Mountain torrent disaster early warning information broadcasting terminal

Abstract

The invention discloses a mountain torrent disaster early warning information broadcasting terminal, and relates to the technical field of mountain torrent disaster early warning, and the terminal comprises an information receiving unit which analyzes, collects and verifies multi-source data; the AI operation processing unit classifies and sorts the data, predicts a trend, judges an early warning level and optimizes pushing; the display output unit converts the data into three-dimensional content and displays the three-dimensional content in multiple forms; the sound-light alarm unit adjusts sound and light and broadcasts according to grades; the user interaction unit records behavior data and supports personalized setting, and all the units cooperate to achieve intelligent early warning; according to the invention, multi-source data is processed through the information receiving unit, and the energy quantification and coupling prediction model of the AI unit is combined, so that the dependence of a traditional single data source is broken through, a full-chain intelligent decision is formed, and the prediction accuracy and the grade division scientificity are improved; meanwhile, the early warning form is adjusted through an energy threshold adaptation algorithm, three-dimensional rendering and grading acousto-optic technologies are integrated, the understanding threshold is reduced through multi-modal interaction, the risk avoiding decision time is shortened, and the disaster prevention efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of flash flood disaster early warning technology, specifically a flash flood disaster early warning information display terminal. Background Technology

[0002] With the acceleration of global climate change and urbanization, the suddenness, destructiveness and complexity of flash floods are becoming increasingly prominent. Traditional disaster prevention systems mainly rely on water level and rainfall data from single monitoring stations and trigger warnings through thresholds. However, they lack the ability to integrate multi-source heterogeneous data and find it difficult to accurately capture the dynamic characteristics of disaster energy evolution. At the same time, traditional warning models are mostly based on statistical regression or empirical thresholds, which fail to quantify the coupling relationship of disaster energy indicators and cannot predict the critical damage time of disasters. This results in warning information being delayed and ambiguous, making it difficult to meet the dual requirements of accuracy and timeliness in high-risk scenarios.

[0003] Existing technologies suffer from several shortcomings: First, they have weak data integration capabilities, requiring manual conversion of multi-source heterogeneous data or reliance on dedicated gateways, resulting in low collection efficiency, poor real-time performance, and difficulty in capturing instantaneous changes in disasters; second, their early warning decision-making models are simplistic, triggering only based on water level or rainfall thresholds without considering the impact of human protective measures on disaster energy, leading to a significant deviation between early warning levels and actual risks; third, they lack user interaction mechanisms, with traditional sound and light alarms using fixed frequencies and uniform volumes, failing to meet the perception needs of specific groups, and primarily relying on text-based notifications, lacking three-dimensional, scenario-based, and multilingual broadcasting methods, making it difficult for the public to quickly understand the disaster situation and take evasive action in emergency situations. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a flash flood disaster early warning information display terminal. This terminal integrates information reception, AI computing and processing, display output, audible and visual alarms, and user interaction units. It achieves multi-source data fusion through multi-protocol parsing and high-frequency data acquisition; utilizes AI algorithms for disaster trend prediction and graded early warning decision-making; and intuitively displays early warning information through visualization rendering and multi-format output. The audible and visual alarm unit adjusts the alarm intensity according to the early warning level; and the user interaction unit records user behavior, supports personalized settings, and improves the efficiency of receiving and responding to early warning information.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a flash flood disaster early warning information broadcasting terminal, the terminal comprising: The information receiving unit consists of a multi-protocol parsing module, a high-frequency data acquisition card, and a data verification module. The multi-protocol parsing module is compatible with multiple information sources from hydrological monitoring stations and meteorological platforms, converting data from different protocols into a standardized format. The high-frequency data acquisition card collects disaster and monitoring data, and the data verification module performs anomaly detection and secondary verification on the collected data. The AI ​​processing unit consists of an information classification and prioritization module, a short-term trend prediction module, a graded early warning decision-making module, and a personalized push optimization module. The information classification and prioritization module classifies and prioritizes the received data from multiple dimensions. The short-term trend prediction module combines real-time data to predict short-term disaster trends. The graded early warning decision-making module determines the early warning level and generates a push strategy. The personalized push optimization module adjusts push parameters based on user behavior data. Display output unit: includes a visualization rendering engine and a multi-format output adaptation module; the visualization rendering engine transforms data into 3D scene-based content; the multi-format output adaptation module displays warning information in the form of charts, animations, and subtitles according to the warning level; The sound and light alarm unit consists of a hierarchical sound and light driving module and a voice broadcasting coordination module. The hierarchical sound and light driving module adjusts the volume, flashing frequency and brightness according to the alarm level. The voice broadcasting coordination module realizes multilingual and dialect broadcasting through a TTS engine and supports content customization. User interaction unit: includes a behavior data acquisition module and a personalization settings module; the behavior data acquisition module records user operations and device status and transmits them in encrypted form; the personalization settings module provides a visual interface and supports users to manually set warning parameters.

