Multifunctional emergency broadcasting system based on Beidou short message

Abstract

The invention discloses a multifunctional emergency broadcasting system based on Beidou short messages. According to the invention, a voice recognition module in the intelligent voice interaction broadcast module can quickly convert a voice instruction input by a user through a microphone into text information, so that the complexity and possible misoperation of manual input are avoided, and user feedback is more direct and efficient; the natural language processing and demand analysis module performs semantic analysis on the recognized text, accurately extracts the core content required by the user, and associates an emergency keyword library in the system, so that the system can quickly understand the intention of the user; the multi-language speech synthesis module can convert a broadcast text generated by the system into a language or dialect familiar to a user, and the three parts act synergistically, so that the interaction efficiency of the system and the user, the accuracy of demand response and the coverage range of information transmission are remarkably improved, emergency broadcast is upgraded from one-way notification to two-way interaction, and the user experience is improved. And the practicability and the reliability of the system in a complex scene are further enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of emergency broadcasting, and in particular relates to a multifunctional emergency broadcasting system based on Beidou short messages. Background Art

[0002] Multifunctional emergency broadcasting is a public safety communication system that integrates audio playback, information dissemination, remote control, and disaster warning. It is widely used in government agencies, communities, schools, hospitals, transportation hubs, and disaster-prone areas. The system usually consists of a central control platform, a transmission network, terminal playback devices, and multiple input interfaces. It supports multiple audio source inputs, such as microphones, CDs, MP3s, and network streaming media, and has functions such as automatic switching, zone broadcasting, emergency interrupts, and remote dispatch. In the event of a public emergency or natural disaster, the multifunctional emergency broadcasting system can quickly issue authoritative instructions and emergency information, ensuring the timeliness and accuracy of information transmission, and effectively enhancing the public's emergency response capabilities. At the same time, the system can also be used for daily broadcast notifications, background music playback, and conference amplification, achieving multi-purpose, flexible and efficient use of a single device. It is one of the important infrastructures for building smart cities and emergency management systems.

[0003] However, in existing technologies, traditional emergency broadcasts mostly rely on ground communication networks, which are easily interrupted when disasters cause base station damage or insufficient signal coverage in remote areas, resulting in poor information transmission reliability. In addition, most broadcasts are one-way and cannot obtain user reception status or specific demand feedback, resulting in a blind spot of "sent but not knowing whether it was received". Resource allocation often relies on manual experience and lacks real-time correlation analysis between user location and disaster scope and resource points, which is prone to response delays or inaccurate resource allocation. Summary of the Invention

[0004] The purpose of the present invention is to provide a multifunctional emergency broadcast system based on Beidou short messages in order to solve the above-mentioned problems.

[0005] The technical solution adopted by the present invention is as follows: a multifunctional emergency broadcast system based on Beidou short message, the system comprising: an information collection and processing module, a Beidou short message communication module, a geographic information system module, a user terminal module, an intelligent voice interactive broadcast module, a data storage and analysis module and an emergency command and control center module; The intelligent voice interactive broadcast module is internally configured with: a voice recognition module, a natural language processing and demand analysis module, and a multilingual voice synthesis module; The standardized broadcast text output by the content generation unit of the information collection and processing module is connected to the protocol encapsulation unit of the Beidou short message communication module; The position data (such as Beidou positioning signal) output of the user terminal module is connected to the position analysis unit of the geographic information system module for coordinate conversion and position information extraction; The voice feedback input of the user terminal module is connected to the voice recognition module of the intelligent voice interactive broadcast module; The user demand information after natural language processing and demand analysis by the intelligent voice interactive broadcast module is connected to the resource coordination unit of the emergency command and control center module; The speech content generated by the multilingual speech synthesis module is output to the speech playback unit connected to the user terminal module; The analysis results of the data storage and analysis module are connected to the operation monitoring unit and resource coordination unit of the emergency command and control center module to provide data support for decision-making; The command dispatching unit of the emergency command and control center module is connected to the protocol encapsulation unit of the Beidou short message communication module, through which command instructions are sent; the operation monitoring unit of the emergency command and control center module is connected to the status of each module.

