Air-ground communication exercise equipment

Abstract

The invention discloses an air-ground communication practice device, and relates to the technical field of communication practice, the air-ground communication practice device comprises a communication selection end, a bilingual interaction end and a judgment guidance end, the communication selection end is used for automatically matching an interaction model by selecting a training scene, and automatically generating a path and a road sign according to an aviation real scene; and the bilingual interaction end is used for performing real-time matching of interaction contents by tracking the interaction contents in real time and aiming at the path of the aviation stage, retrieving conversation interaction contents in real time, performing abnormal identification according to an identification result of the interaction contents, and guiding the contents to be embodied according to aviation training contents and an oral English pronunciation method. According to the air-ground conversation training equipment, the authenticity of the aviation conversation training effect is improved, the stability and the monitoring effect of aviation conversation training are guaranteed, a student is automatically reminded of Chinese version conversation content, whether the English student is blocked or not is automatically retrieved, the conversation content is evaluated in real time according to the training content of the student, and the training efficiency is improved. And students can be timely guided to correct English pronunciation.

Description

Technical Field

[0001] The present invention relates to the technical field of communication practice, and in particular to a land-to-air communication practice device. Background Art

[0002] Ground-to-air communication is a standard professional terminology communication method used between air traffic controllers and pilots. It is mainly used to transmit instructions and information and is directly related to flight safety. As a key communication method in the aviation field, the accuracy and standardization of ground-to-air communication are of vital importance. During the flight, radio calls between pilots and ground controllers need to strictly abide by specific specifications and procedures to ensure accurate communication of information. During air flight training, trainees need to be trained in the content of ground-to-air communication in Chinese and English according to aviation training scenarios, which requires the use of ground-to-air communication training equipment. The ground-to-air communication training equipment is controlled by big data equipment and Internet terminals.

[0003] At present, there are some deficiencies in aviation training: 1. During the land-to-air call interaction process, it is necessary to adapt to the aviation paths and flight stages of different scenarios. At the same time, when restoring the scene, it is impossible to achieve multi-angle scene monitoring, resulting in the limitation of simulation restoration in land-to-air call training, affecting the authenticity of the call training effect; 2. During land-to-air call training, due to the need for English interactive training, students with poor English proficiency will experience interaction freezes or obstructions. The existing technology can only repeat the English training interaction content, but cannot automatically retrieve whether English students are obstructed, resulting in a weakened use effect of land-to-air call practice equipment and affecting students' call training mood; 3. For students with poor English proficiency, inaccurate pronunciation and confusing conversation content will occur. It is impossible to evaluate the call content in real time based on the students' training content, which is not convenient for timely guiding students to correct their English pronunciation, resulting in poor call training results.

[0004] Therefore, a land-to-air communication practice device is now proposed to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a ground-to-air communication training device to solve the problems raised in the above background.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a land-air communication practice device, comprising a communication selection end, a bilingual interaction end and an evaluation and guidance end, wherein the communication selection end, the bilingual interaction end and the evaluation and guidance end are jointly provided with a voice reminder module;

[0007] The call selection terminal is used to automatically match the interaction model by selecting the training scene, generate a virtual AI character, and generate an aviation training simulation scene in real time, and automatically generate a path and landmarks according to the real aviation scene, and track the call duration and call stage in real time to achieve full process tracking of the call practice;

[0008] The bilingual interaction terminal is used to track the interaction content in real time and match the interaction content in real time according to the path of the aviation stage, retrieve the call interaction content in real time, and recognize text or English words. The interaction status is monitored in real time according to the call reaction time. If it is timed out, the call training content is repeated. If it is still timed out, it is automatically converted to Chinese training, and the interaction status is monitored again. If it is still timed out, the call practice is stopped, and the system is reported at the same time to feedback that the real-name authentication trainee’s call training is unqualified;

[0009] The judging and guiding terminal is used to identify the interactive content in real time, mark abnormalities according to the interactive content recognition results, and automatically identify whether the English pronunciation of the interactive content is accurate, and compare the pronunciation difference in real time. If the pronunciation is inaccurate or there is a pronunciation difference, the trainee is given guidance on the pronunciation difference correction. The guidance content is specific according to the aviation training content and the English oral pronunciation method, and the recording playback and recording download are realized after the call training is completed;

[0010] The voice reminder module is used to remind the interlocutor to perform real-time preparation, real-time interaction and interactive feedback of the call practice in real time through a voice alarm.

