摘要
合成纤维土作为工程中的常用材料,其造价预算准确性能够影响企业的成本控制。为研究合成纤维土价格模型的效果,在MATLAB软件平台基于BP神经网络建立不同特性的纤维土价格模型。通过将模型预测值与实际值作对比,以相对误差值判断模型构建是否合理。结果表明,BP神经网络预测相对误差均小于10%,准确性较好,能够预测纤维土价格,可以为工程现场应用提供参考。
As a common material in engineering,synthetic fiber soil needs accurate valuation of the cost budget,which can affect the cost control of enterprises.In order to study the effect of the price model of synthetic fiber soil,such models with different characteristics were established based on BP neural network on the MATLAB software platform.By comparing the predicted value of the models with the actual value,the relative error value is used to judge whether the model is reasonable.The results showed that the relative errors of BP neural network are less than 10%in prediction,with remarkable accuracy,which can predict the price of fiber soil,and provide reference for engineering field application.
作者
杨悦
袁小永
Yang Yue;Yuan Xiaoyong
出处
《重庆建筑》
2022年第3期29-32,共4页
Chongqing Architecture
基金
2019年度安徽省教育厅科学研究项目“基于神经网络的合成纤维在无机稳定土经济模型研究”(KJ2019A1189)。