基于GA-BP神经网络的二河闸病害识别方法

Disease Identification Method of Erhe Sluice Based on GA-BP Neural Network

针对二河闸工程不安全因素和病害识别的问题,在深入分析该工程特点的基础上,利用神经网络理论,对其病害识别模型和主要不安全因素提取方法进行研究,分析二河闸病害识别的特点,结合遗传算法,建立基于GA-BP神经网络的水闸病害识别模型。同时抽取二河闸工作桥上50个碳化样本,对其碳化病害进行预测和识别,验证了二河闸病害识别模型的可行性。

Aiming at the problem of unsafe factors and disease identification of Erhe sluice project,on the basis of in-depth analysis of the characteristics of the project,the disease identification model and the extraction method of main unsafe factors are studied by using neural network theory,the characteristics of Erhe sluice disease identification are analyzed,and combined with genetic algorithm,a sluice disease identification model based on GA-BP neural network is established.Fifty carbonization samples were taken from the working bridge of Erhe sluice,and the carbonization diseases were predicted and identified,and the feasibility of the disease identification model of Erhe sluice was verified.