[0006] Furthermore, in the information receiving unit, the multi-protocol parsing module supports TCP / IP, serial communication, and LoRaWAN transmission protocols. The built-in protocol conversion engine converts Modbus protocol data and custom messages into standardized JSON data. The minimum acquisition interval of the high-frequency data acquisition card is 1 second. The data verification module triggers secondary acquisition verification based on the historical data feature library when the acquired data exceeds the historical extreme value ±30%.

[0007] Furthermore, in the AI ​​processing unit, the short-term trend prediction module calls the real-time water level, rainfall, and meteorological data obtained by the information receiving unit, and substitutes them into the disaster energy chain quantification algorithm to calculate the total disaster energy. The calculation formula of the disaster energy chain quantification algorithm is as follows: ,in, It is the total disaster energy. It is the potential energy of rainwater. It is the deformation energy of the mountain. It is the kinetic energy of water. It is an energy buffer for human protection. , , , These are all weighting coefficients, used to adjust the proportions of rainwater potential energy, mountain deformation energy, water kinetic energy, and human-induced energy buffering in the total disaster energy calculation. They are calibrated and determined based on actual application scenarios and historical data. The numerical range is divided into 5 disaster levels, among which, when At joules, it is classified as level 1; when joule At joules, it is divided into 2 levels; when joule At joules, it is divided into 3 levels; when joule At joules, it is divided into 4 levels; when At joules, it is divided into 5 levels.

[0008] Furthermore, within the AI ​​processing unit, the short-term trend prediction module utilizes real-time water level, rainfall, and meteorological monitoring data acquired by the information receiving unit. It then uses a human-disaster energy coupling prediction algorithm to predict the short-term development trend of flash floods. The calculation formula for this algorithm is as follows: ,in, It is a time parameter related to the short-term development trend of flash flood disasters, representing the predicted time required for the disaster to reach a critical state of destruction from the current moment. It is the critical destructive energy of a disaster. It is the time-varying rate of change of the total energy of the disaster. It is the intervention efficiency coefficient, used to measure the effectiveness of human emergency response measures in intervening in disaster energy. The equivalent energy of human and material resources invested in emergency response includes the peak flood arrival time, water level rise rate, debris flow advance rate, and landslide sliding acceleration. The prediction results are output in the form of time-energy change curves and text descriptions.

[0009] Furthermore, in the AI ​​processing unit, the hierarchical early warning decision module integrates the disaster level, total energy value, and short-term trend prediction results output by the information classification and priority ranking module, and classifies the early warning level according to preset early warning level determination rules: when the corresponding disaster level is 1-2 and the energy change rate is <100 joules / second, it is determined as a blue warning; when the corresponding disaster level is 2-3 and the energy change rate is 100-500 joules / second, it is determined as a yellow warning; when the corresponding disaster level is 3-4 and the energy change rate is 500-1000 joules / second, it is determined as an orange warning; when the corresponding disaster level is 4-5, it is determined as an orange warning; when the corresponding disaster level is 4-5, it is determined as an orange warning. A red alert is issued when the disaster level is ≥1000 joules / second and the energy change rate is ≥1000 joules / second. Simultaneously, a specific push notification strategy is implemented for each level: a blue alert triggers a scrolling text display at the bottom of the display output unit, and the audio-visual alarm unit activates a low-frequency alert sound; a yellow alert triggers an animated warning in the middle of the display output unit, and the audio-visual alarm unit activates a medium-frequency voice broadcast; an orange alert triggers a full-screen semi-transparent warning frame on the display output unit, and the audio-visual alarm unit activates a high-frequency audio-visual linkage; a red alert triggers a full-screen forced pop-up window on the display output unit, and the audio-visual alarm unit activates a continuous audio-visual alarm and a linked, looping voice broadcast.

