TWM679063U - Cross-regional air traffic control voice information prompt system - Google Patents

Abstract

A cross-regional air traffic control voice information prompting system includes a model training server that trains corresponding voice and text models based on accents and speech characteristics of different regions. When pilots communicate with the control tower via onboard equipment, the system instantly receives the tower personnel's voice, compares it to the voice model from the region that best matches, and then converts it into corresponding text displayed on the screen to help pilots quickly understand instructions. This entire system uses wireless connectivity to facilitate smoother communication between the control tower and the aircraft, especially in cross-border or significantly accented situations, greatly reducing misunderstandings and communication errors.

Description

Cross-regional air traffic control voice information prompt system

This invention relates to an air traffic communication technology, and in particular to a system and method for receiving, identifying and analyzing air traffic control (ATC) voice messages and providing the results to pilots in real time.

In current air traffic management systems, air traffic control (ATC) primarily uses radio for voice communication, with ground air controllers issuing flight instructions to pilots. Due to the high timeliness and accuracy requirements of aviation operations, control terminology must conform to standardized specifications, such as those established by the International Civil Aviation Organization (ICAO). However, in practice, air controllers and pilots in different regions may communicate using regional accents, intonations, or varying speaking speeds, making semantic comprehension difficult, especially for non-native speakers, potentially leading to misunderstandings and even flight risks.

While traditional speech recognition systems can be used to convert speech into text, most systems are trained for general speech or standard intonation and cannot effectively recognize air traffic control speech with regional intonation characteristics. Furthermore, it is difficult to maintain recognition accuracy when applied across regions. In addition, most current systems do not integrate real-time aircraft location information and cannot automatically switch to the corresponding speech model based on the aircraft's location, further limiting their applicability on global routes.

While some systems can record or store control conversations, these are mostly for simple recording purposes and are not integrated with voice recognition and prompting mechanisms, nor can they provide personalized recognition support based on regional accents. In conclusion, current technology lacks a cross-regional air traffic control voice system that can effectively combine regional voice characteristics, real-time recognition, and information prompts to improve the timeliness and accuracy of flight communications.

The purpose of this invention is to provide a system and method that can receive, identify, and process air traffic control (ATC) voice messages with regional intonation characteristics in real time, and to provide relevant information to pilots through various means such as text display, so as to assist them in understanding the communication content in real time, thereby reducing the risk of misunderstanding caused by intonation differences or accents, improving communication efficiency and flight safety, and reducing the high dependence on pilots' language experience and listening comprehension ability, thus lowering the operating threshold.

This invention overcomes the human error inherent in existing technologies where ATC voice communication relies entirely on the pilot's real-time interpretation, memorization, and operation, especially in situations involving different regions, intonations, or non-standard pronunciations. Furthermore, this invention provides an integrated system combining regional voice models, real-time voice recognition, personalized voice conversion, and information prompts. It automatically applies the corresponding voice model based on the aircraft's geographical location, providing voice analysis and auxiliary prompts, offering real-time, targeted, and visualized analysis results, significantly improving the accuracy and efficiency of flight communication.

This system is a cross-regional air traffic control voice information prompting system, mainly to help pilots clearly understand local air traffic control (ATC) instructions in different regions.

The system comprises several key components: a model training server and a database. The model training server has a built-in database that collects and stores air traffic control voice data from different regions. This data may be text conversation records or recorded voiceprints, all of which have the characteristics of local accents or intonations.

The model training module uses speech data from these regions to train the model, further establishing two types of speech-to-text models: one that preserves the local intonation and another that converts it to standard speech. These models contain speech and corresponding text content, which can assist in subsequent speech recognition and conversion tasks.

The acquisition and synthesis of audio data involves first extracting key audio segments from local air traffic control conversations using Natural Language Processing (NLP) technology, and then converting these text conversations into speech using speech synthesis technology to create voiceprint data with local intonation.

