基于PSO-BP神经网络算法的设计施工耦合

Design and Construction Coupling Based on PSO-BP Neural Network Algorithm

设计与施工的交互耦合分析对建设工程的发展具有极其重要的研究意义。这种交互耦合关系能从客观上反映出两个系统的耦合程度和耦合协调发展态势。为了较科学地研究两者的耦合态势情况,文章建立了设计与施工两个系统交互耦合的数学模型,并系统性分析了2010—2015年设计与施工耦合度、耦合协调度曲线变化情况。研究表明,设计与施工过程处于高水平耦合阶段,且两系统的整体发展水平表现不均衡。在此基础上利用PSO-BP神经网络算法对2016—2017年的交互耦合态势进行预测,选取了1996—2015年建筑业设计、施工相关数据作为训练样本,从中均匀抽取4组作为测试样本。该测试结果得到的预测值与实际值符合程度比较好,且预测精度较高。最后文章给出了设计与施工两个系统在处于高水平耦合阶段下,提高耦合协调度的建议。

The interaction analysis between design and construction is of great significance to the development of construction project. This interactive coupling relationship can objectively reflect the coupling degree of two systems and the coordinated development trend. In order to study the coupling situation between the two systems more scientifically,a mathematical model of interaction between design and construction system is established and the changes of design and construction coupling degree and coupling coordination degree from 2010-2015 are systematically analyzed. The research shows that the design and construction process are at a high level of coupling stage,and the overall development level of the two systems is not balanced. Based on this,the PSO-BP neural network algorithm is used to predict the interaction coupling trend of 2016-2017,and the relevant data of the design and construction of the construction industry are selected as the training samples from 1996 to2015,and four groups are selected as the test samples. The network output value obtained by this algorithm is in good agreement with the actual value curve,and the coupling trend of 2016-2017 is accurately predicted. Finally,some suggestions are given to improve the coupling coordination degree in the stage of high level coupling of design and construction.