[0010] Furthermore, within the AI ​​processing unit, the personalized push optimization module deeply analyzes the historical behavioral data collected by the user interaction unit. It then uses a perception energy threshold adaptation algorithm to mine the warning viewing habits and response preferences of different users. The calculation formula for the perception energy threshold adaptation algorithm is as follows: ,in, These are the relevant parameters and indicators for early warning push notifications. It is the minimum disaster energy that users can perceive. It is the user's basic perception threshold. It is the memory decay coefficient. It is the energy-weighted sum of the disasters encountered by users throughout their history. It is the urgency coefficient. The system assesses the urgency of the current disaster. For users who frequently engage in outdoor activities, the system increases the sound volume to 80% of the device's maximum volume and shortens the reminder interval to 3 minutes, based on the tiered warning system. For users with hearing impairments, the system enhances the brightness of the flashing warning to 500 nits, extends the duration of a single flash to 5 seconds, and simultaneously triggers a vibration alert. For elderly users, the system enlarges the font size to 48 points and slows down the scrolling speed of the subtitles to 20 characters per second, enabling personalized adjustments to the warning push parameters.

[0011] Furthermore, in the display output unit, the visualization rendering engine is based on the WebGL framework, which maps the flood evolution path and landslide displacement data to a virtual GIS scene, and displays the disaster situation through dynamic water flow effects and color gradient mountain models.

[0012] Furthermore, in the aforementioned sound and light alarm unit, the graded sound and light drive module achieves stepless volume adjustment from 30 to 120 decibels through PWM technology, the flashing frequency of the warning light group is 1-20 times / minute, and the brightness is 100-500 nits; the TTS engine of the voice broadcasting collaboration module supports multilingual broadcasting and can customize personalized reminders including local place names and names of safe havens.

[0013] Furthermore, in the user interaction unit, the behavior data acquisition module records the user's operation trajectory, response time, and device status when viewing the warning through a front-end script, accelerometer, and light sensor. The data is then transmitted to the AI ​​computing unit after being encrypted with AES-256. The personalized settings module supports touch operation and voice control, allows for stepless adjustment of the warning volume from 0 to 100%, selection of sound, light, vibration, and pop-up warning combinations, and setting of do-not-disturb periods accurate to the minute. The data is stored in a local SQLite database.

[0014] Compared with existing technologies, this flash flood disaster early warning information broadcasting terminal has the following beneficial effects: I. This invention achieves protocol conversion and high-frequency acquisition of multi-source heterogeneous hydrological and meteorological data through an information receiving unit. Combined with the disaster energy chain quantification algorithm and human-disaster energy coupling prediction model of the AI ​​computing and processing unit, it breaks through the dependence of traditional early warning systems on a single data source. The system can dynamically calculate core disaster energy indicators and predict the critical damage time of disasters, forming a full-chain intelligent decision-making mechanism of data acquisition, energy analysis, and trend inference. Compared with traditional statistical models, this solution quantifies disaster risk through the energy dimension and combines it with the efficiency coefficient of human protection and intervention, thereby improving the accuracy of short-term trend prediction and the scientific nature of disaster level classification.

[0015] Second, this invention dynamically adjusts the intensity, display format, and push frequency of audible and visual alarms through a sensing energy threshold adaptation algorithm. At the same time, the system integrates a three-dimensional scene rendering engine and hierarchical audible and visual driving technology to transform the disaster situation into a visualized dynamic model. It also supports customized multilingual broadcasts through a voice broadcasting collaboration module. This multimodal interactive design solves the problems of high understanding threshold and low response efficiency of traditional early warning information. In particular, it shortens the public's decision-making time for disaster avoidance in complex disaster scenarios and improves the overall disaster prevention and mitigation efficiency.

[0016] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 A structural diagram of each unit of the flash flood disaster early warning information display terminal; Figure 2 Flowchart of data transmission process for flash flood disaster early warning information broadcasting terminal; Figure 3 A schematic diagram of the working principle of a flash flood disaster early warning information broadcasting terminal. Detailed Implementation

[0019] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0020] Example 1: Structure and working principle of flash flood disaster early warning information display terminal.

