A fully automatic explanation control system, method and application thereof in robots

By collecting and processing exhibition hall data in real time, combining IoT sensors and voice dialogue models, the robot is accurately positioned and natural interaction in the exhibition hall, solving the problem of lack of natural voice interaction in the existing system, and enhancing the sense of participation of tourists and System interactivity.

CN118646774BActive Publication Date: 2024-11-05SHENZHEN KINTO EXHIBITION
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411121735.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-05
Estimated Expiration
2044-08-15

AI Technical Summary

Technical Problem

The existing electronic explanation system lacks the ability to interact and communicate naturally between people, resulting in a weak sense of participation from tourists.

Method used

By collecting exhibition hall explanation data in real time, establishing a voice dialogue model, and installing IoT sensors in the exhibition hall, combining indoor positioning algorithms to realize real-time precise positioning of robots, realizing fully automatic explanation and regulation, and improving interactivity and real-timeness.

Benefits of technology

It enhances the real-time and interactive nature of the robot's fully automatic explanation and regulation, and improves the sense of participation of tourists.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118646774B_ABST
    Figure CN118646774B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of automatic control systems, and discloses a fully automatic explanation control system, method and application thereof in a robot. The method collects explanation data of all collections in each display area of ​​the exhibition hall in real time, saves the collected explanation data to a database and collects voice training data, extracts the collected voice training data, and establishes a connection between voice data and explanation data; at the same time, by installing Internet of Things sensors in each display area of ​​the exhibition hall, and establishing a connection between the installed Internet of Things sensors and the explanation data in the database, and based on the installed Internet of Things sensors, through an indoor positioning algorithm, the real-time and precise positioning of the robot is achieved, and at the same time, based on the real-time and precise positioning of the robot and the connection between the Internet of Things sensors and the explanation data, the fully automatic explanation control is achieved, thereby improving the real-time and interactivity of the fully automatic explanation control of the robot.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of automatic control systems, and in particular to a fully automatic control system and method and application thereof in a robot. Background Art

[0002] With the rapid development of social economy, people have higher and higher requirements for the quality of life, and more and more people are traveling. Many domestic scenic spots have built electronic explanation systems, but these systems are currently broadcast-style explanations activated by infrared sensing, code scanning, and touch screens, lacking the ability to interact with natural voices between people, and tourists do not feel strong in participation.

[0003] The existing patent application document with publication number CN112558567A, the invention receives the user's touch command through the touch panel or collects the user's voice command through the voice collection module, and sends it to the core control mainboard; the core control mainboard converts the touch command or the voice command into a control device command; the core control mainboard obtains the museum guide's instruction requirements according to the control device command, and selects the lights or equipment in the corresponding area according to the museum guide's instruction requirements; the core control mainboard searches for the corresponding equipment or lamps through the WiFi module or Zigbee module, controls the state changes of the relevant equipment or lamps, lacks the ability of natural voice interactive communication between people, and has certain limitations. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a fully automatic explanation control system, method and application in a robot, which has the advantages of real-time and interactivity, and solves the problem of lack of natural voice interactive communication ability between people and weak sense of participation of tourists.

[0006] (II) Technical solution

[0007] In order to solve the above-mentioned technical problems of lack of natural voice interactive communication ability between people and weak sense of participation of tourists, the present invention provides the following technical solutions:

[0008] The present invention discloses a fully automatic explanation control method, which specifically comprises the following steps:

[0009] S1. Collect the explanation data of all collections in each display area of ​​the exhibition hall in real time, and save the collected explanation data to the database;

[0010] S2, collecting voice training data, extracting the collected voice training data, and establishing a voice dialogue model based on the collected explanation data;

[0011] S3, by installing IoT sensors in various display areas of the exhibition hall, and establishing a connection between the installed IoT sensors and the explanation data in the database;

[0012] S4, based on the installed IoT sensors, the robot can be accurately positioned in real time through indoor positioning algorithms;

[0013] S5. Based on the real-time precise positioning of the robot and the connection between the IoT sensor and the explanation data, fully automatic explanation control is achieved.