[0011] The call selection terminal includes a scene selection module, a scene generation module and a scene tracking module;

[0012] The scene selection module includes a scene selection unit and an interactive corresponding unit;

[0013] The scenario selection unit is used to set the aviation training simulation scenario, the simulation scenario is set in real time for the civil aviation flight path, and different call training interaction contents are set for different flight paths;

[0014] The interaction corresponding unit is used to set fixed words for the scenes corresponding to the training interaction contents, and obtain a standard training model for the land-air call training interaction.

[0015] The scene generation module includes a virtual AI generation unit and a scene restoration unit;

[0016] The virtual AI generation unit is used to virtually generate the aviation scene dialogue character through the virtual AI character;

[0017] The scene restoration unit is used to restore the test flight scene of the aviation stage 100% through a camera and a video capture device. The test flight scene includes the taxiing stage, the take-off stage and the route operation stage.

[0018] The scene tracking module includes a scene path unit, a landmark path unit, a call timing unit and a call tracking unit;

[0019] The scenario path unit is used to set a training path through the aviation stage, and the training path includes a taxiing path, a test flight path and a starting path;

[0020] The landmark path unit is used to color-mark landmarks on the real aviation path through landmarks. After color marking, each landmark on the path is visualized and displayed in real time;

[0021] The call timing unit is used to record the training interaction feedback duration in real time through a clock;

[0022] The call tracking unit is used to track the interactive content in real time through a data tracker.

[0023] The bilingual interaction terminal includes a bilingual interaction module, an interactive retrieval module and a bilingual switching module;

[0024] The bilingual interaction module includes a dialogue interaction unit and a phase matching unit;

[0025] The dialogue interaction unit is used to realize the real-time interaction of the land-to-air call practice through the land-to-air call training interaction standard training model;

[0026] The stage matching unit is used to achieve interaction content matching at different stages of the standard training model for air-land call training interaction.

[0027] The interactive retrieval module includes a conversation content retrieval unit, a text recognition unit and a word recognition unit;

[0028] The dialogue content retrieval unit is used to retrieve the interactive dialogue content in real time through the voice converter;

[0029] The text recognition unit is used to recognize the interactive speech and received text in the air-land call training interactive standard training model in real time, as follows:

[0030] The modal confidence parameter MAC is used to quantify the accuracy of the text. The specific formula is:

[0031]

[0032] ·

[0033] ·

[0034] ·

[0035]

[0036] Among them, φ1 is the basic data of the first recognized text at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. represents the inner product of two vectors, represents the first current moment φ1 and The similarity of represents the nth current moment φ n and similarity;

[0037] The word recognition unit is used to recognize the interactive words and received English letters in the air-land communication training interactive standard training model in real time, as follows:

[0038] The modal confidence parameter MAC is used to quantify the accuracy of English letters. The specific formula is:

[0039]

[0040] ·

[0041] ·

[0042] ·

[0043]

[0044] Among them, φ1 is the basic data of the first English letter recognized at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. Represents the inner product of two vectors represents the first current moment φ1 and The similarity of represents the nth current moment φ n and The similarity.

