KR102603035B1 - AI-based virtual chatbot operating system - Google Patents

Abstract

The AI-based virtual chatbot operating system according to the present invention supports chatting with users, and includes a text processor that converts the user's voice into text based on STT (Sound-to-Text), and a text processor that analyzes the text. It includes an analysis unit, an answer generation unit that generates different answers depending on whether conditions for specific sentences or words in the text match, and an output unit that converts the answers into voices based on TTS (Text-to-Sound) and outputs them. chatbot interface; It is characterized by comprising a character management module including a character database storing a plurality of characters, and an execution unit that outputs a character selected by a user from the character database in conjunction with the chatbot interface.
According to the AI-based virtual chatbot operating system according to the present invention, it has the effect of providing users with a chatbot operating system that provides a more intimate communication environment as if communicating with a person through a virtual character.

Description

AI-based virtual chatbot operating system

The present invention relates to an AI-based virtual chatbot operating system. More specifically, when implementing an artificial intelligence-based chatbot interface in a voice output method, a virtual character is provided, and a voice is output through the virtual character, as well as the virtual character. It is about a virtual chatbot operating system that can double the fun of users by incorporating unique means to foster them.

Chatbot is a type of conversational artificial intelligence and refers to a bot that communicates with users through messengers. There are a variety of chatbots, ranging from simple chatbots that output speech by simply entering a message according to established rules to chatbots that analyze the other person's speech and produce speech that is close to artificial intelligence.

In the past, chatbots merely sent messages based on rules, but since the Google DeepMind Challenge Match in 2016 and the emergence of AlphaGo, attempts to incorporate AI technology into chatbots have been actively made. In addition, companies have shown great interest, and many consultation chatbots for various banks, insurance companies, etc. have appeared.

There are also a number of such chatbot-related patents. Domestic Patent No. 2230372, a chatbot device using automatic question generation and its operation method, analyzes the first utterance input by the user and, if the first utterance is a question, pre-generated learning knowledge. The chatbot device generates the first response of the chatbot device to the first utterance, and if the first utterance is not a question, predicts the user's second utterance using learning knowledge and induces the predicted second utterance. It includes generating a second response and providing a first response or a second response of the chatbot device generated as a response to the first utterance, and the learning knowledge is obtained by performing machine learning on a database built from the corpus. By being created, it is revealed that knowledge/language learning oriented towards scalable and logical flow is possible.

However, not only the above technologies, but also many chatbot services follow the method of simply outputting only voice or text to the user, which not only makes it easy to feel boring, but also creates an environment where continuous communication is difficult because you can feel like you are communicating with a machine. .

Therefore, there is a growing need to develop a new and advanced chatbot operating system that performs various communication functions such as answering questions and schedule management through artificial intelligence-based chatbots, but provides a more intimate communication environment like communicating with people. am.

Domestic Patent No. 2230372

The present invention was developed to overcome the problems of the above technology. It performs communication functions such as answering various questions and schedule management through an artificial intelligence-based chatbot, but provides more intimate communication as if communicating with a person through a virtual character. The main purpose is to provide a chatbot operating system that provides an environment.

Another object of the present invention is to provide a unique and diverse means for users to grow and develop their characters.

Another purpose of the present invention is to provide a security measure to prevent personal information from being leaked when a character's answer voice is output and the voice is recorded and leaked to the outside.

In order to achieve the above purpose, the AI-based virtual chatbot operating system according to the present invention supports chatting with users, and includes a text processing unit that converts the user's voice into text based on STT (Sound-to-Text). an analysis unit that analyzes the text, an answer generation unit that generates different answers depending on whether conditions for specific sentences or words in the text match, and the answers are converted into voices based on TTS (Text-to-Sound). Chatbot interface including an output unit that converts and outputs; It is characterized by comprising a character management module including a character database storing a plurality of characters, and an execution unit that outputs a character selected by a user from the character database in conjunction with the chatbot interface.

In addition, the system includes a counting module that generates a counting index by counting at least one of the number of voices of the user and the number of words expressing the user's satisfaction, and the character management module is configured to generate the counting index according to the counting index. It is characterized by including a character breeding unit that varies at least one of the size and shape of the character.

In addition, the chatbot interface is implemented in a mobile communication terminal, and the mobile communication terminal includes a microphone, a directional speaker that outputs the voice converted by the output unit toward the microphone, and a device configured to cover the microphone and the directional speaker. A security control interface including a cover member and a controller that generates a howling while outputting the recorded voice through the directional speaker, synthesizes the howling into the recorded voice, and stores the recorded voice.

According to the AI-based virtual chatbot operating system according to the present invention,

1) It has the advantage of providing users with a chatbot operating system that provides a more intimate communication environment as if communicating with a person through a virtual character,

2) The fun that users can have in raising and nurturing their characters can double the enjoyment.

3) By modifying and storing the character's output voice, security can be strengthened through methods such as forensics.

