A smart voice management system for a scenic spot self-service robot

Abstract

The present application relates to the technical field of intelligent robots, and particularly relates to a wisdom voice management system of a scenic spot self-service robot, which comprises: a control terminal, which is a master control terminal of the system and is used for issuing execution commands; an input module, which is used for inputting dialogue questions and corresponding reply information of the dialogue questions; a collection module, which is used for collecting audio information issued by a user in real time; and a conversion module, which is used for obtaining the audio information collected by the collection module, and converting the audio information into text information. The wisdom voice management system provides answers to consultation questions for the user in a manner of inputting dialogue questions and corresponding reply information of the dialogue questions. In the process of answering, the system can independently query the best reply information to feed back to the user, and can also independently capture corresponding dialogue questions with inaccurate reply information at the same time, and then input and edit corresponding reply information of the dialogue questions, so that the dialogue capability of the dialogue robot is further enhanced and improved, and the use scene requirements are met.

Description

Technical Field

[0001] This invention relates to the field of intelligent robot technology, specifically to a smart voice management system for a self-service robot in a scenic area. Background Technology

[0002] Voice robots are a deep learning-based algorithm application that can accurately recognize user intent and automatically respond, enabling self-service transactions. By using this technology, businesses can free up manpower, improve operational efficiency, and reduce labor costs. The automatic response and transaction processing capabilities of voice robots are crucial for enhancing service levels and efficiency. Furthermore, this technology allows businesses to optimize service processes, leading to higher customer satisfaction. Voice robots can also be integrated with existing systems to achieve intelligent recognition and analysis functions, and they are commonly used in tourist attractions and scenic spots.

[0003] However, voice robots used in scenic areas often answer users' questions using a pre-defined response logic and lack learning capabilities, which limits their functionality and limits their ability to answer users' questions to a limited range. Summary of the Invention

[0004] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a smart voice management system for scenic area self-service robots, which solves the technical problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A smart voice management system for self-service robots in scenic areas includes: The control terminal is the main control terminal of the system, used to issue execution commands; the input module is used to input dialogue questions and corresponding response information; the acquisition module is used to acquire audio information sent by users in real time; the conversion module is used to acquire the audio information acquired by the acquisition module and convert the audio information into text information; the configuration module is used to traverse the text information converted from audio information in the conversion module and configure the response information entered in the input module based on the text information; the monitoring module is used to monitor the number of times the configuration module runs. The control terminal is electrically connected to an input module via a medium. The input module is electrically connected to a storage unit and an addition unit via a medium. The input module is electrically connected to a data acquisition module via a medium. The data acquisition module is electrically connected to an enhancement unit via a medium. The data acquisition module is electrically connected to a conversion module, a configuration module, and a monitoring module via a medium. The data acquisition module is electrically connected to the configuration module via a medium. The conversion module is electrically connected to the addition unit via a medium. The configuration module is electrically connected to the configuration unit via a medium.

[0006] Furthermore, the input module has sub-modules at its lower level, including: The storage unit is used to receive and store the dialogue questions and corresponding response information entered in the input module. The addition unit is used to add dialogue questions and their corresponding responses, and to feed the dialogue questions and their corresponding responses back to the storage unit. When the storage unit receives the dialogue question and the corresponding answer information from the adding unit, it simultaneously stores the dialogue question and the corresponding answer information.

[0007] Furthermore, the acquisition module runs in real time within the system, and the acquisition module internally includes sub-modules, including: The enhancement unit is used to receive audio information emitted by the user collected by the acquisition module and to enhance the audio information. The enhancement unit applies any one of the following speech enhancement algorithms to enhance the audio information: speech enhancement algorithm based on spectral subtraction, speech enhancement algorithm based on wavelet analysis, speech enhancement algorithm based on Kalman filtering, enhancement method based on signal subspace, speech enhancement method based on auditory masking effect, speech enhancement method based on independent component analysis, and speech enhancement method based on neural network.

[0008] Furthermore, during the operation of the conversion module, after converting the audio information into text information, the text information is synchronously fed back to the adding unit. The adding unit uses the received text information as a similarity comparison target and compares it with each set of dialogue questions stored in the storage unit. The text information corresponding to the similarity comparison target with a similarity of no more than 80% is used as the dialogue question fed back to the storage unit by the adding unit. The conversion module sends text information back to the adding unit, which includes the text information of the added dialogue question and the corresponding response information.