[0045] The bilingual switching module includes an answer timing unit, a timeout repetition unit and a bilingual switching unit;

[0046] The answer timing unit is used to record the time taken by the trainee to answer in real time through a timer;

[0047] The timeout repetition unit is used to set the time for the student to answer. If the time is exceeded, the call interaction content is repeated;

[0048] The bilingual switching unit is used to automatically switch to Chinese interactive dialogue when the student's answering time is exceeded for the second time, so as to remind the student of the call steps and call status.

[0049] The judging and guiding end includes a content recognition module, a pronunciation assessment module, a tone difference calculation module and a tone difference guiding module;

[0050] The content identification module includes a content identification unit and an abnormality identification unit;

[0051] The content recognition unit is used to recognize the interactive content in real time through the data sensor and the voice converter, and the recognition method is as follows:

[0052] Step 1: Calculate the similarity between the keywords of the interactive content and the keywords of the standard interactive model. The calculation formula is as follows:

[0053]

[0054] Among them, Y sim (Z, V0) represents the similarity function between the keywords of the current interactive content and the keywords of the standard interactive model, Z represents the word feature vector of the retrieved standard interactive model keywords, S(Z, D) represents the similarity correlation between the keywords of the current interactive content and the keywords of the standard interactive model, α represents the first weight coefficient, β represents the second weight coefficient, V0 represents the standard interactive model keyword feature vector, C i The cluster feature vector representing the keyword of the ith retrieved interactive content and the cluster set of the standard interactive model keywords, M is the number of cluster sets, i and M are positive integers;

[0055] Step 2: If the calculated Y sim The closer the (Z, V0) value is to 1, the higher the similarity of the interaction content is, and the higher the accuracy of the corresponding interaction content call is;

[0056] The abnormality identification unit is used to record the abnormality of the interactive content in real time through color identification.

[0057] The pronunciation assessment module includes a pitch recognition unit, and the pitch recognition unit includes a determination factor unit, and the determination factor unit is used to make pronunciation judgment according to one of English phonetic symbols, vowels and consonants;

[0058] The tone difference calculation module includes a tone difference comparison unit, which is used to calculate the high and low difference of English tone differences in real time through an English pronunciation sensor. If the difference = 0, it is determined that the pronunciation is correct, and if the difference ≠ 0, it is determined that there is a tone difference in the pronunciation.

[0059] The tone difference guidance module includes a tone difference guidance unit, a guidance refinement unit and a recording playback unit;

[0060] The tone difference guidance unit is used to provide conversation guidance to the trainee in real time based on the tone difference comparison result and the interaction content recognition result;

[0061] The guidance refinement unit is used to set a tone difference guidance scheme and an interaction content guidance scheme according to the tone difference comparison result and the interaction content recognition result, and to give voice reminders for the student's pronunciation;

[0062] The recording and playback unit is used to set the USB interface and the data recorder to realize recording and playback and call training download.

[0063] The present invention has the following beneficial effects:

[0064] 1. In the present invention, by setting a call selection terminal, during aviation call training, by selecting a training scene to automatically match the interaction model, and automatically generating paths and landmarks according to the actual aviation scene, it is possible to achieve multi-angle scene monitoring at different stages during the aviation training process, improve the authenticity of the aviation call training effect, and achieve full process tracking of the call practice by real-time tracking of the call duration and call stage, thereby ensuring the stability and monitoring effect of the aviation call practice.

[0065] 2. In the present invention, by setting a bilingual interaction terminal, during aviation call training, the interaction content is tracked in real time and the interaction content is matched in real time according to the path of the aviation stage. The interaction status is monitored in real time according to the call reaction time. When the interaction of students with poor English ability is stuck or blocked, the students are automatically reminded of the Chinese version of the call content, and automatic retrieval of whether English students are blocked is realized, thereby increasing the use effect of the land-to-air call practice equipment and avoiding affecting the students' mood for call training.