[0021] This embodiment details the hardware connection relationships, module collaboration mechanisms, and core working principles of the flash flood disaster early warning information broadcasting terminal, and further explains how each unit achieves real-time monitoring, intelligent analysis, and accurate early warning of flash flood disasters through data interaction.

[0022] Overall terminal connectivity: The terminal adopts a modular architecture design, with each unit connected to the main control chip via a high-speed data bus, forming a closed-loop system of "data reception - intelligent processing - multi-format output - user feedback," such as... Figure 1 As shown: The multi-protocol parsing module, high-frequency data acquisition card, and data verification module of the information receiving unit communicate with the main control chip through a USB 3.0 interface, and are responsible for transmitting external monitoring data to the terminal. The AI ​​computing and processing unit is integrated into the neural network acceleration module (NPU) of the main control chip. It receives standardized data from the information receiving unit through the internal data bus and transmits the processing results (warning level, push strategy) to the display output unit and the audible and visual alarm unit. The visualization rendering engine of the display output unit is connected to the terminal display screen (supporting 4K resolution) via an HDMI 2.1 interface. The multi-format output adapter module is linked with the GPIO interface of the main control chip to control the format and triggering method of the displayed content. The graded sound and light drive module of the sound and light alarm unit is connected to the speaker (supporting 30-120 dB output) through the audio power amplifier, and connected to the warning light group (adjustable from 100-500 nits) through the LED drive circuit. The TTS engine of the voice broadcasting collaboration module is integrated into the audio processing unit of the main control chip. The user interaction unit's behavior data acquisition module collects user operations through the touch screen controller and accelerometer (I²C interface), and transmits them to the AI ​​computing and processing unit via the internal bus after AES-256 encryption; the personalization setting module interacts with the user through a graphical interface (running on a Linux system), and the setting parameters are stored in a local SQLite database.

[0023] Core working principle: The terminal's workflow revolves around "real-time data flow + intelligent dynamic decision-making," such as... Figure 3 As shown: Multi-source data fusion receiving stage: The information receiving unit, as the data entry point of the terminal, simultaneously connects to the hydrological monitoring station (TCP / IP protocol), the meteorological platform (LoRaWAN protocol), and local sensors (serial communication) through the multi-protocol parsing module. It converts the water level data of the Modbus protocol and the rainfall data of the custom message into JSON format. The high-frequency data acquisition card collects real-time data at an interval of 1 second. The data verification module calls the historical data feature library (such as the extreme values ​​of water level in the same period of the past 5 years). When the collected data exceeds the historical extreme value by ±30% (such as the historical highest water level of 10 meters and the current collected water level of 13.5 meters), a secondary collection verification is automatically triggered to ensure the accuracy of the data.

[0024] AI-powered intelligent analysis and decision-making stage: Information classification and priority sorting: The AI ​​computing and processing unit first classifies the received data into multiple dimensions according to "disaster type" (flood, landslide), "data source" (official monitoring station, local sensor) and "real-time" (data within 10 minutes is marked as high priority), and prioritizes the processing of high priority data (such as real-time flood data from official monitoring stations).

[0025] Short-term trend prediction: The short-term trend prediction module calls real-time water level, rainfall, and meteorological data, and substitutes them into the disaster energy chain quantification algorithm to calculate the total disaster energy. The calculation formula for the disaster energy chain quantification algorithm is as follows: ,in, It is the total disaster energy. It is the potential energy of rainwater. It is the deformation energy of the mountain. It is the kinetic energy of water. It is an energy buffer for human protection. , , , These are all weighting coefficients, used to adjust the proportions of rainwater potential energy, mountain deformation energy, water kinetic energy, and human-induced energy buffering in the total disaster energy calculation. They are calibrated and determined based on actual application scenarios and historical data. The numerical range is divided into 5 disaster levels, among which, when At joules, it is classified as level 1; when joule At joules, it is divided into 2 levels; when joule At joules, it is divided into 3 levels; when joule At joules, it is divided into 4 levels; when In Joules, it is divided into 5 levels; simultaneously, the short-term trend prediction module calls the real-time water level, rainfall, and meteorological monitoring data obtained by the information receiving unit, and predicts the short-term development trend of flash floods through the human-disaster energy coupling prediction algorithm. The calculation formula of the human-disaster energy coupling prediction algorithm is: ,in, It is a time parameter related to the short-term development trend of flash flood disasters, representing the predicted time required for the disaster to reach a critical state of destruction from the current moment. It is the critical destructive energy of a disaster. It is the time-varying rate of change of the total energy of the disaster. It is the intervention efficiency coefficient, used to measure the effectiveness of human emergency response measures in intervening in disaster energy. The equivalent energy of human and material resources invested in emergency response includes the peak flood arrival time, water level rise rate, debris flow advance rate, and landslide sliding acceleration. The prediction results are output in the form of time-energy change curves and text descriptions.