[0009] Furthermore, when the adding unit sends feedback on the dialogue question to the storage unit based on the conversion module, the input module runs again, and the system user edits and uploads the corresponding response information for the dialogue question in the input module.

[0010] Furthermore, when performing similarity comparison of dialogue questions, the conversion module applies any one of the following algorithms to complete the similarity calculation of dialogue questions: edit distance algorithm, Jaccard coefficient algorithm, word frequency-inverse document frequency algorithm, WordVec algorithm, and cosine similarity algorithm.

[0011] Furthermore, during the configuration module's operation phase, the storage unit in the sub-module set below the input module is used as the configuration target, and the configuration of the text information corresponding to the response information entered in the input module is completed in the storage unit. During the configuration module's operation phase, it synchronously acquires the similarity comparison of dialogue questions in the conversion module, and uses the corresponding response information of the dialogue questions with the best similarity in the similarity comparison results as the configuration information.

[0012] Furthermore, after the configuration module finishes running, the acquisition module is synchronously switched to run again. While the acquisition module is running again, the conversion module runs synchronously and performs a similarity comparison on the text information obtained from the two consecutive conversions. When the similarity comparison result is 100%, the configuration module runs in the background. The dialogue questions corresponding to the response information other than the response information configured in the previous run are used to form a dataset, and a similarity comparison is performed with the converted text information. The response information corresponding to the dialogue question with the highest similarity in the dataset is used as the configuration information. In this module, after the configuration module configures the response information entered in the text information input module, the robot will broadcast the response information via voice.

[0013] Furthermore, when the monitoring module detects that the configuration module has run continuously for no less than three times based on a set of audio information, the text information obtained by the conversion module from the audio information is transmitted to the adding unit again.

[0014] Compared with known public technologies, the technical solution provided by this invention has the following beneficial effects: This invention provides a smart voice management system for a self-service robot in a scenic area. During operation, the system provides answers to user inquiries by recording dialogue questions and corresponding responses. In the process of answering questions, the system can autonomously search for the best response and provide feedback to the user. Simultaneously, it can also autonomously capture dialogue questions with inaccurate responses and then record and edit the corresponding responses, thereby further enhancing and improving the dialogue capabilities of the robot to meet the needs of the usage scenario. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0016] Figure 1 This is a schematic diagram of the structure of an intelligent voice management system for a self-service robot in a scenic area. The numbers in the diagram represent: 1. Control terminal; 2. Input module; 21. Storage unit; 22. Adding unit; 3. Acquisition module; 31. Enhancement unit; 4. Conversion module; 5. Configuration module; 6. Monitoring module. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0018] The present invention will be further described below with reference to embodiments. Example 1

[0019] This embodiment describes a smart voice management system for a self-service robot in a scenic area, such as... Figure 1 As shown, it includes: Control terminal 1 is the main control terminal of the system, used to issue execution commands; input module 2 is used to input dialogue questions and corresponding response information; acquisition module 3 is used to acquire audio information sent by users in real time; conversion module 4 is used to acquire the audio information acquired by acquisition module 3 and convert the audio information into text information; configuration module 5 is used to traverse the text information converted from audio information in conversion module 4 and configure the response information entered in input module 2 based on the text information; monitoring module 6 is used to monitor the number of times configuration module 5 runs. Input module 2 has sub-modules, including: Storage unit 21 is used to receive the dialogue questions and corresponding response information entered in input module 2, and to store the dialogue questions and corresponding response information. Adding unit 22 is used to add dialogue questions and corresponding answer information, and to feed the dialogue questions and corresponding answer information back to storage unit 21. When the storage unit 21 receives the dialogue question and the corresponding answer information from the addition unit 22, it simultaneously stores the dialogue question and the corresponding answer information. The data acquisition module 3 runs in real time within the system. Internally, the data acquisition module 3 contains sub-modules, including: The enhancement unit 31 is used to receive the audio information emitted by the user collected by the acquisition module 3 and to enhance the audio information. Among them, the enhancement unit 31 applies any one of the following speech enhancement algorithms to enhance the audio information: speech enhancement algorithm based on spectral subtraction, speech enhancement algorithm based on wavelet analysis, speech enhancement algorithm based on Kalman filtering, enhancement method based on signal subspace, speech enhancement method based on auditory masking effect, speech enhancement method based on independent component analysis, and speech enhancement method based on neural network. The control terminal 1 is electrically connected to the input module 2 via a medium. The input module 2 is electrically connected to the storage unit 21 and the addition unit 22 via a medium. The input module 2 is electrically connected to the acquisition module 3 via a medium. The acquisition module 3 is electrically connected to the enhancement unit 31 via a medium. The acquisition module 3 is electrically connected to the conversion module 4, the configuration module 5 and the monitoring module 6 via a medium. The acquisition module 3 is electrically connected to the configuration module 5 via a medium. The conversion module 4 is electrically connected to the addition unit 22 via a medium. The configuration module 5 is electrically connected to the configuration unit 21 via a medium.