[0066] 3. In the present invention, by setting up an evaluation and guidance terminal, during aviation communication training, the interactive content can be identified in real time, and whether the English pronunciation of the interactive content spoken is accurate can be automatically identified, thereby avoiding inaccurate pronunciation and confusing conversation content during communication training. By performing real-time sound difference comparison, the communication content can be evaluated in real time according to the training content of the trainees, which is convenient for timely guiding the trainees to correct their English pronunciation, and the communication training effect is good. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1It is a schematic diagram of the overall system architecture of a land-to-air communication training device of the present invention;

[0068] Figure 2 It is a schematic diagram of the structure of a call selection end of a land-to-air call training device of the present invention;

[0069] Figure 3 A schematic diagram of the structure of a bilingual interaction terminal of a land-air communication practice device of the present invention;

[0070] Figure 4 The present invention is a schematic diagram of the structure of the judging and guiding end of the land-to-air communication training equipment. DETAILED DESCRIPTION

[0071] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0072] Embodiment 1

[0073] Please refer to Figure 1-2 As shown: a land-air communication practice device, including a call selection end, a bilingual interaction end and an evaluation and guidance end, wherein the call selection end, the bilingual interaction end and the evaluation and guidance end are jointly provided with a voice reminder module;

[0074] The call selection terminal is used to automatically match the interaction model by selecting the training scene, generate virtual AI characters, and generate aviation training simulation scenes in real time. It also automatically generates paths and landmarks based on the actual aviation scene, tracks the call duration and call stage in real time, and realizes the whole process tracking of call practice.

[0075] The bilingual interaction terminal is used to track the interaction content in real time and match the interaction content in real time according to the path of the aviation stage, retrieve the call interaction content in real time, and recognize text or English words. It monitors the interaction status in real time according to the call reaction time. If it times out, it will repeat the call training content. If it still times out, it will automatically switch to Chinese training, monitor the interaction status again, and if it still times out, it will stop the call practice and report to the system at the same time, giving feedback that the real-name authentication trainee’s call training is unqualified;

[0076] The judging and guiding end is used to identify the interactive content in real time, mark abnormalities according to the interactive content recognition results, and automatically identify whether the English pronunciation of the interactive content is accurate, and compare the pronunciation difference in real time. If the pronunciation is inaccurate or there is a pronunciation difference, the trainee will be guided to correct the pronunciation difference. The guidance content is specific according to the aviation training content and the English oral pronunciation method, and the recording can be played back and downloaded after the call training is completed;

[0077] The voice reminder module is used to remind the interlocutor in real time to perform real-time preparation, real-time interaction and interactive feedback of the call practice through a voice alarm.

[0078] The call selection end includes a scene selection module, a scene generation module and a scene tracking module;

[0079] The scene selection module includes a scene selection unit and an interactive corresponding unit;

[0080] The scenario selection unit is used to set the aviation training simulation scenario. The simulation scenario is set in real time for civil aviation flight paths, and different call training interaction contents are set for different flight paths.

[0081] The interaction corresponding unit is used to set fixed words for the scenes corresponding to the training interaction contents, and obtain a standard training model for the land-air call training interaction.

[0082] The scene generation module includes a virtual AI generation unit and a scene restoration unit;

[0083] The virtual AI generation unit is used to virtually generate the dialogue characters of the aviation scene through the virtual AI characters;

[0084] The scene restoration unit is used to restore the test flight scenes in the aviation stage 100% through cameras and video capture devices. The test flight scenes include the taxiing stage, take-off stage and route operation stage.

[0085] The scene tracking module includes a scene path unit, a landmark path unit, a call timing unit, and a call tracking unit;

[0086] The scenario path unit is used to set the training path through the aviation phase, and the training path includes the taxi path, the test flight path and the take-off path;

[0087] The landmark path unit is used to color-code landmarks on the real aviation path. After color coding, each landmark on the path can be visualized and displayed in real time.

[0088] The call timing unit is used to record the duration of training interaction feedback in real time through a clock;

[0089] The call tracking unit is used to track the interactive content in real time through a data tracker.