[0026] Tiered early warning decision-making: The tiered early warning decision-making module combines the disaster level (level 3) and the energy change rate (600 joules / second) to determine an orange warning and trigger the corresponding push strategy: the display output unit displays a full-screen semi-transparent warning frame, and the sound and light alarm unit starts high-frequency sound and light linkage (volume 80 decibels, warning light flashing 15 times / minute).

[0027] Personalized push notification optimization: The personalized push notification optimization module analyzes user behavior data (e.g., a user has viewed alerts while outdoors for the past 3 times) and adjusts parameters using a perceived energy threshold adaptation algorithm. The calculation formula for the perceived energy threshold adaptation algorithm is as follows: ,in, These are the relevant parameters and indicators for early warning push notifications. It is the minimum disaster energy that users can perceive. It is the user's basic perception threshold. It is the memory decay coefficient. It is the energy-weighted sum of the disasters encountered by users throughout their history. It is the urgency coefficient. The system assesses the urgency of the current disaster. For users who frequently engage in outdoor activities, the system increases the sound volume to 80% of the device's maximum volume and shortens the reminder interval to 3 minutes, based on the tiered warning system. For users with hearing impairments, the system enhances the brightness of the flashing warning to 500 nits, extends the duration of a single flash to 5 seconds, and simultaneously triggers a vibration alert. For elderly users, the system enlarges the font size to 48 points and slows down the scrolling speed of the subtitles to 20 characters per second, enabling personalized adjustments to the warning push parameters.

[0028] Multi-form early warning output and user interaction stage: The visualization rendering engine of the display output unit is based on the WebGL framework, mapping flood evolution path and landslide displacement data to a virtual GIS scene (such as a 3D water flow animation overlaid with a local map). During a blue alert, scrolling text is displayed in 36-point yellow font on a black background at a speed of 30 characters per second. During a red alert, a forced pop-up window occupying 80% of the screen is triggered, accompanied by device vibration. In the sound and light alarm unit, the graded sound and light drive module adjusts the volume (e.g., 80 decibels for an orange alert) and the brightness of the warning light (500 nits) through PWM technology. The voice broadcasting collaboration module broadcasts in the local dialect through a TTS engine. Users can adjust parameters through the personalization settings module (touch or voice control) (e.g., setting 23:00-6:00 as a do-not-disturb period). The behavior data acquisition module records the operation trajectory and encrypts it to feed back to the AI ​​unit to continuously optimize the push strategy.

[0029] Example 2: An example of a flash flood disaster early warning information display terminal in a mountainous tourist area.

[0030] The information receiving unit is fully operational. Its multi-protocol parsing module is compatible with the serial communication protocols of hydrological monitoring stations around the scenic area, the LoRaWAN protocol of the meteorological department platform, and the TCP / IP protocol of monitoring equipment within the scenic area. Through its built-in protocol conversion engine, it uniformly converts Modbus protocol water level data and custom-formatted rainfall reports from different sources into standardized JSON data. This conversion eliminates format differences between multiple data sources, ensuring efficient subsequent processing. The high-frequency data acquisition card continuously collects monitoring data on rainfall, river levels, and mountain vibration frequency at a minimum interval of 1 second. This high-frequency acquisition captures subtle changes in the data, preventing delayed warnings due to data lag. The data verification module calls upon a historical data feature library to detect anomalies in the collected data. When the 10-minute rainfall in a particular collection exceeds 30% of the historical extreme value for the same period, a secondary acquisition verification is immediately triggered. This verification eliminates errors in data transmission or acquisition, ensuring the reliability of data transmitted to subsequent units. The confirmed data is then transmitted to the AI ​​processing unit. Figure 2 As shown.