The flight-mounted device connects to the model training server and has built-in GPS functionality, enabling real-time confirmation of the aircraft's location. Based on the location results, the system switches to the local air traffic control dialogue environment and displays the real-time dialogue on the screen.

The wireless communication equipment connects to the control tower, and the aircraft is also connected to a dedicated ATC radio to communicate with the local control tower via a pre-set frequency. Ground air controllers communicate with the pilots via voice communication through the control tower. These conversations are recorded and stored in a database linked to the current geographic location.

Voice recognition and information prompts: When a pilot receives voice instructions from the control tower, the system performs the following steps: First, it converts the air traffic controller's voice into a voiceprint; then, it compares this voiceprint with models in the database to find the closest voice model; finally, it extracts the corresponding text content from that model and displays it on the screen for the pilot. Similarly, when a pilot responds to the control tower, the system also: compares the pilot's voiceprint with the voice model of their area; after confirming the content is correct, it converts it into the local standard voice format and sends it to the control tower for playback.

This system is not only usable across tones and regions, but also greatly improves the clarity and accuracy of voice communication, which is very helpful in improving flight safety and efficiency.

The regional voice text model includes: 1. Initial Contact: Controller: Taipei Tower, China Airlines 001, Goodnight. Pilot: Taipei Tower, China Airlines 001, Goodnight. 2. Takeoff Clearance: Controller: China Airlines 001, Runway 05, Takeoff Clearance. Pilot: Runway 05, Takeoff Clearance, China Airlines 001. 3. Descent Clearance: Controller: China Airlines 001, Descent to 3000 feet. Pilot: Descent to 3000 feet, China Airlines 001. 4. Emergency Situation: Pilot: Mayday, Mayday, Mayday, China Airlines 001, Engine Failure. Controller: China Airlines 001, Received, Maintain current altitude, we will provide assistance as soon as possible, and other precautions.

A cross-regional air traffic control voice information prompting system, as shown in Figures 1 and 2, comprises: a model training server 10, which includes a database 11 storing a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of regions, wherein the dialogue texts and/or the voiceprints have a local intonation characteristic of the respective regions; and a model training module 12, which trains itself based on the dialogue texts and/or the voiceprints corresponding to the respective regions, and outputs a plurality of regional voice text models and/or a plurality of standard voice text models containing the local intonation characteristics, wherein each voice text model includes voice data and text data.

In one embodiment, the voiceprints are selected from those regions, have the characteristics of the local dialect, and are recorded according to those dialogue texts.

In one embodiment, the voiceprints are extracted by the model training server 10 from voice segments related to air traffic control dialogues within those regions using Natural Language Processing (NLP), and then the corresponding dialogue texts for those regions are synthesized into speech using a speech text synthesis system.

This speech-to-text synthesis system includes: 1. Voice Feature Extraction (HuBERT): The HuBERT model is used for voice feature extraction. 2. Semantic Feature Extraction (VALL-E): The system adopts the VALL-E architecture, combining text features, voice features, and the Transformer-Based Bidirectional Encoder Representation (BERT) technique. 3. Speech Synthesis Optimization (VITS): The VITS architecture is used to jointly optimize semantic features, voice features, text information, and target voice.

As shown in Figure 3, a flight terminal device 20 is connected to the model training server 10. The flight terminal device 20 includes a Global Positioning System 21 (GPS) to locate the position of the flight terminal device 20 at any time. A processor 22 performs an air traffic control dialogue operation corresponding to the region according to the GPS 21, and a display 23 displays the air traffic control dialogue operation.

A radio device (ATC radio device) (not shown) is connected to the flight terminal device 20 to transmit and receive voice signals with a tower terminal device 30 via an aviation frequency.

The tower-end device 30 is wirelessly connected to the radio device, through which an air traffic controller communicates with a pilot.