[0006] In a preferred embodiment, the information collection and processing module is internally provided with a data access unit, a quality filtering unit, an information integration unit and a content generation unit.

[0007] In a preferred embodiment, the Beidou short message communication module is internally provided with a protocol encapsulation unit, an encryption processing unit, a channel scheduling unit and a receipt verification unit.

[0008] In a preferred embodiment, the geographic information system module is internally provided with a position resolution unit, a spatial association unit, a visualization rendering unit and a data fusion unit.

[0009] In a preferred embodiment, the user terminal module is internally provided with a message receiving unit, a content parsing unit, a multimodal display unit and a feedback interaction unit.

[0010] In a preferred embodiment, the speech recognition module is internally provided with a speech signal preprocessing unit, a feature extraction unit, an acoustic model decoding unit and a domain adaptation language model unit.

[0011] In a preferred embodiment, the natural language processing and demand analysis module integrates the joint optimization objectives of context attention and domain constraints based on the characteristics of limited but parameter-sensitive emergency scenario demand types. The formula is: L=α·L 意图 +(1−α)·L 槽位 +λ·L 领域 ; in: Lintent is the cross entropy loss for intent classification, which is used to optimize the model's ability to discriminate the user's core demand types; The L slot is the conditional random field (CRF) loss for slot filling, which is used to accurately locate and label key parameters in the demand (such as location and material type); L domain is the domain constraint loss, which is calculated by comparing the semantic similarity (such as cosine similarity) between standard instructions in the emergency domain knowledge base and the model output to ensure that the parsing results conform to the professional logic of the emergency scenario; α (0<α<1) is the weight balance coefficient between intention and slot task, which is dynamically adjusted according to the characteristic that intention discrimination takes precedence over parameter extraction in emergency scenarios; λ is the regularization coefficient of the domain constraint, which is used to control the guidance intensity of domain knowledge on model training.

[0012] In a preferred embodiment, the multilingual speech synthesis module is internally provided with a text normalization processing unit, a multilingual phoneme conversion unit, a cross-language prosody generation unit and a speech waveform synthesis unit.

[0013] In a preferred embodiment, the data storage and analysis module is internally provided with a data archiving unit, a structured storage unit, an analysis modeling unit and a result output unit.

[0014] In a preferred embodiment, the emergency command and control center module is internally equipped with an operation monitoring unit, a command dispatch unit, a resource coordination unit, and an interface interaction unit. The operation monitoring unit collects real-time status data from each module, including the "data reception rate" of the information acquisition module, the "number of satellite connections" of the Beidou communication module, and the "number of online user terminals." The data is displayed on the command screen as a dashboard, with green indicating normal operation, yellow indicating a warning, and red indicating a fault. Alarm thresholds are also set.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. In the present invention, the voice recognition module in the intelligent voice interactive broadcast module can quickly convert the voice instructions input by the user through the microphone into text information, avoiding the tediousness of manual input and possible operational errors, and making user feedback more direct and efficient; the natural language processing and demand analysis module performs semantic analysis on the recognized text, accurately extracts the core content of the user's needs, and associates the emergency keyword library in the system, so that the system can quickly understand the user's intention and provide an accurate basis for the subsequent resource allocation and instruction generation of the command center; the multilingual speech synthesis module can convert the broadcast text generated by the system into a language or dialect familiar to the user, solving the problem of information reception barriers caused by language differences in traditional broadcasting, and ensuring that users with different language backgrounds can clearly understand the broadcast content. The synergistic effect of the three significantly improves the interaction efficiency between the system and the user, the accuracy of demand response, and the coverage breadth of information transmission, allowing emergency broadcasting to upgrade from "one-way notification" to "two-way interaction", further enhancing the practicality and reliability of the system in complex scenarios.