[0014] The present invention collects explanation data of all collections in each display area of ​​the exhibition hall in real time, saves the collected explanation data in a database and collects voice training data, extracts the collected voice training data, and establishes a connection between the voice data and the explanation data; at the same time, Internet of Things sensors are installed in each display area of ​​the exhibition hall, and a connection between the installed Internet of Things sensors and the explanation data is established in the database, and based on the installed Internet of Things sensors, through an indoor positioning algorithm, real-time and accurate positioning of the robot is achieved, and based on the real-time and accurate positioning of the robot and the connection between the Internet of Things sensors and the explanation data, full-automatic explanation control is achieved, thereby improving the real-time and interactivity of the robot's full-automatic explanation control.

[0015] Preferably, the collecting of speech training data, extracting the collected speech training data, and establishing a speech dialogue model based on the collected explanation data comprises the following steps:

[0016] S21, perform word segmentation and part-of-speech tagging on the collected speech training data, record the word frequency TF of each word, and arrange them in descending order according to the number of recorded word frequencies TF;

[0017] The word frequency TF of each word in the speech training data is recorded as follows:

[0018]

[0019]

[0020] in, is the feature vector , indicating the collected The speech training data of the group, is the feature vector , indicating the collected The speech training data of the group, Indicates the word with the highest frequency. Indicates the word frequency ranking The words, Indicates the word frequency ranking The word, set the TF value to the word frequency TF number, which represents the number of times the word appears, and map the feature vector into a directed graph based on the number of times each word appears in the speech training data;

[0021] S22, select nouns and verbs in the collected speech training data , ... As a vocabulary set, where Indicates the word ranked nth in frequency, take As the initial lexical chain L;

[0022] S23, selecting a word with a TF value greater than 2 in the vocabulary chain set and calculating its semantic similarity with each word in the vocabulary chain L based on the semantic similarity formula;

[0023] S24, set the semantic similarity threshold, if the semantic similarity is greater than the threshold, Inserted into the vocabulary chain L, where Indicates the word ranked jth in terms of frequency;

[0024] S25, if If the semantic similarity with each word in the chain is less than the threshold, a new vocabulary chain will be created. insert;

[0025] S26, repeating steps S23 to S25 until all words are calculated;

[0026] S27. Calculate the semantic similarity between the created vocabulary chain and the collected explanation data based on the semantic similarity formula, match and combine the created vocabulary chain and the explanation data according to the set semantic similarity threshold, and set each created vocabulary chain after the summary matching combination as a group of voice dialogue models.

[0027] Preferably, the semantic similarity formula is as follows:

[0028] The semantic similarity of two words is determined by calculating the maximum cosine similarity of two feature vectors;

[0029]

[0030] in, , Expressing words and Breadth-first traversal information in directed graphs; Represents the nearest common root node of the word mapping in the directed graph; , Expressing words and The depth information in the mapped directed graph, Expressing words and The cosine similarity between .

[0031] The present invention performs word segmentation and part-of-speech tagging on the collected speech training data, records the word frequency of each word, selects words to construct a vocabulary chain, and establishes a speech dialogue model by using a semantic similarity formula to calculate the semantic similarity between words and constructed vocabulary chains, thereby improving the interactivity of the robot's fully automatic explanation control.

[0032] Preferably, the method of installing IoT sensors in various display areas of the exhibition hall and establishing a connection between the installed IoT sensors and the explanation data in the database includes the following steps:

[0033] S31. Based on the installation of IoT sensors, number q IoT sensors q1, q2, ...q q , set each IoT sensor to be associated with a collection;

[0034] S32, taking the IoT sensor q1 as the coordinate origin, establishing a three-dimensional coordinate system according to the coordinates of q1, and determining the coordinates of other IoT sensors based on the IoT sensor q1;

[0035] S33, saving the coordinates of each IoT sensor in a database of the system;

[0036] S34. Match the explanation data stored in the database with the collections associated with the IoT sensor, and merge the matched explanation data with the coordinates of each IoT sensor, represented by G(V,E), where V represents the explanation data, E represents the coordinates of the IoT sensor, and G represents the combined data.

[0037] The present invention installs Internet of Things sensors in various display areas of the exhibition hall and numbers them, and determines the coordinates of each Internet of Things sensor by establishing a three-dimensional coordinate system; at the same time, by matching the explanation data stored in the database with the collections associated with the Internet of Things sensors, the accuracy of the robot's fully automatic explanation control is improved.

[0038] Preferably, the method of realizing real-time and accurate positioning of the robot based on the installed IoT sensor through an indoor positioning algorithm comprises the following steps:

[0039] S41, performing preliminary positioning of the robot through offline acquisition;

[0040] S42. Accurately position the robot through online positioning.