When the flight terminal device 20 receives a command signal executed by the pilot on the radio device or on the flight terminal device 20, the flight terminal device 20 initiates the air traffic control dialogue operation.

The database 11 stores the dialogue between the pilot and the air traffic controller during the air traffic control dialogue operation and associates it with the region.

In one embodiment, the command signal includes transmitting the voice signal to the flight end device 20 and/or the tower end device 30 by the pilot pressing a call button on the radio device.

In one embodiment, the command signal also includes the pilot clicking a call icon on the flight terminal device 20 or speaking a command slogan to the flight terminal device 20. The command slogan can be but is not limited to opening a traffic control dialogue, starting a traffic control dialogue, requesting a dialogue, connecting to a ground tower, etc.

The air traffic control communication operation includes the following stages, but is not limited to: pre-flight (announcement and taxiing), takeoff and departure, cruise (contacting the flight information area), approach and landing, taxiing and parking, etc.

The processor 22 performs a voice conversion operation, converting a voice signal of the air traffic controller in the air traffic control dialogue into a voiceprint of the air traffic controller. It then compares the voiceprint of the air traffic controller with the voice data in the regional voice text models, extracts the voice data of the regional voice text model with the smallest threshold difference, and outputs the text data corresponding to the regional voice text model to the display 23.

In one embodiment, the speech conversion operation described above also includes outputting standard speech text models corresponding to the local speech text model through a headset.

In one embodiment, the speech conversion operation also includes the processor 22 performing an automatic speech recognition and semantic analysis (ASR) to analyze the traffic controller's speech and compare it to the most suitable text of the speech.

In one embodiment, the threshold calculation method includes the difference between the voiceprint of the voice signal and the average difference between those voiceprints.

The model training server 10 also includes a registration module 13, which allows the pilot to set a pilot language and/or a pilot region, and the voice conversion operation is performed according to the pilot language and/or the pilot region.

The voice conversion operation also includes comparing the voice signal of the pilot in the air traffic control dialogue with the voice data in the voice text models of the pilot's region, extracting the voice data with the smallest threshold difference, converting it into the standard voice text models of the region located by the global positioning system 21, and outputting the voice data corresponding to the standard voice text model to the tower end device 30.

The display 23 includes, but is not limited to, image display, text display, voice playback, head-up display (HUD) or helmet display device.

The flight terminal device 20 can be any type of smartphone, tablet, laptop, etc., but it must have a communication interface for receiving data transmitted from the ATC communication system and an audio interface for outputting sound corresponding to the data received from the ATC communication system.

The flight terminal device 20 and/or the radio device includes analog-to-digital (ADC) and/or digital-to-analog (DAC) conversion functions.

A cross-regional air traffic control voice information prompting system, as shown in Figure 4, includes the following preliminary operations: A1 A model training server 10 stores a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of regions, wherein the dialogue texts and/or the voiceprints have the local intonation characteristics of the regions; A2 Through a model training module 12, based on the dialogue texts and/or the voiceprints, train a plurality of regional voice text models and/or a plurality of standard voice text models with the local intonation characteristics, wherein the voice models include voice data and text data.

The pre-processing also includes the model training server 10 providing a pilot with the option to set a pilot language and/or a pilot region in a registration module 13.

A cross-regional air traffic control voice information prompting system, as shown in Figure 5, includes the following operations: A10 A flight terminal device 20 determines its flight position via a Global Positioning System 21 (GPS) and identifies the current air traffic control area; A15 When the flight terminal device 20 receives a command signal executed by a pilot on the flight terminal device 20, the flight terminal device 20 initiates an air traffic control dialogue operation; A20 During the air traffic control dialogue operation, when the flight terminal device 20 receives a voice signal from an air traffic controller on a tower terminal device 30, a processor 22 converts the voice signal into an air traffic controller's voiceprint according to a voice conversion operation; A25 The processor 22 compares the traffic controller's voiceprint with the voice data stored in a database 11 that corresponds to a region with a local accent, and calculates a threshold; A30 The processor 22 extracts the voice data with the smallest threshold difference from the database 11, outputs the text data corresponding to the regional voice text model, and displays it on a display 23.