[0016] 2. In this invention, the information collection and processing module integrates multi-source data from sensors, government platforms, and other sources to ensure comprehensive information sources. The Beidou short message communication module leverages the advantages of satellite coverage to address the problem of traditional communications being easily interrupted in remote areas and disaster scenarios. Even in areas where mobile phone signals are not available, emergency notifications can be accurately sent to user terminals via satellite channels. The user terminal module supports multi-modal displays such as voice, text, and map annotations, allowing users of different ages and cultural levels to quickly understand information. The feedback interaction function allows the command center to grasp user status in real time, forming a closed loop of "send-receive-feedback", avoiding the problem of "one-way broadcast without knowing the effect", and significantly improving the effectiveness of information transmission.

[0017] 3. In this invention, the system enhances the accuracy of emergency response and the efficiency of resource allocation. The geographic information system module can associate user location with the scope of the disaster and, based on the specific needs reported by users, generate a matching solution for "user location - demand type - nearest resource point." The emergency command and control center module monitors the operating status of each module to promptly detect anomalies. Dispatch instructions can be directly connected to the Beidou communication module for rapid issuance, and the resource coordination function can dynamically adjust the allocation plan based on real-time data. The historical data recorded by the data storage and analysis module can also provide a basis for the optimization of subsequent emergency strategies, making each disaster response more efficient and more in line with actual needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a block diagram of the overall system of the present invention; Figure 2 This is a system block diagram of the intelligent voice interactive broadcast module in the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] Reference Figure 1-2 A multifunctional emergency broadcast system based on Beidou short message, the system includes: information collection and processing module, Beidou short message communication module, geographic information system module, user terminal module, intelligent voice interactive broadcast module, data storage and analysis module and emergency command and control center module; The internal settings of the intelligent voice interactive broadcast module include: voice recognition module, natural language processing and demand analysis module and multi-language speech synthesis module; The standardized broadcast text output by the content generation unit of the information collection and processing module is connected to the protocol encapsulation unit of the Beidou short message communication module; The position data (such as Beidou positioning signal) output of the user terminal module is connected to the position analysis unit of the geographic information system module for coordinate conversion and position information extraction; The voice feedback input of the user terminal module is connected to the voice recognition module of the intelligent voice interactive broadcast module; The intelligent voice interactive broadcast module connects the user demand information after natural language processing and demand analysis to the resource coordination unit of the emergency command and control center module; The speech content generated by the multilingual speech synthesis module is output to the speech playback unit connected to the user terminal module; The analysis results of the data storage and analysis module are connected to the operation monitoring unit and resource coordination unit of the emergency command and control center module to provide data support for decision-making; The command dispatching unit of the emergency command and control center module is connected to the protocol encapsulation unit of the Beidou short message communication module, through which command instructions are sent; the operation monitoring unit of the emergency command and control center module is connected to the status data output end of each module to obtain the operation status of each module in real time.

[0021] The information collection and processing module consists of a data access unit, a quality filtering unit, an information integration unit, and a content generation unit. The data access unit connects to various sensors, including seismic accelerometers and hydrological level gauges, through standardized interfaces. It also connects to government data platforms, including the Meteorological Bureau's disaster warning system, the Earthquake Bureau's monitoring network, and third-party monitoring equipment. It supports multi-protocol data reception, including Modbus, MQTT, and HTTP, ensuring unified input of raw data from diverse sources, including weather cloud maps, seismic wave signals, and water level monitoring values. The quality filtering unit uses sliding window filtering and threshold detection to eliminate outliers, such as extreme values ​​caused by sensor failures. It also uses timestamp verification to eliminate duplicate data and, based on pre-set rules, such as data integrity thresholds to flag missing fields, ensures the authenticity and reliability of information entering subsequent processes. The information integration unit categorizes the filtered data by disaster type (earthquake, flood, or typhoon), and by data dimensions (time, location, and intensity). It then uses domain ontology libraries, such as mapping rules between earthquake magnitude and emergency response level, to establish relationships between data, forming a structured collection of emergency information. The content generation unit is based on a template engine, such as Freemarker, which converts the integrated information into standardized broadcast text. The template contains fixed fields such as disaster type, affected area, and recommended measures, and supports dynamic filling of specific data. For example, Typhoon Haikui will land at 14:00 on the 22nd and affect XX County. Residents are advised to move to disaster shelters to ensure the standardization and readability of the broadcast content.

