基于BP神经网络的大规模建设项目投资估算方法

A Method for Evaluating the Rationality of Large-scale Construction Project Cost Based on BP Neural Network

目前大规模建设项目投资估算方法的估算结果与实际结果误差较大,估算速率较低。针对此问题,文章引入BP神经网络研究了一种新的大规模建设项目投资估算方法,分析市场要求,建立指数指标矩阵,确定指标和投资之间的关联度,根据关联度计算结果,通过归一化处理得到非线性算子,在BP神经网络内部通过并行处理得到输出值和输出值阈值,通过逆向分析得到连接系数,确定误差数据节点,建立估算模型,通过多次筛选得到特征算子,完成信息提取,与阈值进行对比,判断估算结果的合理性。实验结果表明,基于BP神经网络的大规模建设项目投资估算方法采用并行计算,数据拟合度能够达到99%,与传统WSR投资估算方法相比,估算速率提高了30%以上,与BIM投资估算方法相比,估算速率提高了35%,更适合于实际应用。

At present,the error between the estimation results and the actual results of the large-scale construction project investment estimation method is large and the estimation rate is low.In view of the above problems,this paper introduces BP neural network to study a new large-scale construction project investment estimation method,analyzes the market requirements,establishes the index index matrix,determines the correlation degree between the index and the investment,and calculates the results according to the correlation degree.The nonlinear operator is obtained through normalization processing,the output value and output value threshold are obtained through parallel processing inside the BP neural network,the connection coefficient is obtained through reverse analysis,the error data node is determined,the estimation model is established,the feature operator is obtained through multiple screening,the information extraction is completed,and the threshold is compared to judge the rationality of the estimation results.The experimental results show that the BP neural network-based large-scale construction project investment estimation method adopts parallel computing,and the data fitting degree can reach 99%.Compared with the traditional WSR investment estimation method,the estimation speed increases by more than 30%,and compared with the BIM investment estimation method,the estimation speed increases by 35%,which is more suitable for practical application.