基于软计算的电力线路工程造价预测模型

采用软计算和数据挖掘技术来解决电力线路工程造价问题,提出了一种工程造价的智能预测模型.首先根据工程造价的专业要求对搜集的实际历史数据进行预处理,并应用包装算法优选属性使数据结构符合建模要求,然后采用基于神经网络集成的模糊系统构建工程造价的预测模型,在建模过程中,利用实验对比方法对模型参数进行了优选.最后在Matlab平台上对该模型进行了仿真,实验结果说明该模型在预测准确性及泛化能力上优于BP网络.模型对于工程建设业主方控制初期的投资规模具有一定的指导意义,也能为电力部门在历史工程的基础上形成客观全面的工程造价经验提供辅助手段.

An Engineering Cost Prediction Model for Power Construction Based on Soft Computing Technology

The article provides an intelligence forecast model for solving the problem of engineering cost of power construction based on soft computing and data mining technologies.According to the requirement of engineering cost prediction,data preprocessing is firstly made for collected data to normalize the real data,and a wrapper algorithm is used for feature selection.Later,a neural network driven Takagi-Sugeno fuzzy system is applied to construct the forecast model,and contrast test is used for character select...