【摘要】 近年来，随着中国基建事业的大力发展，公路桥梁数量与规模不断扩大，随之而来生成了海量监测数据，桥梁健康监测平台作为数据的持有者，在桥梁数据的有效利用方面还有很大提升空间。因此，为了提高桥梁数据的利用率，缓解平台面临的数据灾难问题，实现了一种基于神经网络的桥梁健康监测与预警平台。该平台不仅拥有数据采集、传输、持久化等功能，还支持多种类、多数量的桥梁数据可视化交互与桥梁健康评估功能，而且该平台集成了基于神经网络的桥梁监测数据预测功能，实现了桥梁安全预警。预测功能将监测数据划分为三部分：观察窗口、预警窗口、预测窗口，通过机器学习算法，在预测窗口预测到异常值后，预警窗口的设置可以保证充足的时间(该平台中为24h)进行问题处理。实验验证该平台可以有效地完成桥梁状态监测与安全预警，实现桥梁安全运行。更多还原

【Abstract】 In recent years, with the vigorous development of China’s infrastructure, the number and scale of highway bridges have been expanding, which has generated a huge amount of monitoring and surveillance data, and the bridge health monitoring platform, as the data holder, has much room for improvement in the effective utilization of bridge data. Therefore, in order to improve the utilization of bridge data and alleviate the data disaster problem faced by the platform, a neural network-based bridge health monitoring and early warning platform has been implemented. The platform not only has bridge data collection, transmission, and persistence functions, and supports multiple types and quantities of bridge data visualization interaction and bridge health assessment functions, but also, the platform integrates the bridge monitoring data prediction function based on neural network. The prediction function divides the monitoring data into three parts: observation window, warning window and prediction window. Through machine learning algorithm, the setting of warning window can ensure sufficient time(24 h in this platform) for problem processing after the abnormal value is predicted in the prediction window. The experiment verifies that the platform can more effectively ensure the monitoring and prediction of bridge status, complete the full utilization of data, and ensure the safety of the bridge at the same time.更多还原