基于GA-BP神经网络的装配式高层住宅成本预测

Research on Cost Forcasting of Prefabricated High-Rise Residential Buildings Based on GA-BP Neural Network

装配式高层住宅成本受诸多因素影响,成本与各个因素之间存在复杂的非线性关系,BP神经网络难以对其进行准确的成本预测。文章提出了一种遗传算法(GA)优化BP神经网络的预测方法,利用GA-BP神经网络模型提高装配式高层住宅成本预测的准确率。通过对GA-BP神经网络模型的权值和阈值进行优化,构建了GA-BP神经网络装配式高层住宅成本预测模型,并以50组装配式高层住宅样本数据为例,分别运用GA-BP神经网络与BP神经网络预测模型进行了对比试验。试验结果表明,GA-BP神经网络预测模型具有较强的稳定性和更高的预测准确率。

The cost of prefabricated high-rise residential buildings is influenced by various factors,and there is a complex nonlinear relationship between the cost and each factor.Thus,it's difficult for BP neural network to accurately predict its cost.Accordingly,a genetic algorithm(GA)is proposed to optimize BP neural network,which utilizes the GA-BP neural network model to enhance the accuracy of cost prediction.In this thesis,the proposed model was constructed by optimizing the weights and thresholds,and the new model and BP neural network model were applied to.50 prefabricated high-rise residential sample data for comparative experiments,respectively.The experimental results suggest that the GA-BP neural network prediction model has both excellent stability and higher accuracy.