[0031] The information classification and prioritization module categorizes the received rainfall, water level, landslide vibration, and wind direction and force data according to multiple dimensions such as "threat level to tourist safety" and "timeliness of data updates," and marks data on short-term heavy rainfall and water level rise rate as the highest priority. This classification and prioritization ensures that the terminal processes the most critical information first, ensuring that core risk data is not overlooked. The short-term trend prediction module calls upon the above real-time data and first calculates the total disaster energy using the disaster energy chain quantification algorithm. The calculation formula for the disaster energy chain quantification algorithm is as follows: This is used to determine the current disaster level; then, the human-disaster energy coupling prediction algorithm is used. The calculation formula for the human-disaster energy coupling prediction algorithm is as follows: The system analyzes the time from the current moment until the disaster reaches a critical state of destruction, predicting that a flash flood may form within the next 1.5 hours, and its destructive power will gradually increase. The generated time-energy change curve and text description serve to provide a scientific basis for subsequent early warning decisions. The tiered early warning decision module integrates the above results: the total disaster energy corresponds to a disaster level of 4, and the time change rate of the total disaster energy is 1200 joules / second, which meets the judgment rules for a red alert. A red alert push strategy is then generated. This strategy clarifies the early warning methods for different terminals, ensuring that early warning information accurately reaches all personnel. The personalized push optimization module analyzes historical behavioral data collected by the user interaction unit and uses a perceived energy threshold adaptation algorithm. For outdoor tourists within the scenic area, the calculation formula for the perceived energy threshold adaptation algorithm is: To ensure timely detection of warnings in noisy environments, the volume of the warning sound has been increased to 80% of the device's maximum volume, and the reminder interval has been shortened to 3 minutes. For hearing-impaired tourists in the scenic area, the brightness of the flashing warning has been increased to 500 nits, the duration of a single flash has been extended to 5 seconds, and the device vibration has been triggered simultaneously, ensuring that they receive the warnings through multi-sensory stimulation. For elderly tourists in the scenic area, the display font size has been enlarged to 48 points, and the scrolling speed of the subtitles has been slowed down to 20 characters per second for easier reading. These personalized adjustments are designed to cater to the needs of different groups and improve the efficiency of receiving warning information.

[0032] The visualization rendering engine of the display output unit is based on the WebGL framework. It maps the real-time collected data on the possible evolution path of the flood and the displacement data of the surrounding mountains onto the virtual GIS scene. Through dynamic water flow effects (such as blue water flow simulating the spread of flood) and color gradient mountain models (such as a gradient from green to red to indicate an increase in the degree of danger), the disaster situation is displayed intuitively. The purpose of this visualization is to allow tourists and staff to quickly understand the development of the disaster and enhance their awareness of the risks. The multi-format output adaptation module forces an 80% screen-sized warning interface to pop up on all terminal screens according to the red warning level. The interface contains information such as the specific location of the disaster, the expected impact range, and the suggested evacuation routes. At the same time, the screen vibrates as feedback. The pop-up and vibration are intended to forcefully attract people's attention through strong visual and tactile stimulation to ensure that the warning information is not ignored.

[0033] The tiered audio-visual alarm unit's drive module adjusts the volume to 120 decibels using PWM technology based on the red alert level, ensuring clear audibility in open areas and buildings within the scenic area. The warning light group's flashing frequency is set to 20 times per minute, with a brightness maintained at 500 nits. This high-frequency, high-brightness flashing transmits emergency signals. This audio-visual setup aims to convey danger information over a wide area through strong auditory and visual signals, reinforcing the sense of urgency in the warning. The voice broadcasting coordination module's TTS engine activates continuous audio-visual linkage, alternately broadcasting warning information in Mandarin and local dialects, with strict synchronization between the broadcasts and the flashing lights. The multilingual and dialectal broadcasts ensure that people with different language habits can understand the warning content, and continuous linkage repeatedly reinforces the warning information, urging personnel to take immediate action.