1 is a conceptual diagram showing the basic configuration of the system of the present invention.
Figure 2 is a block diagram showing the detailed configuration of the system of the present invention.
Figure 3 is a conceptual diagram showing various characters stored in the character database of the present invention.
Figure 4 is a conceptual diagram showing various facial expressions of the character of the present invention.
Figure 5 is a conceptual diagram showing a security control interface included in a mobile communication terminal in which a chatbot interface is implemented.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. The accompanying drawings are not drawn to scale, and like reference numbers in each drawing refer to like elements.

1 is a conceptual diagram showing the basic configuration of the system of the present invention.

As can be seen from Figure 1, the system of the present invention can be implemented through the relationship between the user terminal 10 and the main server 20 owned by the user, that is, the user terminal 10 and the main server 20 ) is possible to include.

First, the user terminal 10 can be composed of a mobile device capable of communication, that is, a smartphone, or, as will be described later, a mobile communication terminal equipped with a security control interface with enhanced security (a smartphone is also possible). .

This user terminal 10 is installed with a chatbot interface 100 supported in the form of an app application, which provides the function of performing AI-based chatting to be described later while communicating with the main server 20, and additionally provides a security control interface 400. ) can be installed.

The main server 20 communicates with the user terminal 10 and plays a central role in managing and operating the chatbot interface 100.

That is, in the system of the present invention, the main server 220 is the entity that outputs the virtual character (character) on the display (touch screen) of the user terminal 10 as well as AI-based chat. ) is a series of entities for implementing the system of the present invention and includes a server PC and a network communication network.

In addition, the main server 20 is based on hardware equipped with a central processing unit (CPU) and storage means such as memory and hard disk, and a program that can be executed on the central processing unit, that is, software, is installed and can execute this software. A series of specific configurations for the software will be described later as structural units such as 'module', 'part', and 'part'.

These configurations such as 'modules' or 'parts' or 'interfaces' or 'parts' are hardware such as software or FPGA or ASIC that runs through CPU and memory while being installed and stored in the storage means of the main server 20. refers to the work composition of . At this time, the meaning of 'module', 'unit', or 'interface' is not limited to hardware, and may be configured to reside in an addressable storage medium or to reproduce one or more processors.

As an example, 'module' or 'part' or 'interface' refers to components such as software components, object-oriented software components, class components and task components, processes, functions, and properties. Contains fields, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

The functions provided by these 'modules' or 'parts' or 'interfaces' can be combined into smaller numbers of components and 'parts' or 'modules' or can be divided into additional components and 'parts' or 'modules'. Could be further separated.

In addition, the main server 20 refers to all types of hardware devices including at least one processor, and depending on the embodiment, it may be understood as encompassing software configurations operating on the hardware device. For example, a computing device as an example of a server may be understood as including, but is not limited to, a smartphone, tablet PC, desktop, laptop, and user clients and applications running on each device.

Hereinafter, based on this configuration, the system of the present invention will be described with the attached drawings.

Figure 2 is a block diagram showing the detailed configuration of the system of the present invention.

Referring to Figure 2, it can be seen that the system of the present invention is based on including a chatbot interface 100 and a character management module 200.

The chatbot interface 100 of the present invention has a basis for providing a virtual avatar, that is, a virtual character (hereinafter referred to as 'character'), that can chat with the user in conjunction with the character management module 200. At the same time, it provides a chat function that supports search functions as well as communication with users.

This chatbot interface 100 provides a so-called virtual assistant service, but in this virtual assistant service, a character is visually provided and provides a function such as a so-called 'virtual chatbot' in which the character is having a conversation with the user.

In other words, unlike a typical chatbot, the chatbot interface 100 of the present invention provides a conversation function with a character supported by a GUI in a virtual space (e.g., a messenger running in a virtual space), and provides a conversation function with a character supported in a virtual space GUI (e.g., a messenger running in a virtual space), and supports normal communication, that is, daily life. It can perform both conversation and AI-based search functions.

This chatbot interface 100 can incorporate voice recognition technology, and specifically includes a text processing unit 110, an analysis unit 120, an answer generation unit 130, and an output unit 140.

The text processing unit 110 provides a function to convert the user's voice into text based on STT (Sound-to-Text).

At this time, the text processing unit 110 can provide not only the user's voice but also a function that allows the user to input sentences or characters through the keypad of the user terminal 10, making it convenient for the user to use the chatbot interface 100. As an intermediary, you can have a conversation or input text with a character that will be described later.

In addition, the text processing unit 110 can provide a function to record the user's voice in order to accurately recognize the user's voice, which provides a basis for analyzing the recorded voice file in more detail in the analysis unit to be described later. Alternatively, it is possible to create an environment where the user can communicate with the character with more accurate pronunciation by playing the recording file once so that the user can accurately understand the intention of the conversation while listening to the user's own voice.

The analysis unit 120 performs the function of analyzing text obtained from the user's voice. In other words, its main function is to analyze the meaning of the text.

This analysis unit 120 may apply various language analysis algorithms based on computational technology used to process, analyze, understand, and generate human language, including natural language processing (NLP), a field of artificial intelligence. It can have functions.