[0090] By selecting training scenarios to automatically match interactive models, generating virtual AI characters and real-time aviation training simulation scenarios, and automatically generating paths and landmarks based on actual aviation scenes, it is possible to achieve multi-angle scene monitoring at different stages of aviation training, reduce the limitations of land-to-air call training restoration, and improve the authenticity of aviation call training results. By real-time tracking of call duration and call stages, it is possible to track the entire process of call practice, thereby ensuring the stability and monitoring effect of aviation call practice.

[0091] Embodiment 2

[0092] Please refer to Figure 3 As shown: Based on the first embodiment, the bilingual interaction terminal includes a bilingual interaction module, an interactive retrieval module and a bilingual switching module;

[0093] The bilingual interaction module includes a dialogue interaction unit and a phase matching unit;

[0094] The dialogue interaction unit is used to realize the real-time interaction of the air-ground call training through the standard training model of the air-ground call training interaction;

[0095] The phase matching unit is used to achieve interaction content matching at different phases of the standard training model for air-ground call training interaction.

[0096] The interactive retrieval module includes a conversation content retrieval unit, a text recognition unit and a word recognition unit;

[0097] The dialogue content retrieval unit is used to retrieve the interactive dialogue content in real time through the voice converter;

[0098] The text recognition unit is used to recognize the interactive speech and received text in the standard training model of the air-land call training interaction in real time, as follows:

[0099] The modal confidence parameter MAC is used to quantify the accuracy of the text. The specific formula is:

[0100]

[0101] ·

[0102] ·

[0103] ·

[0104]

[0105] Among them, φ1 is the basic data of the first recognized text at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. represents the inner product of two vectors, represents the first current moment φ1 and The similarity of represents the nth current moment φ n and similarity;

[0106] The word recognition unit is used to recognize the interactive words and received English letters in real time by training the interactive standard training model of air-land communication, as follows:

[0107] The modal confidence parameter MAC is used to quantify the accuracy of English letters. The specific formula is:

[0108]

[0109] ·

[0110] ·

[0111] ·

[0112]

[0113] Among them, φ1 is the basic data of the first English letter recognized at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. Represents the inner product of two vectors represents the first current moment φ1 and The similarity of represents the nth current moment φ n and The similarity.

[0114] The bilingual switching module includes an answer timing unit, a timeout repetition unit, and a bilingual switching unit;

[0115] The answer timing unit is used to record the time taken by the students to answer in real time through a timer;

[0116] The timeout repeat unit is used to set the time for the student to answer. If the answer time is exceeded, the call interaction content will be repeated;

[0117] The bilingual switching unit is used to automatically switch to Chinese interactive dialogue when the student's answer time is exceeded for the second time, and is used to remind the student of the call steps and call status.

[0118] By real-time tracking of interactive content and real-time matching of interactive content against the path of the aviation stage, real-time retrieval of call interactive content, and recognition of text or English words, real-time monitoring of interactive status based on call reaction time, timing of call training interactive process, and real-time monitoring of interactive status, when students with poor English proficiency experience interaction freezes or obstructions, students are automatically reminded of the Chinese version of the call content, and automatic retrieval of whether English students are obstructed, the use effect of land-to-air call practice equipment is increased to avoid affecting students' mood for call training.

[0119] Embodiment 3

[0120] Please refer to Figure 4 As shown: Based on the first embodiment, the judging and guiding end includes a content recognition module, a pronunciation assessment module, a tone difference calculation module and a tone difference guiding module;

[0121] The content identification module includes a content identification unit and an abnormality identification unit;

[0122] The content recognition unit is used to recognize the interactive content in real time through the data sensor and the voice converter. The recognition method is as follows:

[0123] Step 1: Calculate the similarity between the keywords of the interactive content and the keywords of the standard interactive model. The calculation formula is as follows:

[0124]