[0022] The Beidou short message communication module is internally equipped with a protocol encapsulation unit, an encryption processing unit, a channel scheduling unit, and a receipt verification unit. The protocol encapsulation unit converts the emergency broadcast text into the Beidou short message communication protocol, such as the RDSS protocol, and adds a message header containing the sender ID, receiver ID, timestamp, and message trailer, including a checksum, to ensure that the data conforms to the transmission format requirements of satellite communications. The encryption processing unit encrypts the encapsulated message content using an encryption algorithm, such as the national secret SM4 algorithm. The key is dynamically distributed via the Beidou satellite's secure channel to ensure the confidentiality of the information during transmission and prevent unauthorized interception or tampering. The channel scheduling unit selects the optimal satellite forwarding channel based on current satellite coverage, such as the number of visible satellites in the user terminal's area, and communication priority (e.g., emergency broadcasts take precedence over regular notifications). It supports single-satellite point-to-point forwarding and multi-satellite forwarding broadcast modes to improve communication efficiency. The receipt verification unit starts timing after sending the message and receives the confirmation message returned by the user terminal, which includes the original message ID and timestamp. It confirms the integrity of the receipt through verification methods such as hash verification. If the receipt is not received within the time limit or the verification fails, the retransmission mechanism is triggered to ensure the reliability of information transmission.

[0023] The GIS module consists of a location resolution unit, a spatial correlation unit, a visualization rendering unit, and a data fusion unit. The location resolution unit receives Beidou positioning signals from the user terminal, including longitude, latitude, and altitude, or base station positioning data. It converts these signals into coordinates using a map coordinate system, such as WGS-84, and outputs the user's precise location information, e.g., "XX Village, XX Town, coordinates 118.5°E, 28.3°N." The spatial correlation unit spatially overlays the user's location with emergency information, such as the extent of the flood inundation zone and earthquake intensity zones. It utilizes buffer zone analysis, such as generating a 5-kilometer buffer zone centered on the user's location, and spatial queries, such as whether the user is within a red alert zone, to determine the targeted content of emergency broadcasts, e.g., "You are in a flood inundation zone, evacuate immediately." The visualization rendering unit uses map service interfaces, such as the AutoNavi and Baidu Maps APIs, to load base map layers, including roads, rivers, and buildings. It then overlays emergency thematic layers, including the disaster area and rescue point distribution. This intuitive display of geographic information is achieved through color coding (red for high risk, yellow for medium risk), and symbols (triangles for shelters, and crosses for affected areas). The data fusion unit associates GIS data user location, disaster scope, and other module data, such as user feedback on the need for medical assistance, to generate comprehensive emergency response recommendations. For example, if a user is at point A and needs medical assistance, the nearest hospital is at point B, 3 kilometers away, to provide support for subsequent command decisions.

[0024] The user terminal module is internally equipped with a message receiving unit, content parsing unit, multimodal display unit, and feedback interaction unit. The message receiving unit integrates a Beidou short message receiving chip. It captures satellite signals through a directional antenna and parses the encrypted data in the message. This data is encrypted by the Beidou short message communication module and extracts the emergency broadcast payload. The content parsing unit verifies the received text format, such as whether it contains the disaster type and location fields. It then uses a local vocabulary to perform semantic recognition of emergency keywords such as "evacuate" and "rescue," distinguishing between emergency notifications and general reminders, and extracting key information such as the typhoon landfall time and the address of the disaster shelter. The multimodal display unit converts the parsed information into various user-friendly formats. The voice playback unit outputs the broadcast content through a speaker, supporting adjustable speech speed, especially for emergency broadcasts. The text display unit scrolls key information on the terminal screen, such as the disaster shelter "XX Primary School." The image display unit uses the GIS module interface to plot the user's location and the path to the disaster shelter on a map. The feedback interaction unit provides multiple input methods: voice feedback collects user commands through the microphone, such as received, medical assistance required, and generates feedback messages after processing by the intelligent voice interaction module; key feedback selects preset options through the terminal's physical buttons, such as evacuated, not evacuated; image feedback supports users to take pictures of the on-site situation, such as photos of damaged houses, which are compressed and sent to the emergency command and control center along with the feedback message.