The answer generator 130 provides a function to generate different answers depending on whether or not a condition for a specific sentence or word in the text matches, and this may also be based on the previously mentioned natural language processing (NLP) function. You can.

For example, the analysis unit 120 and the answer generation unit 130 use Bag of Words (BoW), a simple and commonly used model in NLP, and erm Frequency-Inverse Document Frequency (TF-IDF), which is frequently used in text mining. , Word2Vec, which can capture the semantic meaning of words by expressing them as high-dimensional vectors, GPT (Generative Pre-trained Transformer), which is trained to predict the next word in a sentence when the previous word is presented while using lip learning technology, and The same algorithm or program can be linked/applied.

In other words, the answer generator 130 not only searches for information about the user's question and provides an answer, but also has the function of having a daily conversation with the user. In addition, it receives the user's schedule in advance and then generates an answer at a specific time. It is also possible to provide a so-called schedule management function that informs the schedule, or to analyze the user's voice and generate a schedule related to the analyzed content as a kind of answer.

The output unit 140 converts the answer generated by the answer generation unit 130 into voice based on TTS (Text-to-Sound) and outputs it through a speaker (not shown) mounted on the user terminal 10. provides.

In other words, the chatbot interface 100 of the present invention can provide a function such as a kind of user-only assistant by applying a language analysis model/algorithm trained and learned based on artificial intelligence (AI) or Lipnering, and this chatbot interface Since (100) is the same or similar to the function of the currently known artificial intelligence-based chatbot service platform, a separate detailed explanation is omitted.

Figure 3 is a conceptual diagram showing various characters stored in the character database of the present invention.

The character management module 200 of the present invention goes beyond simply outputting only voice for the chatbot service described above and can have a visual output function in addition to auditory output, such as when a virtual avatar, that is, a virtual character, speaks. Specifically, the character management module 200 It is based on including a database 210 and an execution unit 220.

As can be seen from FIG. 3, the character database 210 stores a plurality of characters with various shapes and characteristics. At this time, the character database 210 not only stores various types of characters, but also allows for variable processing in one character. Various shapes can be saved together.

For example, based on the character A, various derived characters can be stored dependently by size, clothing worn, facial expression, etc., and the form of the character can be stored in a moving video format, 2D or 2D as well as a still image. It can be expressed in various forms such as 3D content.

These characters can be purchased in paid/free form through the character online store, and it is possible to give a separate name to the character selected by the user. For example, if the character name is "JUNE", It is also possible that when the user calls "Jun~", the character responds and is ready to chat with the user.

The execution unit 220 provides a function to output a character selected by the user from the character database 210 in conjunction with the chatbot interface 100.

In other words, when the chatbot interface 100 outputs a response to the user's voice, the character is visually exposed and the response is output as if the character is answering, as well as the overall chat process of listening to or responding to the user's voice. It is also possible to process it so that it is always exposed.

In summary, the system of the present invention provides users with a chatbot that can provide secretarial services based on artificial functions based on a virtual chatbot. Unlike chatbots that simply provide auditory voice answers, the system provides auditory voice through visual images. It has the characteristic of being able to provide chat/search functions in a fun and interesting environment without being boring by printing it out.

As explained earlier, the above-mentioned characters can not only be selected from the character database 210, but also various sizes and expressions can be stored based on one character, just as the user is raising the character, that is, breeding. It is possible to provide a similar environment.

To this end, specifically, the system of the present invention may additionally include a counting module 300 that generates a counting index by counting at least one of the number of user's voices and the number of words indicating user's satisfaction.

Here, the number of times the user's voice refers to the number of times the user runs the chatbot interface 100, calls the character, and then chats with the character, and the number of times the user expresses the user's satisfaction is "(character name), it was good." This refers to the number of times a user vocally expresses words with a meaning related to satisfaction, such as “Good job!” and “Wow, that’s great.”

In other words, it can be understood that the counting module 300 counts the number of times a user frequently encounters a character or expresses an intimate feeling and performs its function.

In addition, the counting index can be generated as a number proportional to the above-mentioned counting number. For example, if the number of voices is 10 and the number of satisfaction word expressions is 5, any one of them, number 10, is taken and expressed as 10, or all of them are expressed as 10. It can be reflected and expressed as the arithmetic average value of 10+5/2=7.5, or it can be derived using another arithmetic method proportional to the number of counting times.

Correspondingly, the character management module 200 may include a character breeding unit 230 that variably processes at least one of the size and shape of the character according to the counting index.

In other words, the character breeding unit 230 can give the user the feeling of growing their own character. As the counting index increases, the character's size is increased or a specific shape such as the character's hands, feet, and face is raised. or provide a function to change it.

In other words, when you say to an actual pet, ‘It’s pretty.’ Good job. Expressing satisfaction such as 'good job' can provide a process for the character to grow, making it possible to provide an environment in which the user cares more for and cares for his or her character.