[0020] In this embodiment, the control terminal 1 controls the input module 2 to input dialogue questions and corresponding answers. The storage unit 21 receives the dialogue questions and corresponding answers input by the input module 2 in real time and stores them. The adding unit 22 adds dialogue questions and corresponding answers synchronously and feeds them back to the storage unit 21. The acquisition module 3 runs in real time to acquire audio information sent by the user. The enhancement unit 31 receives the audio information acquired by the acquisition module 3 and enhances it. The conversion module 4 then acquires the audio information acquired by the acquisition module 3 and converts it into text information. The configuration module 5 runs through the text information converted from the audio information in the conversion module 4 and configures the answers input by the input module 2 based on the text information. Finally, the monitoring module 6 monitors the number of times the configuration module 5 runs. Example 2

[0021] At the implementation level, based on Example 1, this example refers to... Figure 1 The intelligent voice management system for a scenic area self-service robot in Example 1 will be further described in detail below: During the operation of the conversion module 4, after converting the audio information into text information, the text information is synchronously fed back to the adding unit 22. The adding unit 22 uses the received text information as a similarity comparison target and compares it with each set of dialogue questions stored in the storage unit 21. The text information corresponding to the similarity comparison target with a similarity of no more than 80% is used as the dialogue question fed back to the storage unit 21 by the adding unit 22. Among them, the text information fed back from the conversion module 4 to the adding unit 22 is the text information that the adding unit 22 runs to add dialogue questions and corresponding reply information for the dialogue questions; When the adding unit 22 sends feedback on the dialogue question to the storage unit 21 based on the conversion module 4, the input module 2 runs again, and the system user edits and uploads the corresponding answer information for the dialogue question in the input module 2; When performing similarity comparison of dialogue questions, the conversion module 4 applies any one of the following algorithms to complete the similarity calculation of dialogue questions: edit distance algorithm, Jaccard coefficient algorithm, word frequency-inverse document frequency algorithm, Word2Vec algorithm, and cosine similarity algorithm.

[0022] In this embodiment, the above settings provide further operational logic support for the operation of the system in Embodiment 1. Example 3

[0023] At the implementation level, based on Example 1, this example refers to... Figure 1 The intelligent voice management system for a scenic area self-service robot in Example 1 will be further described in detail below: During the operation phase of the configuration module 5, the storage unit 21 in the sub-module set by the input module 2 is used as the configuration target, and the configuration of the text information corresponding to the reply information entered in the input module 2 is completed in the storage unit 21. During the operation phase of the configuration module 5, the similarity comparison of the dialogue questions in the conversion module 4 is obtained synchronously, and the corresponding response information of the dialogue questions with the best similarity in the similarity comparison results is used as the configuration information.

[0024] like Figure 1 As shown, after the configuration module 5 finishes running, the acquisition module 3 runs again synchronously. While the acquisition module 3 is running again, the conversion module 4 runs synchronously and compares the similarity of the text information obtained from the two consecutive runs. When the similarity comparison result is 100%, the configuration module 5 runs in the background. The dialogue questions corresponding to the response information other than the response information configured in the previous run are used to form a dataset, and the similarity is compared with the converted text information. The response information corresponding to the dialogue question with the highest similarity in the dataset is used as the configuration information. In this module, after the configuration module 5 configures the response information entered in the text information input module 2, the robot will broadcast the response information via voice.

[0025] like Figure 1 As shown, when the monitoring module 6 detects that the configuration module 5 has been running continuously for no less than three times based on a set of audio information, the text information obtained by the conversion module 4 is transmitted to the addition unit 22 again.

[0026] In this embodiment, the above settings provide further operational data support for the operation of the system in Embodiment 1, and provide security for the stable operation of the system.

