基于模糊神经网络的护坡植被优选

Selection of Optimal Slope Protection Plants Based on Fuzzy Neural Network

以寒区黏土(pH=6.97)、粉质黏土(pH=7.81)、砂土(pH=8.33)3种土质作为试验对象,并分别种植紫花苜蓿、小冠花、无芒雀麦。从植被高度、颜色、均一度、抗病性、抗旱性、抗贫瘠性、复生率、盖度、植株密度、抗拉强度、抗剪强度11个指标建立护坡植被评价体系,利用模糊神经网络理论,构建了植被护坡优选模型并应用。结果表明,黏土上小冠花的护坡效果最好,其次为粉质黏土的无芒雀麦。

Three common soil types, clay （ pH 6.97）, silty clay （pH 7.81 ） and sandy soil （ pH 8.33） were selected as the experimental soil in the cold area （Harbin）. Medicago sativa, Coronilla varia and Bromus inermis were planted in these three types of soil. A system for the evaluation of slope protection plants was established in terms of plant height, color, uniformity level, disease resistance, drought resistance, barren resistance, survival rate, cover degree, density, tensile strength, and shear strength. A model for selecting the optimal slope protection plants was constructed according to the fuzzy neural network theory. Results show that C. varia is the optimal vegetation for slope protection in clay and B. inermis in silty clay.