Furthermore, the counting module 300 counts the number of words expressing the user's emotions in addition to the number of voices described above and the number of words expressing the user's satisfaction, and calculates, for example, their arithmetic average value as mentioned above. It is also possible to generate a counting exponent while calculating.

Words expressing the user's emotions, for example, mean words such as 'happy, grateful, proud...'. If there are, for example, 5 emotion expressing words in the user's voice, the counting module will calculate the number of words expressing the user's emotions. It is possible to count the number of words as '5'.

At this time, words expressing the user's emotions may include not only positive emotions but also negative emotions (sad, depressed...), and the counting module 300 has the function of distinguishing between positive and negative emotion words and processing them. It is possible to provide, and correspondingly, the character breeding unit 230 may reflect only the number of positive emotion expression words among the user's emotion expression words.

Additionally, the counting module 300 counts at least one of the above-described number of voices of the user, the number of words indicating the user's satisfaction, and the number of times the user expresses emotions, as well as the number of touches on the touch screen on which the character is displayed. A counting index can be created by counting and summing the results.

To this end, the premise is that the chatbot interface 100 must be implemented on a smartphone including a touch screen.

In other words, when a user touches a character through the touch screen as if stroking it, the counting module can count the number of touches and reflect this processed value in the counting index.

This reflects the intention of encouraging users to pay as much attention and sincerity to their characters as possible, in the various ways described above, such as the number of times a word with a specific meaning is used, the number of times a voice is output by calling the character, etc., as well as by touching the character. You can provide characteristics to develop your character by comprehensively reflecting the number of times you do it.

Figure 4 is a conceptual diagram showing various facial expressions of the character of the present invention.

Referring to FIG. 4, a character can make a specific facial expression, and the character management module 200 can store characters with various facial expressions in the character database 210 and then load and output characters with specific facial expressions. , it is possible to variably process these facial expressions if the special conditions provided by the present invention are satisfied.

Specifically, the character management module 200 may include a needs request unit 240.

The needs request unit 240 includes a needs storage part 241 that stores the character's needs, and an alarm part 242 that outputs words related to the needs as a voice to the user at a certain cycle.

At this time, the character's needs are either due to the individual characteristics of the user or related to specific conditions required in the process of raising the character like a pet. For example, in conjunction with a schedule management application, the needs are specified according to the user's schedule. Calling the character by voice at the right time, touching the touch screen on which the character is displayed n times per day, and writing a word with a specific meaning such as 'I had a hard time' to the character n times per day. Speaking by voice means speaking a voice containing a specific word or touching a touch screen to a character under certain conditions, such as calling a character by voice when waking up in the morning.

These needs may be preset in the needs storage part 241 or the user may set the needs directly, or further, they may be individually set for each user by identifying the user's behavior or life patterns based on artificial intelligence.

The alarm part 242 provides a function to output words related to the above-mentioned needs as a voice to the user at a certain period.

In other words, the user may be able to fulfill the above-mentioned needs on their own, but in case this is not the case, 'please pat me once' every three hours, for example, at 9:00 AM, 12:00 PM, and 3:00 PM. It is possible to output a voice to the user that contains words that mean the character is requesting a touch or that have a consoling meaning, such as 'Please tell me I've had a hard time'.

In response to this, the counting module 300 may provide a function to generate a needs satisfaction index by counting words related to the need among the user's voice.

In other words, the counting module 300 checks/counts how well the user complied with the needs request of the alarm part 242 and spoke words related to the needs by voice, and generates a needs satisfaction index proportional to the number of checks/counts. is to carry out.

In addition, the character breeding unit 230 provides a function for variable processing of the character's facial expression according to the above-described needs satisfaction index.

In other words, the higher or larger the needs satisfaction index is, the more colorful and diverse the character's facial expression can be. In this case, the character's facial expression can be happy and happy rather than sad or depressed.

In summary, the system of the present invention does not just passively wait for a specific voice or action from the user, but actively outputs the needs of the character to the user, allowing the user to say actions or voices related to the needs, thereby creating a connection with the character. It can not only strengthen communication but also double the pleasure of developing characters.

As previously explained, it may be more desirable to ensure the user's convenience if the chatbot interface 10 is implemented in a mobile communication terminal such as a smartphone that the user always carries.

At this time, it is possible to further include a mobile communication terminal, specifically the circuitry of the mobile communication terminal, including a security control interface 400.

Specifically, this security control interface includes a microphone 410, a directional speaker 420, a cover member 430, and a controller 440.

The microphone 410 provides a basis for generating howling, and is supported to generate howling by allowing the character's voice output through the directional speaker 420 to be input again through the microphone 410.

The directional speaker 420 provides the function of outputting sound, specifically the voice converted by the output unit 140, toward the microphone 410. It may be a conventional directional speaker or a microphone 410 with the arrangement of the speaker described in detail. ), and by allowing the speaker's voice to be output toward the microphone 410, the function as a directional speaker 420 can be implemented.