[0025] The speech recognition module is internally equipped with a speech signal preprocessing unit, a feature extraction unit, an acoustic model decoding unit, and a domain-adapted language model unit. The speech signal preprocessing unit is responsible for the preliminary processing of the original speech signal collected by the user terminal, including environmental noise suppression, speech endpoint detection, and signal normalization: Noise suppression reduces background interference (such as wind and alarm sounds at disaster sites) through adaptive filtering technology to ensure the clarity of effective speech segments; endpoint detection identifies the start and end positions of speech and filters out meaningless silence or noise; signal normalization adjusts the speech amplitude to avoid signal strength fluctuations caused by differences in acquisition equipment. The feature extraction unit converts the preprocessed time-domain speech signal into frequency-domain features, using Mel-frequency cepstral coefficients (MFCCs) as the core feature parameters. Through steps such as frame windowing, fast Fourier transform, Mel filter bank filtering, and discrete cosine transform, it extracts speech feature vectors that reflect the characteristics of human auditory perception, providing a stable input representation for subsequent models. The acoustic model decoding unit is based on a deep neural network (e.g., a combination of a convolutional neural network and a long short-term memory network). It maps feature vectors into phoneme sequences using trained network parameters. The model training data covers typical speech in emergency scenarios (e.g., urgent calls for help, dialect expressions, and mixed multilingual input), ensuring robust recognition of complex speech. The domain-adaptive language model unit incorporates a specialized vocabulary for emergency response (e.g., high-frequency terms like "medical assistance," "supply shortage," and "rescue coordinates") to construct a statistical language model for large-vocabulary continuous speech recognition. Using n-gram or pre-trained language model techniques, it optimizes the decoding process from phoneme sequences to text, improving the recognition accuracy of proper nouns and task-oriented sentences in emergency scenarios.

[0026] The natural language processing and demand analysis module first preprocesses the input text (word segmentation, stop word removal, and abnormal expression standardization) to extract key semantic elements such as "demand type" (medical care / supplies / rescue, etc.), "location" (XX Village), and "urgency" (implicit or explicit). Second, it uses a pre-trained language model (such as an improved BERT) to obtain the contextual semantic representation of the text. Combined with the emergency domain knowledge base (such as the preset emergency resource dispatch rules for "medical help" and the matching logic for "supply demand" to supply storage points), it constructs a feature vector that integrates general semantics and domain knowledge. Finally, it uses an intent classifier (such as a multi-layer perceptron) to output the user's core intent (such as "medical help request") and uses a slot filling model to extract key parameters in the demand (such as location coordinates and quantity required). Ultimately, it generates a structured instruction containing "intent type + parameter set" (such as {intent: medical help, location: XX Village, priority: high}) for subsequent GIS module and emergency command center calls.

[0027] In view of the limited types of emergency scenarios but sensitive parameters, this module designs a joint optimization objective that integrates contextual attention and domain constraints. The formula is as follows: L=α·L 意图 +(1−α)·L 槽位 +λ·L 领域 ; in: Lintent is the cross entropy loss for intent classification, which is used to optimize the model's ability to discriminate the user's core demand types; The L slot is the conditional random field (CRF) loss for slot filling, which is used to accurately locate and label key parameters in the demand (such as location and material type); L domain is the domain constraint loss, which is calculated by comparing the semantic similarity (such as cosine similarity) between standard instructions in the emergency domain knowledge base and the model output to ensure that the parsing results conform to the professional logic of the emergency scenario; α (0<α<1) is the weight balance coefficient between intention and slot task, which is dynamically adjusted according to the characteristic that intention discrimination takes precedence over parameter extraction in emergency scenarios; λ is the regularization coefficient of the domain constraint, which is used to control the guidance intensity of domain knowledge on model training.