[0034] The user interaction unit's behavior data collection module uses front-end scripts and the device's built-in accelerometer and light sensor to record the operation trajectory of tourists and staff when viewing warning information (such as clicking the screen to confirm, swiping to view details), response time (the time from the warning being issued to the person operating the device), and device status (such as whether the volume and brightness are manually adjusted). This data is transmitted to the AI ​​computing and processing unit after being encrypted with AES-256. The purpose of encrypted transmission is to protect user privacy and data security, while the collected data serves to provide a basis for the subsequent personalized push optimization module to continuously improve the push strategy. At the same time, the personalized settings module provides a visual interface, allowing tourists and staff to manually adjust the warning volume (such as some tourists adjusting the volume from 120% to 90% to avoid being too loud), select the "sound and light + vibration" combination warning mode, or set a temporary do-not-disturb period (such as staff setting a 5-minute do-not-disturb period during evacuation to avoid interfering with command) via touch or voice control. All setting data is stored in a local SQLite database. The purpose of personalized settings is to allow users to adjust the warning mode according to their own needs, improving the user experience without affecting the overall effect of the warning.

[0035] In summary, the flash flood disaster early warning information display terminal of the present invention, in mountainous tourist scenic areas, efficiently processes multi-source data through the information receiving unit, ensuring the foundation for subsequent analysis; the AI ​​computing and processing unit accurately classifies, predicts, and generates personalized strategies, improving the scientific nature and pertinence of the early warning; the display output and audible and visual alarm units transmit information in various forms, ensuring timely perception by personnel; and the user interaction unit records data and supports personalized settings, optimizing the experience and privacy security. All units work together to achieve efficient early warning throughout the entire process, providing strong support for disaster prevention and mitigation.

[0036] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A terminal for displaying and broadcasting early warning information for flash floods, characterized in that, The terminal includes: The information receiving unit consists of a multi-protocol parsing module, a high-frequency data acquisition card, and a data verification module. The multi-protocol parsing module is compatible with multiple information sources from hydrological monitoring stations and meteorological platforms, converting data from different protocols into a standardized format. The high-frequency data acquisition card collects disaster and monitoring data, and the data verification module performs anomaly detection and secondary verification on the collected data. The AI ​​processing unit consists of an information classification and prioritization module, a short-term trend prediction module, a graded early warning decision-making module, and a personalized push optimization module. The information classification and prioritization module classifies and prioritizes the received data from multiple dimensions. The short-term trend prediction module combines real-time data to predict short-term disaster trends. The graded early warning decision-making module determines the early warning level and generates a push strategy. The personalized push optimization module adjusts push parameters based on user behavior data. Display output unit: includes a visualization rendering engine and a multi-format output adaptation module; the visualization rendering engine transforms data into 3D scene-based content; the multi-format output adaptation module displays warning information in the form of charts, animations, and subtitles according to the warning level; The sound and light alarm unit consists of a hierarchical sound and light driving module and a voice broadcasting coordination module. The hierarchical sound and light driving module adjusts the volume, flashing frequency and brightness according to the alarm level. The voice broadcasting coordination module realizes multilingual and dialect broadcasting through a TTS engine and supports content customization. User interaction unit: includes a behavior data acquisition module and a personalization settings module; the behavior data acquisition module records user operations and device status and transmits them in encrypted form; the personalization settings module provides a visual interface and supports users to manually set warning parameters.

2. The flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the information receiving unit, the multi-protocol parsing module supports TCP / IP, serial communication, and LoRaWAN transmission protocols. The built-in protocol conversion engine converts Modbus protocol data and custom messages into standardized JSON data. The minimum acquisition interval of the high-frequency data acquisition card is 1 second. The data verification module is based on the historical data feature library. When the acquired data exceeds the historical extreme value of ±30%, a secondary acquisition verification is triggered.

3. The flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the AI ​​processing unit, the short-term trend prediction module calls the real-time water level, rainfall, and meteorological data obtained by the information receiving unit, and substitutes them into the disaster energy chain quantification algorithm to calculate the total disaster energy. The calculation formula of the disaster energy chain quantification algorithm is as follows: ,in, It is the total disaster energy. It is the potential energy of rainwater. It is the deformation energy of the mountain. It is the kinetic energy of water. It is an energy buffer for human protection. , , , These are all weighting coefficients, used to adjust the proportions of rainwater potential energy, mountain deformation energy, water kinetic energy, and human-induced energy buffering in the total disaster energy calculation. They are calibrated and determined based on actual application scenarios and historical data. The numerical range is divided into 5 disaster levels, among which, when At joules, it is classified as level 1; when joule At joules, it is divided into 2 levels; when joule At joules, it is divided into 3 levels; when joule At joules, it is divided into 4 levels; when At joules, it is divided into 5 levels.