In particular, the directional speaker 420 is provided separately from the speaker that outputs the voice converted by the output unit 140 to reduce the difference between the user's ability to actually hear the converted voice through the speaker and the function for generating howling. put it

Furthermore, the sound output from the directional speaker 420 is not necessarily limited to the voice of the output unit, but can be any variety of sound to generate howling, so it is of course possible to include general sounds such as music and specific sounds. However, since the voice converted by the output unit already exists, it can be said to be more efficient to use it to output it from a directional speaker.

The surface of the directional speaker 420 and the microphone 410 is wrapped and sealed through a cover member 430, and the cover member 430 covers the entire microphone 410 and the directional speaker 420. It is configured to have a function to prevent leakage to the outside of the mobile communication terminal.

The controller 440 controls the operation of the directional speaker 420 to output conversation information through the directional speaker 420, and also controls the operation of the microphone 410 to output voice (output) through the directional speaker 420. It performs the function of causing howling as a result by allowing the negative voice) to be re-input through the microphone 410.

In addition, when howling occurs, it is synthesized again into the voice of the output unit and stored in an overwrite manner, thereby performing a function of removing the original voice of the output unit 140 and storing only the modified version with the howling overlaid.

According to the configuration of the security control interface 400, the voice of the output unit is output while the directional speaker 420 is facing the microphone 410, and the voice of the output unit 140 is output (voice or sound of the output unit). As the sound is input again through the microphone 410, the input voice is amplified and output again to the directional speaker 420, creating a circular loop in which sound is absorbed again through the microphone 410, and a specific frequency is significantly amplified, resulting in howling. Occurs.

By synthesizing this howling into the voice of the output unit and storing it, the content of the voice of the output unit is modulated so that it is unrecognizable. When the voice of the output unit is simply output and then deleted, the voice may remain in the temporary file storage or may be used in the forensic process. It is possible to further strengthen the user's individual security without causing security problems such as restoring the voice of the deleted output unit.

Furthermore, this circulation of sound occurs to a small extent in the security control interface 400, but the directional speaker 420 and the cover member 430 are wrapped through the cover member 430, so the sound is not exposed to the outside, so the directional speaker 420 and the cover member 430 are surrounded by the cover member 430. Of course, it is possible to prevent the sound from overlapping with the sound of the output unit output from a speaker separate from the speaker 420.

Therefore, this novel method allows modulation and encryption of the voice of the output unit 140, so that the voice of the output unit 140 is output even when it is leaked to the outside while stored (temporarily stored) in the mobile communication terminal. By making it difficult to identify the voice content of the unit 140, the function to strengthen the security of the user's personal information can be strengthened.

Furthermore, the cover member 430 can be said to be important in the configuration of this security control interface 400. The cover member 430 covers the directional speaker 420 and the microphone 410, which generate howling, to transmit sound to the outside. It is basic to prevent leakage.

For this purpose, a sound-absorbing coating layer 450 is laminated on the inner peripheral surface of the cover member 430 to prevent the sound of the output unit (voice or sound of the output unit) from leaking to the outside. This sound-absorbing coating layer 450 is preferably It may include 90 to 99 parts by weight of a sound-absorbing base containing glass fiber and 1 to 10 parts by weight of a sound-absorbing enhancer containing graphene oxide.

The sound-absorbing base includes glass fiber, which is a material widely used as a conventional sound-absorbing material and is known to have high strength and excellent sound absorption. Furthermore, it has the advantages of being resistant to high temperatures, not burning, not being absorbent, having little hygroscopicity, not corroding because it is chemically durable, having high strength, especially tensile strength, and high electrical insulation, so it has a sound-absorbing coating layer (450) containing a sound-absorbing base. ), when laminated on the inner peripheral surface of the cover member 430, not only can increase the sound absorption of the cover member 430, but also have the effect of increasing high temperature durability, chemical durability, strength, and electrical insulation.

Furthermore, the sound absorption enhancer includes graphene oxide, and it is known that when graphene oxide is included in a sound absorption material, the sound absorption performance is further strengthened, and related experimental results also prove this.

Therefore, the sound-absorbing coating layer 450, which is a mixture of the sound-absorbing enhancer containing graphene oxide and the sound-absorbing base, is laminated on the inner peripheral surface of the cover member 430, thereby strengthening the sound absorption of the cover member 430 and exposing the output portion to the outside. This has the effect of preventing voice (voice or sound from the output unit) from leaking and increasing howling generation performance by integrating sound into the inner peripheral surface of the cover member 430.

At this time, preferably, the sound absorption enhancer may further include additional compositions in addition to graphene oxide, and this sound absorption enhancer is used in the step of preparing the first solution (S11), the step of preparing the second solution (S12), and the step of preparing the tertiary solution. Step (S13), preparing a fourth solution (S14), and obtaining an intermediate complex (S15). It can be prepared through the step of preparing the 5th solution (S15), the step of preparing the 6th solution (S16), and the step of obtaining the sound absorption enhancer (S17).

