Secondary Disaster Prevention Detection and Early Warning Command System (I)

For disasters such as mudslides and reservoir dam breaches caused by earthquakes, detailed data collection is conducted for key areas and key locations to form automated and intelligent early warning measures, preventing the occurrence of secondary disasters in advance.

Landslide and mudslide detection: in areas prone to danger on mountain slopes, multiple boreholes will be vertically arranged along the mountain trend. Each borehole will have a liquid level sensor deployed at the bottommost end, and several inclination sensors deployed at different depths. Since landslides in this area are mainly caused by rainwater erosion, the groundwater level depth is the primary indicator of landslide danger. Inclination sensors can monitor the movement status of the mountain body.

The main issues with reservoir dam bodies are that internal stress and uplift pressure in the dam body and dam foundation exceed design limits, manifested as cracks, excessive and uneven dam displacement, and water seepage, as well as related dam collapse and surrounding bank slope landslides. The main observation items include deformation, stress, temperature, seepage flow, uplift pressure, and expansion joints.

Deformation monitoring observes the overall or local deformation of the dam through instrumental means, understanding the deformation patterns of the dam under environmental factors such as self-weight, water pressure, uplift pressure, and temperature, and studying trends of cracks, landslides, sliding, and tilting. Deformation monitoring mainly includes: surface deformation, internal deformation, dam foundation deformation, cracks and joints, concrete face deformation, and bank slope displacement.

Mudslide and dam safety monitoring mainly relies on the Internet of Things for daily real-time monitoring, as well as multi-parameter fusion calculation on servers, and intelligent analysis based on big data and knowledge graphs. It also relies on daily manual instrument inspection. Inspection and measurement methods are commanded in real time through an APP, which uploads measurement data from designated locations in real time, with model calculations performed in the system backend. On the basis of data reporting, video surveillance functions are also relied upon to perform AI intelligent analysis of observation data and issue video model alarms; at the same time, the public can report alarm information, leveraging the call center function.

Since sensors are installed in inconvenient areas, Lora IoT is typically used first to transmit sensor data to gateways, where the data is aggregated and then transmitted to the data center via 4G/5G. NBIOT or 4G/5G IoT can also be used. NBIOT generally only transmits detection data, while 4G/5G networks can simultaneously transmit surveillance video and IoT data. When fiber optic networks or twisted pair networks are available, fixed networks or WIFI are preferred for more stable connectivity.

Since Lora transmission is power-efficient, solar cells and rechargeable batteries can be used for power supply, making deployment easy and cost-effective.

All IoT servers of the emergency system must be built on the government's private network private cloud, which facilitates big data collection and organization, and is also conducive to confidentiality and convenient maintenance. It is not convenient to directly use commercial cloud platforms or other specialized cloud platforms.