Step A25 also includes the processor 22 performing an automatic speech recognition (ASR) and semantic analysis to help compare the traffic controller's voiceprint and find the most suitable text data.

Step A30 also includes converting the regional voice text model according to a pilot region set in a registration module 13, converting it into the text data of the regional voice text model corresponding to the pilot region, and displaying it on the display 23.

Step A30 further includes outputting the speech data from the standard speech text models corresponding to the regional speech text model through an earpiece, or outputting the speech data from the regional speech text model corresponding to the pilot's region.

Step A30 Continuing A40 The processor 22 compares the voice signal of the pilot in the air traffic control dialogue with the voice data of the regional voice text models of the pilot's region, extracts the voice data with the smallest threshold difference, and converts it into the standard voice text models of the region located by the global positioning system 21, and outputs the voice data corresponding to the standard voice text model to the tower end device 30.

A cross-regional air traffic control voice information prompting system, as shown in Figure 6, is combined with a radio device (ATC radio device). The method includes: B10 A flight terminal device 20 determines its flight position via a global positioning system 21 and determines the current air traffic control area; B15 When the flight terminal device 20 receives a signal transmitted by a pilot pressing a call button on the radio device, the flight terminal device 20 initiates an air traffic control dialogue operation; B20 The radio device transmits a voice signal received from an air traffic controller at a tower terminal device 30 to the flight terminal device 20, and a processor 22 converts the voice signal into an air traffic controller's voiceprint according to a voice conversion operation; B25 The processor 22 compares the traffic controller's voiceprint with a voice data in a database 11 that stores a plurality of regional voice text models corresponding to a local dialect feature of the region, and calculates a threshold; B30 The processor 22 extracts the voice data with the smallest threshold difference from the database 11, outputs a text data corresponding to the regional voice text model, and displays it on a display 23.

Step B25 also includes the processor 22 performing an automatic speech recognition (ASR) and semantic analysis to help compare the traffic controller's voiceprint and find the most suitable text data.

Step B30 also includes converting the regional voice text model according to a pilot region set in a registration module 13, converting it into the text data of the regional voice text model corresponding to the pilot region, and displaying it on the display 23.

Step B30 further includes the flight end device 20 transmitting the voice data in the standard voice text models corresponding to the regional voice text model to the radio device, or transmitting the voice data in the regional voice text model corresponding to the pilot's region to the radio device, and further outputting it through an earpiece of the radio device.

Step B30 continues to B40. The radio device transmits the pilot's voice signal during the conversation to the processor 22. The processor 22 compares the voice data with the regional voice text models of the pilot's region, extracts the voice data with the smallest threshold difference, and converts it into the standard voice text models of the region located by the global positioning system 21. The processor outputs the voice data corresponding to the standard voice text model and transmits it to the tower device 30 via the radio device.

In one embodiment, a cross-regional air traffic control voice information prompting system is provided. The system includes: a model training server 10, which includes a model training module 12 trained based on a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of regions, to output a plurality of regional voice models containing local dialect characteristics and corresponding plurality of regional text models; and a flight terminal device 20 connected to the model training server 10, the flight terminal device 20 including: A global positioning system 21 is used to locate the region where the flight terminal device 20 is located at any time; a processor 22 is used to perform an air traffic control dialogue operation based on the region; a display 23 is used to display information of the air traffic control dialogue operation; a tower terminal device 30 is wirelessly connected to the flight terminal device 20, and an air traffic controller conducts a voice dialogue with a pilot through the tower terminal device 30; wherein, the air traffic control dialogue operation includes a voice conversion operation performed by the processor 22 to compare a voice signal issued by the air traffic controller in the dialogue with some regional voice models, to extract the regional voice model with the minimum threshold difference between the two, and output the corresponding regional text model to the display 23 for display.