(S11) Step of preparing the first solution

First, prepare a primary solution by mixing 20 to 40 parts by weight of propylene glycol and 60 to 80 parts by weight of graphene oxide (GO). Graphene oxide can become a uniform dispersion under propylene glycol, where sufficient stirring using a stirrer, etc. must be performed when preparing the primary solution.

In this case, volatilization of the solvent, propylene glycol, may occur during the stirring process. In this process, propylene glycol in the amount evaporated is added to the primary solution to maintain the weight of propylene glycol.

(S12) Step of preparing secondary solution

When the preparation of the primary solution is completed, a secondary solution is prepared by mixing 30 to 40 parts by weight of propylene glycol, 10 to 20 parts by weight of distilled water, and 40 to 60 parts by weight of silver nitrate. At this time, separate heating may be added to completely dissolve the silver nitrate. Silver nitrate is used to add silver to graphene oxide.

(S13) Step of preparing tertiary solution

Once the preparation of the primary and secondary solutions is completed, 30 to 50 parts by weight of the prepared primary solution and 50 to 70 parts by weight of the secondary solution are mixed to prepare a tertiary solution, which is a graphene-silver nitrate mixed solution.

The tertiary solution prepared through this is a silver nitrate solution in which propylene glycol is added and graphene oxide is evenly dispersed. In addition, to prepare the tertiary solution, stirring is performed under nitrogen at a temperature of 40 to 60°C for 1 to 3 hours. During this process, silver ions contained in the silver nitrate solution primarily react with graphene oxide. This makes it possible to obtain a graphene oxide complex to which silver ions are added.

(S14) Step of preparing the fourth solution

When the preparation of the tertiary solution is completed, the tertiary solution is heated to 100 to 120 ° C., 80 to 95 parts by weight of the tertiary solution and 5 to 20 parts by weight of sodium borohydride are mixed and left for 1 to 3 hours. A fourth solution is prepared by stirring.

At this time, preferably, the fourth solution is prepared by slowly adding sodium borohydride dropwise to the third solution and stirring. In this process, reduction of graphene oxide occurs by sodium borohydride. Through this, it is possible to prepare a quaternary solution in which the finally reduced graphene is dispersed in propylene glycol and distilled water.

(S15) Obtaining intermediate complex

When preparation of the quaternary solution is completed, the quaternary solution is washed with distilled water or any one of ethanol and methanol and dried at 70 to 90° C. for 20 to 40 hours to obtain a particulate intermediate composite.

This intermediate composite is a form in which silver is added to reduced graphene, and can be considered an intermediate material for manufacturing the graphene oxide-based porous material of the present invention.

(S16) Step of preparing the 5th solution

Next, a fifth solution is prepared by mixing 10 to 40 parts by weight of the prepared particulate intermediate composite with 60 to 90 parts by weight of distilled water. At this time, the intermediate complex has the property of sinking rather than being evenly dispersed in distilled water over time, so it must be stirred at a high speed, and the subsequent process must also be performed at a high speed.

(S17) Step of preparing the 6th solution

Next, 80 to 95 parts by weight of distilled water, 5 to 10 parts by weight of acrylamide (AAm: Acrylamide), 0.1 to 5 parts by weight of heptafluorobutylamide acrylamide (MBAA: N, N-Methylbis Acrylamide), 2-ketoglu A sixth solution is prepared by mixing 0.1 to 1 part by weight of 2-ketoglutaric acid.

Acrylamide is highly soluble in water, ethanol, ether, and chloroform, so it has the property of forming an aqueous solution using distilled water as a medium. When forming a hydrogel, it has high fluidity due to the steric repulsion of the hydrogel surface, and is soft and flexible. It has character.

Heptafluorobutylamide acrylamide (MBAA: N,N-Methylbis Acrylamide) is added to act as a crosslinker and curing agent, and can increase the hardness of the hydrogel produced by adding heptafluorobutylamide acrylamide.

2-ketoglutaric acid is a crystalline substance that is soluble in water and is a precursor of L-glutamic acid. In the case of examples of the present invention, it is added to promote hydrogelation through acrylamide.

In addition, when mixing acrylamide, heptafluorobutylamide, acrylamide, and 2-ketoglutaric acid, high-speed stirring using ultrasound, etc. can be performed. However, if stirring is carried out for too long, heptafluorobutylamide Care must be taken regarding this point because the carbon double bond contained in acrylamide is broken and the hydrogelation reaction does not proceed.

(S18) Step of preparing the 7th solution

Once the preparation of the 5th solution and the 6th solution is completed, 30 to 70 parts by weight of the 5th solution and 30 to 70 parts by weight of the 6th solution are mixed to prepare the 7th solution.

That is, in the present invention, by mixing the 5th solution and the 6th solution, an intermediate composite is added inside the acrylamide-based hydrogel to strengthen the mechanical, electrical, and thermal properties, and further to increase sound absorption by producing a porous composite material. It has an effect.