[0125] Among them, Y sim (Z, V0) represents the similarity function between the keywords of the current interactive content and the keywords of the standard interactive model, Z represents the word feature vector of the retrieved standard interactive model keywords, S(Z, D) represents the similarity correlation between the keywords of the current interactive content and the keywords of the standard interactive model, α represents the first weight coefficient, β represents the second weight coefficient, V0 represents the standard interactive model keyword feature vector, C i The cluster feature vector representing the keyword of the ith retrieved interactive content and the cluster set of the standard interactive model keywords, M is the number of cluster sets, i and M are positive integers;

[0126] Step 2: If the calculated Y sim The closer the (Z, V0) value is to 1, the higher the similarity of the interaction content is, and the higher the accuracy of the corresponding interaction content call is;

[0127] The abnormality identification unit is used to record the abnormality of the interactive content in real time through color identification.

[0128] The pronunciation assessment module includes a pitch recognition unit, which includes a determination factor unit, which is used to make pronunciation judgments based on one of English phonetic symbols, vowels and consonants;

[0129] The tone difference calculation module includes a tone difference comparison unit, which is used to calculate the high and low difference of English tone in real time through an English pronunciation sensor. If the difference = 0, it is judged that the pronunciation is correct; if the difference ≠ 0, it is judged that there is a tone difference in the pronunciation.

[0130] The tone difference guidance module includes a tone difference guidance unit, a guidance refinement unit and a recording playback unit;

[0131] The tone difference guidance unit is used to provide real-time communication guidance to students based on tone difference comparison results and interactive content recognition results;

[0132] The guidance refinement unit is used to set a tone difference guidance plan and an interaction content guidance plan according to the tone difference comparison result and the interaction content recognition result, and to give voice reminders for the students' pronunciation;

[0133] The recording and playback unit is used to set up the USB interface and data recorder to realize recording playback and call training download.

[0134] By identifying the interactive content in real time and marking anomalies based on the interactive content recognition results, and automatically identifying whether the English pronunciation of the interactive content is accurate, inaccurate pronunciation and confusing conversation content can be avoided during call training. In real time, the sound difference comparison is performed. If inaccurate pronunciation or pronunciation difference occurs, the trainee will be given guidance on how to correct the sound difference. The guidance content is specific based on the aviation training content and the English oral pronunciation method. The call content can be evaluated in real time based on the trainee's training content, so as to guide the trainee to correct his English pronunciation in a timely manner, and the call training effect is good.

[0135] In the present invention, a land-to-air call practice device is provided. When the system is in operation, for aviation call training, firstly, the system automatically matches the interaction model by selecting the training scene, generates a virtual AI character and generates the aviation training simulation scene in real time, and automatically generates the path and road signs according to the aviation real scene, so as to realize multi-angle scene monitoring at different stages in the aviation training process, reduce the limitation of the restoration of land-to-air call training, improve the authenticity of the aviation call training effect, and realize the whole process tracking of the call practice by real-time tracking of the call duration and the call stage, so as to ensure the stability and monitoring effect of the aviation call practice; secondly, by real-time tracking of the interaction content and real-time matching of the interaction content for the path of the aviation stage, real-time retrieval of the call interaction content, and recognition of text or English words, real-time monitoring of the interaction status according to the call reaction time, for the call training interaction process It can time and monitor the interaction status in real time. When students with poor English ability encounter interaction jams or obstructions, it will automatically remind students of the Chinese version of the call content, automatically search for whether English students are obstructed, increase the use effect of air-ground call practice equipment, and avoid affecting students' call training mood. Finally, it can identify the interaction content in real time and mark abnormalities based on the recognition results of the interaction content. At the same time, it can automatically identify whether the English pronunciation of the interactive content is accurate, avoid inaccurate pronunciation and confusion in conversation content during call training, and compare the sound difference in real time. If there is inaccurate pronunciation or pronunciation difference, the students will be guided to correct the sound difference. The guidance content is specific according to the aviation training content and the English oral pronunciation method. It can evaluate the call content in real time according to the students' training content, so as to guide students to correct their English pronunciation in time and achieve good call training effect.

