经网络模型要远远好于多元逐步回归模型。更多还原

【Abstract】 To satisfy the need of Water-Soil Engineering Intensive Study and Modeling and Optimization of Complex Manufacturing System,in accordance with the complexity of soil nutrient and particle composition changes after Sand-controlling and Soil-retaining by Hyperconcentrated Water,ANN Model and Traditional Multiple Regression Techniques are adopted to analyze the relationship between soil nutrient and particle composition after Sand-controlling and Soil-retaining by Hyperconcentrated Water and then to establish a relationship model of them.According to comparative analysis,both of these two models can forecast soil nutrient by soil particles.However,ANN Model far better than Stepwise Multiple Regression Model.更多还原