(S19) Obtaining a sound absorption enhancer

Finally, light with a wavelength of 300 to 400 nm is irradiated to the seventh solution in a dark room for 3 to 5 hours. Through this process, polyacrylamide is obtained and at the same time, it is possible to increase the strength of the prepared porous graphene oxide-silver composite. Additionally, the remaining unreacted acrylamide is removed through methods such as using osmotic pressure to produce the final product. The phosphorus sound absorption enhancer is finally obtained. In addition, when removing the remaining amount of acrylamide, only osmosis was used as an example in this embodiment, but of course, other known methods can be sufficiently applied to remove the remaining amount of acrylamide and solvent.

The sound absorption enhancer manufactured in this way not only has excellent sound absorption performance as it basically contains graphene oxide, but also forms an intermediate complex by adding silver to graphene oxide to increase thermal, mechanical, and electrical stability, and the intermediate manufactured in this way improves thermal, mechanical, and electrical stability. By infiltrating the composite into the acrylamide-based hydrogel to manufacture a porous composite, not only is the adhesion on the inner surface of the cover member 43 improved, but the effect of improving sound absorption through the porous structure is maximized.

Hereinafter, in order to explain the physical properties according to the configuration of the sound absorption enhancer of the present invention, the evaluation results of Examples and Comparative Examples will be compared and explained. The examples consist of Examples 1 and 2 containing the sound absorption enhancer of the present invention, and the comparative example is a conventional sound absorption material composition.

<Example 1>

100g of the primary solution was prepared by mixing 40g of propylene glycol and 60g of graphene oxide.

35g propleglycol, 20g distilled water. 100 g of secondary solution was prepared by mixing 45 g of silver nitrate.

40 g of the prepared primary solution and 60 g of the secondary solution were mixed and stirred for 2 hours in a nitrogen atmosphere at 50°C to prepare 100 g of the tertiary solution.

While the prepared tertiary solution was heated to 110°C, 10 g of sodium borohydride was mixed with 90 g of the tertiary solution and stirred for 2 hours to prepare 100 g of the fourth solution.

The prepared fourth solution was washed with distilled water and dried at 80°C for 30 hours to obtain an intermediate complex.

100g of the 5th solution was prepared by mixing 20g of the prepared intermediate complex and 80g of distilled water.

90g distilled water, 8g acrylamide, 1.5g heptafluorobutylamide acrylamide. 100g of the 6th solution was prepared by mixing 0.5g of 2-ketoglutaric acid.

100g of the 7th solution was prepared by mixing 60g of the prepared 5th solution with 40g of the 6th solution.

The 7th solution was irradiated with 350 nm light for 4 hours in a dark room, and unreacted acrylamide was removed through osmosis to obtain a sound absorption enhancer.

A sound-absorbing coating layer was prepared by mixing 90 g of glass fiber with 10 g of a sound-absorbing reinforcing agent, weaving it into a fiber form, and molding it into a 3 mm sheet shape.

The test specimen used a 9.5mm gypsum board as a conductor, and the manufactured sound-absorbing coating layer was attached with an epoxy adhesive and cured at room temperature.

<Example 2>

90 g of glass fiber was mixed with 10 g of graphene oxide, woven into a fiber form, and molded into a 3 mm sheet to prepare a sound-absorbing coating layer.

The test specimen used a 9.5mm gypsum board as a conductor, and the manufactured sound-absorbing coating layer was attached with an epoxy adhesive and cured at room temperature.

<Comparative Example 1>

Glass fiber was woven into a fiber form and molded into a 3 mm sheet shape to produce a sound-absorbing coating layer.

The test specimen used a 9.5mm gypsum board as a conductor, and the manufactured sound-absorbing coating layer was attached with an epoxy adhesive and cured at room temperature.

1. Sound absorption rate test: A sound absorption rate test was performed for a frequency of 800Hz according to KS F 2805 “Method for measuring sound absorption rate in a reverberant room”.

2. Scratch resistance evaluation: A scratch resistance test using #0000 steel wool was performed (1Kg load, 10 round trips). The test results were shown as follows.

○: No surface damage

△: There is some surface damage.

×: There is a lot of surface damage

3. Adhesion evaluation: Cross Cut Cellotape peel test was performed on the coating layer film. That is, 11 cutting lines of the coating film reaching the substrate at 1mm intervals are inserted into the coating film, horizontally and vertically, to make 100 1mm2 eyes, then attach Sellotape on top of them and quickly peel them off. After repeating this operation 5 times, the results are shown as follows. It was.

○: No peeling of the coating film.

△: When the number of delaminated buds is 1-50

×: When the number of delaminated buds is 51-100

Table 1 shows the test results. sound absorption rate scratch resistance Adhesion Example 1 0.46 ○ ○ Example 2 0.41 ○ △ Comparative Example 1 0.24 △ ×

As shown in Table 2 above, it can be seen that the compositions of Examples 1 and 2 of the present invention have superior sound absorption, scratch resistance, and adhesion compared to the comparative examples. Therefore, not only is the laminated coating film not easily separated and strong adhesion is possible. It can be seen that the sound absorption rate is superior.