[0028] This formula embeds prior knowledge of emergency scenarios into the model training process through domain constraints, solving the problem of "semantically reasonable but invalid instructions" of general NLP models in professional scenarios (for example, avoiding misinterpreting "I need water" as "drinking water demand" instead of "fire water demand"), and significantly improving the accuracy and practicality of emergency demand analysis.

[0029] The internal settings of the multilingual speech synthesis module include a text normalization processing unit, a multilingual phoneme conversion unit, a cross-lingual prosody generation unit, and a speech waveform synthesis unit. The text normalization processing unit performs normalization processing according to the text characteristics of different languages: for Chinese text, it completes simplified-to-traditional conversion, number-to-Chinese conversion, and punctuation unification; for English text, it processes abbreviation expansion (such as converting "Dr." to "Doctor") and case normalization; for connected writing languages such as Arabic, it identifies word boundaries and adds separation marks; for minority language texts (such as Tibetan, Uyghur), it performs format correction based on predefined text encoding rules to ensure the consistency of the input text. The multilingual phoneme conversion unit converts the normalized text into a phoneme sequence corresponding to the language through the phoneme mapping table of each language, such as the pinyin phonemes of Chinese, the International Phonetic Alphabet (IPA) phonemes of English, and the kana phonemes of Japanese. At the same time, it processes polyphonic characters (such as the different pronunciations of the Chinese character "行" in "行动" and "银行") and cross-lingual loanwords (such as the pronunciation variants of the English word "ambulance" in French) to ensure the accuracy of phoneme annotation. The cross-lingual prosody generation unit generates the prosody features of speech according to the context semantics and language characteristics, including tones (such as the four tones of Chinese), stresses (such as word stress in English), pauses (such as the inter-sentence pause rules in Japanese), and speech rates (such as accelerating the speech rate for emergency instructions). It learns the prosody patterns of different languages through statistical models or deep learning models (such as the Transformer architecture) and adjusts the prosody parameters according to the emotional needs of emergency scenarios (such as soothing tone, emergency prompt). The speech waveform synthesis unit converts the phoneme sequence and prosody features into audible speech waveforms based on vocoder technology (such as WaveNet or HiFi-GAN). It optimizes the vocoder parameters for multilingual synthesis to ensure the naturalness of Chinese tones, the fluency of English liaison, and the pronunciation accuracy of minority languages, and finally outputs synthetic speech that conforms to the user's language habits.

[0030] The data storage and analysis module consists of a data archiving unit, a structured storage unit, an analysis and modeling unit, and a result output unit. The data archiving unit categorizes the raw data generated by each module, including sensor data from the information acquisition module, user terminal feedback records, and map logs from the GIS module, by timestamp. It generates daily and monthly data files in CSV or JSON format and stores them redundantly on RAID to prevent data loss. The structured storage unit establishes an emergency database, which includes a disaster information table recording disaster type, time, and impact area; a user feedback table recording user ID, feedback content, and location; and a device status table recording the operating status of sensors and terminal devices. A relational database, such as MySQL, manages inter-table associations, such as linking user feedback to corresponding disaster events. The analysis and modeling unit applies analytical algorithms tailored to specific needs: trend analysis uses time series models, such as ARIMA, to predict disaster development, such as rising flood levels; user behavior analysis uses clustering algorithms, such as K-means, to identify high-frequency feedback patterns, such as areas of concentrated medical assistance; and equipment reliability analysis uses fault tree analysis (FTA) to locate vulnerable modules, such as periods of high sensor communication failure. The result output unit converts the analysis results into visual reports, including line charts showing disaster frequency, heat maps showing feedback-intensive areas, and structured reports containing key conclusions and recommended measures for decision-making reference by the emergency command and control center.

