The invention discloses a bridge overweight vehicle dynamic recognition method based on deep learning. The method comprises the steps of: 1), obtaining bridge structure response data, and marking an overweight vehicle label for each sampling time; 2) constructing and training an overweight vehicle dynamic identification network based on a time convolutional network; and 3) inputting to-be-identified structure response data into the trained overweight vehicle dynamic identification network, and outputting an overweight vehicle identification result. According to the method of the invention, theapplication of a deep learning algorithm in an overweight vehicle identification task in the field of bridge health monitoring is realized; compared with an existing bridge dynamic weighing method, the method of the invention does not need the domain knowledge of a static influence line, vehicle speed and axle load information; the method is suitable for a scene where multiple vehicles appear ona bridge, increase the value of bridge health monitoring data, and has a wide prospect in the field of bridge health monitoring.
本发明公开了一种基于深度学习的桥梁超重车动态识别方法,包括步骤:1)获取桥梁结构响应数据,为每个采样时间标注超重车标签;2)构建并训练基于时间卷积网络的超重车动态识别网络;3)将待识别的结构响应数据输入至训练好的超重车动态识别网络中,输出超重车识别结果。本发明实现了深度学习算法在桥梁健康监测领域超重车识别任务上的应用,相较于现有的桥梁动态称重方法相比,无需静态影响线的领域知识、无需车速和轴重信息,且适用于桥上有多辆车出现的场景,使桥梁健康监测数据产生更大价值,在桥梁健康监测领域有广阔前景。
Bridge overweight vehicle dynamic identification method based on deep learning
一种基于深度学习的桥梁超重车动态识别方法
2020-07-28
Patent
Electronic Resource
Chinese
SNAPSHOT APPARATUS BASED ON BRIDGE OVERWEIGHT MONITORING
European Patent Office | 2023
|