One embodiment, a cross-regional air traffic control voice recognition and information prompting method, as shown in Figure 7, includes: C10. A model training module 12 of a model training server 10 trains the model training module based on a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of regions, to output a plurality of regional voice models containing local dialect characteristics and corresponding plurality of regional text models; C20. A flight terminal device 20 determines its flight position via a global positioning system 21 and determines its current region; C30. When the flight terminal device 20 receives a voice signal from an air traffic controller from a tower terminal device 30, it initiates an air traffic control dialogue operation, and a processor 22 performs a voice conversion operation; C40. The processor 22 compares the speech signal with the regional speech models stored in a database 11 that correspond to the region and have the characteristics of the local accent, and calculates a threshold difference; C50. The processor 22 extracts the regional speech model with the smallest threshold difference from the database 11, outputs the corresponding regional text model, and displays it on a display 23.

The model training module 12 also outputs a plurality of air traffic control standard voice models, and the voice conversion operation also includes outputting the air traffic control standard voice models or the voice signals generated by the local voice models corresponding to the local voice text model through an earpiece.

The voice conversion operation also includes comparing the voice signal emitted by the pilot in the conversation with a pilot region set by the pilot in a registration module 13, extracting the regional voice model with the smallest threshold difference from the voice signal according to the regional voice models corresponding to the pilot region, converting it into the air traffic control standard voice model corresponding to the region located by the global positioning system 21, and outputting the voice signal generated by the air traffic control standard voice model or the regional voice model to the tower end device 30.

One embodiment discloses a cross-regional air traffic control voice information prompting system, comprising: a model training server 10, trained based on a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of local dialect features, to output a plurality of regional voice models and corresponding a plurality of regional text models; a flight terminal device 20 connected to the model training server 10, the flight terminal device 20 performing an air traffic control dialogue operation; a display 23 for displaying information from the air traffic control dialogue operation; and a tower terminal device. 30 is wirelessly connected to the flight terminal device 20, and an air traffic controller conducts a voice conversation with a pilot through the tower terminal device 30; wherein, the air traffic control conversation operation performs a voice conversion operation to compare a voice signal issued by the air traffic controller in the conversation with some regional voice models, to extract the regional voice model with the smallest difference between the two, and output the corresponding regional text model to the display 23 for display.

In one implementation, the voiceprints and dialogue texts included pre-flight (announcement and taxiing) dialogue: Pilot: "Taichung ground station, this is Bravo-Nav-2345. A Cessna 172, two pilots, in the general airspace, visual flight plan to Hualien, requesting taxiing." Ground controller: "Bravvo-Nav-2345, Received. Taxi to runway 36, via Alfa Runway, wind direction 330, wind speed 8 knots, maintain visual flight status. Please notify the transfer tower after departure." Pilot: "Taxi to runway 36, via Alfa Runway, maintain visual flight status, Bravo-Nav-2345."

In one embodiment, the voiceprints and dialogue texts included takeoff and departure dialogue: Pilot: "Taichung Tower, Bravo-Naf-2345, take position on runway 36, preparing for takeoff to Hualien." Tower Controller: "Bravo-Naf-2345, wind direction 330, wind speed 8 knots, cleared for takeoff from runway 36, after takeoff, veer east away from the city, maintain visual contact." Pilot: "Roger, cleared for takeoff from runway 36, after takeoff, veer east, Bravo-Naf-2345."

In one embodiment, the voiceprints and dialogue texts included the following conversation during cruise (contacting the Flight Information Region): Pilot: "Taipei Flight Information Region, this is Bravo-Nav-2345, currently over the Dajia River at an altitude of 4,500 feet, visually heading towards Hualien. Please provide airspace traffic information." Taipei FIR: "Bravvo-Nav-2345, received. Traffic conditions in your vicinity are currently normal. Three nautical miles to the 8 o'clock direction, there is an aircraft at an altitude of 5,000 feet heading south. Please maintain visual distancing." Pilot: "Received, maintaining visual distancing, Bravo-Nav-2345."