[0031] The emergency command and control center module is internally equipped with an operation monitoring unit, a command dispatch unit, a resource coordination unit, and an interface interaction unit. The operation monitoring unit collects real-time status data from each module, including the "data reception rate" of the information acquisition module, the "number of satellite connections" of the Beidou communication module, and the "number of online user terminals." This data is displayed on the command screen as a dashboard, with green indicating normal, yellow indicating a warning, and red indicating a fault. Alarm thresholds are set, such as "the number of satellite connections falling below 2" triggering an audible and visual alarm. The command dispatch unit receives operational commands entered by the command personnel, such as "send an evacuation notice to area XX," and calls the Beidou short message communication module's interface to encapsulate and send the command. It also records a command log, including the sending time, receiving range, and content summary, to support command tracing, such as "which terminals were covered by the typhoon warning issued on July 1st?" The resource coordination unit maintains an emergency resource database, including "medical material storage points," "rescue team locations," and "disaster shelter capacity." When receiving user needs, such as "XX village needs medical help," it generates the optimal deployment plan, such as "dispatching XX hospital ambulances to XX village," through spatial distance calculations (for example, the nearest rescue team is 3 kilometers away from the user) and resource surplus inquiries (for example, "XX hospital has 2 ambulances remaining"). The plan is then synchronized with relevant modules, such as notifying the rescue team and updating feedback information on the user terminal. The interface interaction unit provides a visual operation interface, including a map control area that supports zooming in / out on the map and switching layers, a command input area that supports text box input of broadcast content, a resource query area that supports a drop-down menu to select material types, an alarm prompt area, and a pop-up window to display abnormal status. Commanders can complete the entire process through mouse clicks, keyboard input, and other methods to ensure the efficiency and operability of emergency response.

[0032] From the above we can know: In the present invention, the voice recognition module in the intelligent voice interactive broadcast module can quickly convert the voice commands input by the user through the microphone (such as "notification received" and "medical help needed") into text information, avoiding the tedious manual input and possible operational errors, making user feedback more direct and efficient; the natural language processing and demand analysis module performs semantic analysis on the recognized text, accurately extracts the core content of the user's needs (such as distinguishing between "emergency help" and "regular feedback"), and associates the emergency keyword library in the system (such as "medical" and "supplies"), so that the system can quickly understand the user's intention and provide an accurate basis for the subsequent resource allocation and command generation of the command center; the multilingual speech synthesis module can convert the broadcast text generated by the system (such as disaster warnings and disaster avoidance guidelines) into a language or dialect familiar to the user (such as the native language of ethnic minority areas and the local dialect of dialect areas), solving the problem of information reception barriers caused by language differences in traditional broadcasting, and ensuring that users with different language backgrounds can clearly understand the broadcast content. The synergistic effect of the three significantly improves the interaction efficiency between the system and users, the accuracy of demand response, and the coverage of information transmission, upgrading emergency broadcasts from "one-way notification" to "two-way interaction", further enhancing the practicality and reliability of the system in complex scenarios.

[0033] In this invention, the information collection and processing module integrates data from multiple sources, including sensors and government platforms, to ensure comprehensive information sources. The Beidou short message communication module leverages satellite coverage to address the vulnerability of traditional communications to interruption in remote areas and disaster scenarios. Even in areas without mobile phone coverage, emergency notifications (such as typhoon paths and flood warnings) can be accurately transmitted to user terminals via satellite channels. The user terminal module supports multimodal presentations, including voice, text, and map annotations, enabling users of all ages and educational levels to quickly understand information. The feedback interaction function allows the command center to monitor user status (e.g., whether notifications have been received or whether assistance is needed) in real time, forming a closed "send-receive-feedback" loop. This avoids the problem of "one-way broadcast without an understanding of the effect" and significantly improves the effectiveness of information transmission.

[0034] This system enhances the accuracy of emergency response and the efficiency of resource allocation. The Geographic Information System module correlates user locations with disaster areas (e.g., flood inundation zones, earthquake intensity zones), and combines specific user feedback (e.g., medical assistance, supply shortages) to generate a matching solution based on "user location - demand type - nearest resource point." The Emergency Command and Control Center module monitors the operating status of each module (e.g., number of satellite connections, terminal online rate) to promptly detect anomalies. Dispatch commands can be directly connected to the Beidou communication module for rapid issuance. The resource coordination function dynamically adjusts deployment plans based on real-time data (e.g., shelter capacity, rescue team locations). Historical data recorded by the Data Storage and Analysis module (e.g., high-risk disaster areas, high-frequency user requests) also provides a basis for optimizing subsequent emergency strategies, making each disaster response more efficient and more tailored to actual needs.

