基于灰色关联度及支持向量机理论的公路工程造价预测研究

Research on Highway Engineering Cost Forecasting Based on GRA and SVM

公路工程造价估算在项目可行性研究、融资方式选择、线路方案优选等方面具有非常重要的作用。但传统的造价估算涉及大量不确定工程特征因素,造价与这些影响因素间呈现出非线性回归关系,无法有效估算出准确结果。为此应用支持向量机这种机器学习方法,根据公路工程造价的各种影响特征和特征指标的构建原则建立工程造价预算指标体系,通过灰色关联理论确定各工程的关联程度,构建支持向量机理论进行造价估算并运用MATLAB实现公路工程建安费预测,结果验证了该方法的有效性。

Highway engineering cost estimation plays a very important role in project feasibility study,financing method selection,route plan optimization,etc.However,traditional cost estimation involves a large number of uncertain engineering characteristics,and there is a non-linear regression relationship between cost and these influencing factors,and accurate results cannot be effectively estimated.Therefore,this paper applies the machine learning method of support vector machine.Firstly,it analyzes various characteristic factors affecting highway engineering cost and establishes engineering cost estimation indicators.Second,it uses grey correlation theory to determine the weight of each indicator.Finally,it constructs support vector machine theory to estimate the cost.The model and the use of MATLAB to realize highway engineering cost prediction,the results show the effectiveness of the method.