[0027] In summary, during operation, the system in the above embodiments provides answers to users' questions by inputting dialogue questions and corresponding response information. During the answering process, it can autonomously query the best response information and provide feedback to the user. Simultaneously, it can also autonomously capture dialogue questions with inaccurate responses and then input and edit the corresponding response information for the dialogue questions, thereby further enhancing and improving the dialogue capabilities of the chatbot to meet the needs of the usage scenario.

[0028] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart voice management system for a self-service robot in a scenic area, characterized in that, include: The control terminal (1) is the main control terminal of the system, used to issue execution commands; The input module (2) is used to input dialogue questions and corresponding answers to the dialogue questions; The acquisition module (3) is used to acquire audio information emitted by the user in real time; The conversion module (4) is used to acquire the audio information acquired by the acquisition module (3) and convert the audio information into text information; The configuration module (5) is used to traverse the text information converted from audio information in the conversion module (4) and configure the reply information entered in the input module (2) based on the text information. The monitoring module (6) is used to monitor the number of times the configuration module (5) runs; The input module (2) has sub-modules at its lower level, including: Storage unit (21) is used to receive the dialogue questions and corresponding answer information entered in the input module (2) and store the dialogue questions and corresponding answer information. The addition unit (22) is used to add dialogue questions and corresponding response information, and to feed back the dialogue questions and corresponding response information to the storage unit (21); When the storage unit (21) receives the dialogue question and the corresponding answer information from the addition unit (22), it simultaneously stores the dialogue question and the corresponding answer information. During the operation phase of the conversion module (4), after converting the audio information into text information, the text information is synchronously fed back to the adding unit (22). The text information received by the adding unit (22) is used as the similarity comparison target and compared with the dialogue questions stored in the storage unit (21). The text information corresponding to the similarity comparison target with a similarity of no more than 80% is used as the dialogue question fed back to the storage unit (21) by the adding unit (22). Among them, the text information fed back by the conversion module (4) to the adding unit (22) is the text information that the adding unit (22) runs to add dialogue questions and corresponding answer information for the dialogue questions; When the adding unit (22) sends feedback on the dialogue question to the storage unit (21) based on the conversion module (4), the input module (2) runs again, and the system user edits and uploads the corresponding reply information of the dialogue question in the input module (2); During the operation phase of the configuration module (5), the storage unit (21) in the sub-module set under the input module (2) is used as the configuration target, and the configuration of the text information corresponding to the reply information entered in the input module (2) is completed in the storage unit (21); During the configuration module (5) operation phase, the similarity comparison of the dialogue questions in the conversion module (4) is obtained synchronously, and the corresponding answer information of the dialogue questions with the best similarity in the similarity comparison results is used as the configuration information. After the configuration module (5) finishes running, the acquisition module (3) will run again. When the acquisition module (3) is running again, the conversion module (4) will run synchronously and perform a similarity comparison on the text information obtained from the two consecutive runs. When the similarity comparison result is 100%, the configuration module (5) will run in the next step. The dialogue questions corresponding to the response information other than the response information configured in the previous run will form a dataset and perform a similarity comparison with the converted text information. The response information corresponding to the dialogue question with the highest similarity in the dataset will be used as the configuration information. In this process, after the configuration module (5) inputs the reply information based on the text information in the input module (2), the robot will broadcast the reply information by voice. When the monitoring module (6) detects that the configuration module (5) has been running continuously for no less than three times based on a set of audio information, the text information obtained by the conversion module (4) is transmitted to the addition unit (22) again.

2. The intelligent voice management system for a scenic area self-service robot according to claim 1, characterized in that, The acquisition module (3) runs in real time within the system. The acquisition module (3) contains sub-modules, including: The enhancement unit (31) is used to receive the audio information sent by the user collected by the acquisition module (3) and to enhance the audio information. The enhancement unit (31) applies any one of the following speech enhancement algorithms to enhance audio information: speech enhancement algorithm based on spectral subtraction, speech enhancement algorithm based on wavelet analysis, speech enhancement algorithm based on Kalman filtering, enhancement method based on signal subspace, speech enhancement method based on auditory masking effect, speech enhancement method based on independent component analysis, and speech enhancement method based on neural network.

3. The intelligent voice management system for a scenic area self-service robot according to claim 2, characterized in that, When performing similarity comparison of dialogue questions, the conversion module (4) applies any one of the following algorithms to complete the similarity calculation of dialogue questions: edit distance algorithm, Jaccard coefficient algorithm, word frequency-inverse document frequency algorithm, Word2Vec algorithm and cosine similarity algorithm.

Images