[0136] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A ground-to-air communication training device, characterized in that: The practice device comprises a call selection terminal, a bilingual interaction terminal and an evaluation and guidance terminal, wherein the call selection terminal, the bilingual interaction terminal and the evaluation and guidance terminal are jointly provided with a voice reminder module; The call selection terminal is used to automatically match the interaction model by selecting the training scene, generate a virtual AI character, and generate an aviation training simulation scene in real time, and automatically generate a path and landmarks according to the real aviation scene, and track the call duration and call stage in real time to achieve full process tracking of the call practice; The bilingual interaction terminal is used to track the interaction content in real time and match the interaction content in real time according to the path of the aviation stage, retrieve the call interaction content in real time, and recognize text or English words. The interaction status is monitored in real time according to the call reaction time. If it is timed out, the call training content is repeated. If it is still timed out, it is automatically converted to Chinese training, and the interaction status is monitored again. If it is still timed out, the call practice is stopped, and the system is reported at the same time to feedback that the real-name authentication trainee’s call training is unqualified; The judging and guiding terminal is used to identify the interactive content in real time, mark abnormalities according to the interactive content recognition results, and automatically identify whether the English pronunciation of the interactive content is accurate, and compare the pronunciation difference in real time. If the pronunciation is inaccurate or there is a pronunciation difference, the trainee is given guidance on the pronunciation difference correction. The guidance content is specific according to the aviation training content and the English oral pronunciation method, and the recording playback and recording download are realized after the call training is completed; The voice reminder module is used to remind the interlocutor to perform real-time preparation, real-time interaction and interactive feedback of the call practice in real time through a voice alarm.

2. The exercise device according to claim 1, characterized in that The call selection terminal includes a scene selection module, a scene generation module and a scene tracking module; The scene selection module includes a scene selection unit and an interactive corresponding unit; The scenario selection unit is used to set the aviation training simulation scenario, the simulation scenario is set in real time for the civil aviation flight path, and different call training interaction contents are set for different flight paths; The interaction corresponding unit is used to set fixed words for the scenes corresponding to the training interaction contents, and obtain a standard training model for the land-air call training interaction.

3. The exercise equipment according to claim 2, characterized in that The scene generation module includes a virtual AI generation unit and a scene restoration unit; The virtual AI generation unit is used to virtually generate the aviation scene dialogue character through the virtual AI character; The scene restoration unit is used to restore the test flight scene of the aviation stage 100% through a camera and a video capture device. The test flight scene includes the taxiing stage, the take-off stage and the route operation stage.

4. The exercise equipment according to claim 3, characterized in that The scene tracking module includes a scene path unit, a landmark path unit, a call timing unit and a call tracking unit; The scenario path unit is used to set a training path through the aviation stage, and the training path includes a taxiing path, a test flight path and a starting path; The landmark path unit is used to color-code landmarks on the real aviation path through landmarks. After color coding, each landmark on the path is visualized and displayed in real time. The call timing unit is used to record the training interaction feedback duration in real time through a clock; The call tracking unit is used to track the interactive content in real time through a data tracker.

5. The exercise equipment according to claim 1, characterized in that The bilingual interaction terminal includes a bilingual interaction module, an interactive retrieval module and a bilingual switching module; The bilingual interaction module includes a dialogue interaction unit and a phase matching unit; The dialogue interaction unit is used to realize the real-time interaction of the land-to-air call practice through the land-to-air call training interaction standard training model; The stage matching unit is used to achieve interaction content matching at different stages of the standard training model for air-land call training interaction.