In one embodiment, the voiceprints and dialogue texts included an approach and landing (Hualien Airport) dialogue: Pilot: "Hualien Tower, this is Bravo-Naf-2345, 10 nautical miles southwest of the airport, altitude 1,500 feet, preparing to approach and land on runway 21." Tower Controller: "Bravo-Naf-2345, approach clearance, runway 21, wind direction 210, wind speed 6 knots, report long terminal line position." Pilot: "Roger, approach runway 21, report back when you reach the long terminal line, Bravo-Naf-2345." ... "Bravo-Naf-2345, entered the long terminal line." Tower Controller: "Bravo-Naf-2345, landing clearance, runway 21." Pilot: "Roger, runway 21 cleared for landing, Bravo-Nav-2345."

In one embodiment, the voiceprints and dialogue texts include taxiing and parking dialogue: Pilot: "Tower, Bravo-Nav-2345 has landed, requesting to taxi to the general aviation parking area." Tower Controller: "Bravo-Nav-2345, taxi to the general aviation area, via taxiways E and T, be aware of ground traffic." Pilot: "Taxi to the general aviation area, via taxiways E and T, Bravo-Nav-2345."

Modeling of the training module 12: Extracting the vocal features of the air traffic controllers in those areas into comparable vectors (such as x-vectors and d-vectors). Dynamic sound source separation or filtering: After identifying the most similar target voiceprint, the system automatically amplifies or separates the target sound source while suppressing other sound sources. Real-time speech recognition: The separated target speech is input into the ASR (Automatic Speech Recognition) module for transcription, and the results are applied to the air traffic control dialogue operation or subsequent applications.

The flight terminal device 20 also includes a conflict message detection module, which analyzes the historical air traffic control dialogue operations in the region and the current dialogue between the pilot and the air traffic controller to identify any conflicts or duplications and alerts the pilot accordingly.

The flight terminal device 20 also includes a unit conversion control item, which can automatically convert the relevant values such as altitude and speed in the air traffic control dialogue operation into units (such as feet to meters).

The flight terminal device 20 also includes a history query interface for querying the dialogue between the pilot and the air traffic controller during the air traffic control dialogue operation, including filterable ATC information, the region, the flight code, and the emergency indicator.

The air traffic control dialogue operation also includes a flight information booklet. The pilot enters a flight code (such as flight number or call sign) into the registration module 13. When the display 23 displays the air traffic control dialogue operation, any dialogue related to the flight code is highlighted by color changes, flashing, icon prompts, etc.

The air traffic control dialogue operation also includes an emergency sign. For dialogues containing emergency terms such as "MAYDAY" and "PAN-PAN", the emergency sign is displayed on the top of the display 23.

The air traffic control communication also includes a frequency suggestion, prompting the pilot to switch to a designated radio frequency based on the region.

The system also includes an information display switching module (not shown) for switching the text data of the air traffic control dialogue operation to be displayed on the display 23 and/or the flight helmet display device, and can also switch the voice data of the air traffic control dialogue operation to be transmitted through the flight terminal device 20 and/or the radio device.

The channel selection for aviation frequencies, such as air traffic control using specific Very High Frequency (VHF) channels, is typically between 118.00 and 136.975 MHz.

Although certain exemplary embodiments and implementations have been described herein, other embodiments and modifications will be apparent from that description. Therefore, the conception of this invention is not limited to these exemplary embodiments, but covers a broader scope in accordance with the proposed claims and various obvious modifications and equivalent arrangements.