In addition, through comparison of Examples 1 and 2, it can be confirmed that the composition of Example 1, which contains a more diverse composition, has superior overall physical properties compared to Example 2.

As explained so far, the configuration and operation of the AI-based virtual chatbot operating system according to the present invention have been expressed in the above description and drawings, but this is only an example and the spirit of the present invention is not limited to the above description and drawings. , Of course, various changes and modifications are possible without departing from the technical spirit of the present invention.

10: User terminal 20: Main server
100: chatbot interface 110: text processing unit
120: analysis unit 130: answer generation unit
140: output unit 200: character management module
210: character interface 220: execution unit
230: Character breeding department 240: Needs request department
241: Needs storage part 242: Alarm part
300: Counting module 400: Security control interface
410: Microphone 420: Directional speaker
430: cover member 440: controller
450: Sound-absorbing coating layer

Claims (8)

As an AI-based virtual chatbot operating system,
Supporting chatting with users, it includes a text processing unit that converts the user's voice into text based on STT (Sound-to-Text), an analysis unit that analyzes the text, and conditions for specific sentences or words in the text. A chatbot interface including an answer generation unit that generates different answers depending on whether there is a match, and an output unit that converts the answers into voices based on TTS (Text-to-Sound) and outputs them;
It includes a character management module that includes a character database storing a plurality of characters, and an execution unit that outputs a character selected by a user from the character database in conjunction with the chatbot interface,
The chatbot interface is implemented in a mobile communication terminal,
The mobile communication terminal is,
A microphone and a directional speaker that outputs the voice converted by the output unit toward the microphone, a cover member formed to cover the microphone and the directional speaker, and a howling is generated while outputting the voice recorded through the directional speaker. A security control interface including a controller that synthesizes the howling into the recorded voice and stores it,
On the inner peripheral surface of the cover member,
90 to 99 parts by weight of a sound-absorbing base containing glass fiber and a sound-absorbing coating layer containing 1 to 10 parts by weight of a sound-absorbing reinforcing agent are laminated,
The sound absorption enhancer is,
Preparing a primary solution by mixing 20 to 40 parts by weight of propylene glycol and 60 to 80 parts by weight of graphene oxide (GO);
Preparing a secondary solution containing 30 to 40 parts by weight of propylene glycol, 10 to 20 parts by weight of distilled water, and 40 to 60 parts by weight of silver nitrate;
Preparing a tertiary solution by mixing 30 to 50 parts by weight of the primary solution and 50 to 70 parts by weight of the secondary solution and stirring for 1 to 3 hours under nitrogen at a temperature of 40 to 60°C;
Heat the tertiary solution to 100 to 120°C, mix 80 to 95 parts by weight of the tertiary solution with 5 to 20 parts by weight of sodium borohydride, and then stir for 1 to 3 hours to prepare a fourth solution. steps;
Obtaining an intermediate composite by washing the fourth solution and drying it at 70 to 90° C. for 20 to 40 hours;
Preparing a fifth solution by mixing 10 to 40 parts by weight of the intermediate complex and 60 to 90 parts by weight of distilled water;
80 to 95 parts by weight of distilled water, 5 to 10 parts by weight of acrylamide (AAm), 0.1 to 5 parts by weight of heptafluorobutylamide acrylamide (MBAA: N,N-Methylbis Acrylamide), 2-ketoglutaric acid ( Preparing a sixth solution by mixing 0.1 to 1 part by weight of 2-ketoglutaric acid;
Preparing a 7th solution by mixing 30 to 70 parts by weight of the 5th solution and 30 to 70 parts by weight of the 6th solution;
Irradiating light with a wavelength of 300 to 400 nm to the 7th solution in a dark room for 3 to 5 hours and removing unreacted acrylamide to obtain a sound absorption enhancer; A virtual chatbot operating system characterized in that it is manufactured through . According to clause 1,
The system is,
A counting module that generates a counting index by counting at least one of the number of user voices and the number of words expressing user satisfaction,
The character management module is,
A virtual chatbot operating system comprising a character breeding unit that varies at least one of the size and shape of the character according to the counting index. According to clause 2,
The counting module is,
A virtual chatbot operating system that generates a counting index by counting at least one of the number of voices of the user, the number of words expressing the user's satisfaction, and the number of words expressing the user's emotions. According to clause 3,
The chatbot interface is,
Implemented in smartphones including touch screens,
The counting module is,
Characterized in generating a counting index by counting at least one of the number of voices of the user, the number of words expressing the user's satisfaction, the number of expressions of the user's emotions, and the number of touches on the touch screen on which the character is output. , virtual chatbot operating system. According to clause 2,
The character management module is,
A needs request unit including a needs storage part that stores the character's needs, and an alarm part that outputs words related to the needs as a voice to the user at a certain period,
The counting module is,
A needs satisfaction index is generated by counting words related to the needs in the user's voice,
The character breeding department,
A virtual chatbot operating system, characterized in that the facial expression of the character is variably processed according to the needs satisfaction index. delete delete delete