6. The exercise device according to claim 5, characterized in that The interactive retrieval module includes a conversation content retrieval unit, a text recognition unit and a word recognition unit; The dialogue content retrieval unit is used to retrieve the interactive dialogue content in real time through the voice converter; The text recognition unit is used to recognize the interactive speech and received text in the air-land call training interactive standard training model in real time, as follows: The modal confidence parameter MAC is used to quantify the accuracy of the text. The specific formula is: Among them, φ1 is the basic data of the first recognized text at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. represents the inner product of two vectors, represents the first current moment φ1 and The similarity of represents the nth current moment φ n and similarity; The word recognition unit is used to recognize the interactive words and received English letters in the air-land communication training interactive standard training model in real time, as follows: The modal confidence parameter MAC is used to quantify the accuracy of English letters. The specific formula is: Among them, φ1 is the basic data of the first English letter recognized at the current moment, Represents the standard basic data of the interactive speech in the standard training model of the current land-air call training interaction, φ n and The principle is the same, φ1 T and Represent φ1 and The signal is transferred. Represents the inner product of two vectors represents the first current moment φ1 and The similarity of represents the nth current moment φ n and The similarity.

7. The exercise device according to claim 6, characterized in that The bilingual switching module includes an answer timing unit, a timeout repetition unit and a bilingual switching unit; The answer timing unit is used to record the time taken by the trainee to answer in real time through a timer; The timeout repetition unit is used to set the time for the student to answer. If the time is exceeded, the call interaction content is repeated; The bilingual switching unit is used to automatically switch to Chinese interactive dialogue when the student's answering time is exceeded for the second time, so as to remind the student of the call steps and call status.

8. The exercise equipment according to claim 1, characterized in that The judging and guiding end includes a content recognition module, a pronunciation assessment module, a tone difference calculation module and a tone difference guiding module; The content identification module includes a content identification unit and an abnormality identification unit; The content recognition unit is used to recognize the interactive content in real time through the data sensor and the voice converter, and the recognition method is as follows: Step 1: Calculate the similarity between the keywords of the interactive content and the keywords of the standard interactive model. The calculation formula is as follows: Among them, Y sim (Z, V0) represents the similarity function between the keywords of the current interactive content and the keywords of the standard interactive model, Z represents the word feature vector of the retrieved standard interactive model keywords, S(Z, D) represents the similarity correlation between the keywords of the current interactive content and the keywords of the standard interactive model, α represents the first weight coefficient, β represents the second weight coefficient, V0 represents the standard interactive model keyword feature vector, C i The cluster feature vector representing the keyword of the ith retrieved interactive content and the cluster set of the standard interactive model keywords, M is the number of cluster sets, i and M are positive integers; Step 2: If the calculated Y sim The closer the (Z, V0) value is to 1, the higher the similarity of the interaction content is, and the higher the accuracy of the corresponding interaction content call is; The abnormality identification unit is used to record the abnormality of the interactive content in real time through color identification.

9. The exercise device according to claim 8, characterized in that The pronunciation assessment module includes a pitch recognition unit, and the pitch recognition unit includes a determination factor unit, and the determination factor unit is used to make pronunciation judgment according to one of English phonetic symbols, vowels and consonants; The tone difference calculation module includes a tone difference comparison unit, which is used to calculate the high and low difference of English tone differences in real time through an English pronunciation sensor. If the difference = 0, it is determined that the pronunciation is correct, and if the difference ≠ 0, it is determined that there is a tone difference in the pronunciation.

10. The exercise device according to claim 9, characterized in that The tone difference guidance module includes a tone difference guidance unit, a guidance refinement unit and a recording playback unit; The tone difference guidance unit is used to provide conversation guidance to the trainee in real time based on the tone difference comparison result and the interaction content recognition result; The guidance refinement unit is used to set a tone difference guidance scheme and an interaction content guidance scheme according to the tone difference comparison result and the interaction content recognition result, and to give voice reminders for the student's pronunciation; The recording and playback unit is used to set the USB interface and the data recorder to realize recording and playback and call training download.

Images