Machine-learning-based cross-sea bridge fog monitoring system and application method thereof

Abstract

The invention relates to a machine-learning-based cross-sea bridge fog monitoring system comprising a user interaction module, a cross-sea bridge information remote-sensing extraction module, a remotesensing image processing module, a manual remote sensing image fog-zone identification module, a deep-learning-based fog recognition training module, a real-time fog monitoring module, a fog estimation test module, a bridge fog monitoring information processing module, and an interface service module. In addition, an application method of the machine-learning-based cross-sea bridge fog monitoringsystem includes the following nine steps of digital modeling, remote sensing image processing based on geostationary satellite constellation, remote sensing image processing based on polar orbit satellite constellation, treatment of ground fog and other meteorological elements, artificial fog zone sample labeling, large fog zone identification based on a deep convolutional neural network, fog monitoring information providing, TS scoring., and interface service providing. Therefore, the passing function guaranteeing capability of the cross-sea bridge is realized while the high safety is ensured.

Description

technical field [0001] The invention relates to the field of data analysis systems and analysis methods, in particular to a machine learning-based fog monitoring system for cross-sea bridges and an application method thereof. Background technique [0002] The sea-crossing bridge has brought great convenience to the travel of coastal people, but because of the high cost of the sea-crossing bridge, it needs a high enough traffic volume to effectively recover the cost, and the sea-crossing bridge requires a higher level of traffic than conventional roads security. Heavy fog (or dense fog) is one of the more common disastrous weathers that affect traffic. It has the characteristics of high probability of occurrence, wide range of occurrence, and high degree of harm. It can be seen everywhere all year round. Its harm; in addition, due to the abundant water vapor on the sea surface and the frequent impact of cold air from the south, the sea-crossing bridge is very seriously affec...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the disadvantages that the existing local area fog that occurs in the surrounding sea area of ​​the cross-sea bridge cannot be effectively and accurately predicted, and the relevant departments cannot carry out early warning and road closure control of the cross-sea bridge based on the local area fog forecast. The high-time-resolution geostationary meteorological satellites are combined with high-altitude resolution polar-orbiting meteorological satellites and bridge visibility observations, and observation data such as visibility, temperature, and relative humidity observed by surrounding weather stations. Under the joint action of relevant modules, big data and Machine learning technology can quickly realize the automatic extraction of heavy fog information based on static and polar-orbiting meteorological satellites, and can effectively monitor the sea fog and land fog in and around the cross-sea bridge, and can realize the update frequency of 10-30 minutes according to the demand , can achieve effective monitoring of local sudden cloud fog, and bring convenience to relevant departments for cross-sea bridge fog early warning and road closure control based on local area cloud forecast System and its application method

Images