[0035] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further limitations, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0036] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multifunctional emergency broadcast system based on Beidou short message, characterized by: The system includes: information collection and processing module, Beidou short message communication module, geographic information system module, user terminal module, intelligent voice interactive broadcast module, data storage and analysis module and emergency command and control center module; The intelligent voice interactive broadcast module is internally configured with: a voice recognition module, a natural language processing and demand analysis module, and a multilingual voice synthesis module; The standardized broadcast text output by the content generation unit of the information collection and processing module is connected to the protocol encapsulation unit of the Beidou short message communication module; The position data output of the user terminal module is connected to the position analysis unit of the geographic information system module for coordinate conversion and position information extraction; The voice feedback input of the user terminal module is connected to the voice recognition module of the intelligent voice interactive broadcast module; The user demand information after natural language processing and demand analysis by the intelligent voice interactive broadcast module is connected to the resource coordination unit of the emergency command and control center module; The speech content generated by the multilingual speech synthesis module is output to the speech playback unit connected to the user terminal module; The analysis results of the data storage and analysis module are connected to the operation monitoring unit and resource coordination unit of the emergency command and control center module to provide data support for decision-making; The command scheduling unit of the emergency command and control center module is connected to the protocol encapsulation unit of the Beidou short message communication module, through which command instructions are sent; the operation monitoring unit of the emergency command and control center module is connected to the status data output end of each module to obtain the operation status of each module in real time.

2. A multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The information collection and processing module is internally provided with a data access unit, a quality filtering unit, an information integration unit and a content generation unit.

3. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The Beidou short message communication module is internally provided with a protocol encapsulation unit, an encryption processing unit, a channel scheduling unit and a receipt verification unit.

4. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The geographic information system module is internally provided with a position parsing unit, a spatial association unit, a visualization rendering unit and a data fusion unit.

5. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The user terminal module is internally provided with a message receiving unit, a content parsing unit, a multimodal display unit and a feedback interaction unit.

6. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The speech recognition module is internally provided with a speech signal preprocessing unit, a feature extraction unit, an acoustic model decoding unit and a domain adaptation language model unit.

7. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The natural language processing and demand analysis module takes into account the limited types of emergency scenario requirements but sensitive parameters, and integrates the joint optimization objectives of context attention and domain constraints. The formula is: L=α·L 意图 +(1−α)·L 槽位 +λ·L 领域 ; in: Lintent is the cross entropy loss for intent classification, which is used to optimize the model's ability to discriminate the user's core demand types; The L slot is the conditional random field loss for slot filling, which is used to accurately locate and mark the key parameters in the requirements; L domain is the domain constraint loss. By comparing the semantic similarity between standard instructions in the emergency domain knowledge base and the model output, it ensures that the parsing results conform to the professional logic of the emergency scenario. α (0<α<1) is the weight balance coefficient between intention and slot task, which is dynamically adjusted according to the characteristic that intention discrimination takes precedence over parameter extraction in emergency scenarios; λ is the regularization coefficient of the domain constraint, which is used to control the guidance intensity of domain knowledge on model training.

8. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The multilingual speech synthesis module is internally provided with a text normalization processing unit, a multilingual phoneme conversion unit, a cross-language prosody generation unit and a speech waveform synthesis unit.

9. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The data storage and analysis module is internally provided with a data archiving unit, a structured storage unit, an analysis modeling unit and a result output unit.

10. The multifunctional emergency broadcast system based on Beidou short message according to claim 1, characterized in that: The emergency command and control center module is internally equipped with an operation monitoring unit, a command scheduling unit, a resource coordination unit and an interface interaction unit. The operation monitoring unit collects status data of each module in real time, including the "data receiving rate" of the information acquisition module, the "number of satellite connections" of the Beidou communication module, and the "number of online users" of the user terminals, and displays them in the form of a dashboard on the command screen. Green indicates normal, yellow indicates warning, and red indicates fault, and threshold alarms are set.

Images