4. The flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the AI ​​processing unit, the short-term trend prediction module calls upon real-time water level, rainfall, and meteorological monitoring data obtained by the information receiving unit, and predicts the short-term development trend of flash floods using a human-disaster energy coupling prediction algorithm. The calculation formula for the human-disaster energy coupling prediction algorithm is as follows: ,in, It is a time parameter related to the short-term development trend of flash flood disasters, representing the predicted time required for the disaster to reach a critical state of destruction from the current moment. It is the critical destructive energy of a disaster. It is the time-varying rate of change of the total energy of the disaster. It is the intervention efficiency coefficient, used to measure the effectiveness of human emergency response measures in intervening in disaster energy. The equivalent energy of human and material resources invested in emergency response includes the time to peak flood arrival, the rate of water level rise, the rate of debris flow advance, and the acceleration of landslide sliding. The prediction results are output in the form of time-energy change curves and text descriptions.

5. A flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the AI ​​processing unit, the hierarchical early warning decision module integrates the disaster level, total energy value, and short-term trend prediction results output by the information classification and priority sorting module, and classifies the early warning level according to the preset early warning level determination rules: when the corresponding disaster level is 1-2 and the energy change rate is <100 joules / second, it is determined as a blue warning; when the corresponding disaster level is 2-3 and the energy change rate is 100-500 joules / second, it is determined as a yellow warning; when the corresponding disaster level is 3-4 and the energy change rate is 500-1000 joules / second, it is determined as an orange warning; when the corresponding disaster level is 4-5, it is determined as an orange warning. Furthermore, when the energy change rate is ≥1000 joules / second, it is judged as a red alert; at the same time, a special push strategy is formulated for each level: a blue alert triggers the scrolling text at the bottom of the display output unit and the sound and light alarm unit starts a low-frequency prompt sound; a yellow alert triggers the animated warning in the middle of the display output unit and the sound and light alarm unit starts a medium-frequency voice broadcast; an orange alert triggers the full-screen semi-transparent warning box of the display output unit and the sound and light alarm unit starts a high-frequency sound and light linkage; a red alert triggers the full-screen forced pop-up window of the display output unit and the sound and light alarm unit starts a continuous sound and light alarm and links to a looping voice broadcast.

6. A flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the AI ​​processing unit, the personalized push optimization module deeply analyzes the historical behavior data collected by the user interaction unit, and mines the warning viewing habits and response preferences of different users through the perception energy threshold adaptation algorithm. The calculation formula of the perception energy threshold adaptation algorithm is as follows: ,in, These are the relevant parameters and indicators for early warning push notifications. It is the minimum disaster energy that users can perceive. It is the user's basic perception threshold. It is the memory decay coefficient. It is the energy-weighted sum of the disasters encountered by users throughout their history. It is the urgency coefficient. It represents the energy urgency of the current disaster, particularly for users who frequently engage in outdoor activities.

7. A flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the display output unit, the visualization rendering engine is based on the WebGL framework, which maps the flood evolution path and landslide displacement data to the virtual GIS scene, and displays the disaster situation through dynamic water flow effects and color gradient mountain models.

8. A flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the aforementioned sound and light alarm unit, the graded sound and light drive module achieves stepless volume adjustment from 30 to 120 decibels through PWM technology, the flashing frequency of the warning light group is 1-20 times / minute, and the brightness is 100-500 nits; the TTS engine of the voice broadcasting collaboration module supports multilingual broadcasting and can customize personalized reminders including local place names and names of safe havens.

9. A flash flood disaster early warning information broadcasting terminal according to claim 1, characterized in that, In the user interaction unit, the behavior data acquisition module records the user's operation trajectory, response time, and device status when viewing the warning through the front-end script, accelerometer, and light sensor. After AES-256 encryption, the data is transmitted to the AI ​​computing and processing unit. The personalized settings module supports touch operation and voice control, and can adjust the warning volume steplessly from 0 to 100%. Users can select a combination of sound, light, vibration, and pop-up warning methods, set a do-not-disturb period accurate to the minute, and store the data in a local SQLite database.

Images