10: Model Training Server 11: Database 12: Model Training Module 13: Registration Module 20: Flight Terminal 21: Global Positioning System 22: Processor 23: Display 30: Tower Terminal A1~A40: Methods and Procedures for Cross-Regional Air Traffic Control Voice Recognition and Information Prompt B10~B40: Methods and Procedures for Cross-Regional Air Traffic Control Voice Recognition and Information Prompts Including ATC Radio Devices C10~C50: Methods and Procedures for Cross-Regional Air Traffic Control Voice Recognition and Information Prompts

[Figure 1] Block diagram of the cross-regional air traffic control voice information prompting system; [Figure 2] Block diagram of the model training server structure; [Figure 3] Block diagram of the flight terminal device structure; [Figure 4] Flowchart of the pre-operation method for cross-regional air traffic control voice recognition and information prompting; [Figure 5] Flowchart of the application method for cross-regional air traffic control voice recognition and information prompting; [Figure 6] Flowchart of the method for cross-regional air traffic control voice recognition and information prompting including ATC radio device; [Figure 7] Flowchart of the application method for cross-regional air traffic control voice recognition and information prompting.

10: Model Training Server

20: Flight terminal device

30: Tower-side device

Claims (10)

A cross-regional air traffic control voice information prompting system, the system comprising: a model training server, trained based on a plurality of dialogue texts and/or a plurality of voiceprints corresponding to a plurality of local dialect features, to output a plurality of regional voice models and corresponding a plurality of regional text models; a flight terminal device connected to the model training server, the flight terminal device performing an air traffic control dialogue operation, the flight terminal device including a display for displaying information of the air traffic control dialogue operation; and a tower terminal device wirelessly connected to the flight terminal device, through which an air traffic controller conducts voice dialogue with a pilot; and The air traffic control dialogue operation performs a voice conversion operation to compare a voice signal issued by the air traffic controller in the dialogue with the regional voice models, extract the regional voice model with the smallest difference between the two, and output the corresponding regional text model to the display for display. As described in Request 1, the cross-regional air traffic control voice information prompting system includes a flight information booklet. When the pilot enters a flight code into a registration module, the display shows the air traffic control dialogue operation. Whenever dialogue related to the flight code is displayed, it is emphasized by color change, flashing, graphic prompts, etc. As described in Request 1, the cross-regional air traffic control voice information prompting system, the flight terminal device also includes an emergency indicator, which is displayed on the top of the display for dialogues containing emergency terms such as "MAYDAY" and "PAN-PAN". As described in Request 1, the cross-regional air traffic control voice information prompting system includes a conflict message detection module that analyzes the historical air traffic control dialogue operations in the region and the current dialogue between the pilot and the air traffic controller to identify any conflicts or duplications and prompts the pilot accordingly. As described in Request 1, the cross-regional air traffic control voice information prompting system includes a unit conversion control item that can automatically convert the relevant values such as altitude and speed in the air traffic control dialogue operation into units. As described in claim 1, the cross-regional air traffic control voice information prompting system, the flight terminal device also includes a frequency suggestion to prompt the pilot to switch to a specified radio frequency based on the region. As described in claim 1, the cross-regional air traffic control voice information prompting system further includes a global positioning system for locating the region where the aircraft terminal is located at any time, and a processor for performing air traffic control dialogue operations based on the region. As described in claim 7, the cross-regional air traffic control voice information prompting system, the model training server also outputs a plurality of standard air traffic control voice models, the voice conversion operation converts the voice signal issued by the pilot in the conversation into the standard air traffic control voice model corresponding to the region, and outputs the voice signal generated by the standard air traffic control voice model to the tower end device. As described in claim 8, the cross-regional air traffic control voice information prompting system further includes outputting, via an earpiece, a voice signal generated by the standard air traffic control voice models corresponding to the air traffic controller's voice signal. As described in claim 1, in the cross-regional air traffic control voice information prompting system, the flight end device is further connected to a radio device for transmitting and receiving voice signals for air traffic control dialogue operations with the tower end device.