[0041] Preferably, the preliminary positioning of the robot by offline acquisition includes the following steps:

[0042] S411, dividing numerous sampling points to fully cover the area to be detected or positioned, and collecting signal strength data of the IoT sensor at one sampling point multiple times;

[0043] S412, performing preliminary positioning of the robot based on the collected signal strength data of the IoT sensor;

[0044] It is assumed that the signal strength of IoT sensors will decay evenly as the distance increases in space;

[0045] Signal strength in space propagation loss function The formula is as follows:

[0046] ;

[0047] in, Represents the signal strength propagation loss function in space, Indicates the signal output frequency of the IoT sensor, Indicates the distance between the robot and the IoT sensor;

[0048] The output signal of the IoT sensor is set to be constant, and the distance between the sampling point and each IoT sensor is determined based on the signal strength of each IoT sensor collected at each sampling point;

[0049] S413, integrating the signal strength data of the IoT sensors collected at all sampling points and the distances between the IoT sensors to construct a location fingerprint and save it to a database;

[0050] After receiving the IoT sensor signal, the sampling point stores a set of triplet data including the location coordinates of the collection point, the IoT sensor signal strength and the IoT sensor number in the database;

[0051] A set of triplet data is set as a location fingerprint.

[0052] Preferably, the precise positioning of the robot by online positioning comprises the following steps:

[0053] In the online positioning stage, when the robot receives the IoT sensor signal, it compares the received IoT sensor signal strength and the IoT sensor number with the location fingerprint in the database to find the same or similar location fingerprints, and uses the location coordinates of the collection points in the same or similar location fingerprints as the robot's location coordinates.

[0054] The present invention sets numerous sampling points through offline collection, and collects signal strength data of the Internet of Things sensor multiple times at one sampling point, thereby preliminarily determining the positioning of the robot and constructing a position fingerprint. At the same time, the real-time and precise positioning of the robot is determined online through the constructed position fingerprint, thereby improving the accuracy of the robot's fully automatic explanation and control.

[0055] Preferably, the automatic explanation control based on the real-time accurate positioning of the robot and the connection between the IoT sensor and the explanation data comprises the following steps:

[0056] S51, voice control;

[0057] Collect user voice data in real time, and perform word segmentation and part-of-speech tagging on the collected user voice data to obtain user voice words collected in real time;

[0058] The similarity between the real-time collected user voice words and the established voice dialogue model is calculated through the semantic similarity formula;

[0059] Select a group of voice dialogue models with the highest similarity, and play the explanation data corresponding to the selected voice dialogue models;

[0060] S52, automatic explanation and control;

[0061] The robot's position coordinates are determined in real time, and a distance threshold between the robot and the IoT sensor is set. When the distance between the robot's position and the IoT sensor is less than the set distance threshold, the explanation data associated with the IoT sensor is played.

[0062] The present invention divides the modes of voice control and automatic explanation control, and respectively uses a voice recognition control mode and a sensor distance threshold setting mode to realize full-automatic explanation control of the robot, thereby improving the interactivity and real-time performance of the robot's full-automatic explanation control.

[0063] The present invention discloses a fully automatic explanation control system, comprising: a database, an Internet of Things sensor, a data acquisition module, a robot and a positioning module;

[0064] The data collection module is used to collect the explanation data of all collections in real time;

[0065] The database is used to store the collected data in real time;

[0066] The Internet of Things sensor includes a signal output module;

[0067] The signal output module is used to constantly output wireless signals;

[0068] The robot includes a signal receiving module, a voice receiving module and a voice control module;

[0069] The signal receiving module is used to receive the wireless signal output by the signal output module in real time;

[0070] The voice receiving module is used to receive voice data for input in real time;

[0071] The voice control module is used to identify, extract and control the received voice data;

[0072] The positioning module is used to perform real-time positioning of the robot according to the collected signal strength.

[0073] The invention also discloses an application in a robot, wherein the application uses the above-mentioned fully automatic explanation and control method.

[0074] (III) Beneficial effects

[0075] Compared with the prior art, the present invention has the following beneficial effects:

[0076] 1. The present invention collects the explanation data of all collections in each display area of ​​the exhibition hall in real time, saves the collected explanation data to a database and collects voice training data, extracts the collected voice training data, and establishes a connection between the voice data and the explanation data; at the same time, by installing Internet of Things sensors in each display area of ​​the exhibition hall, and establishing a connection between the installed Internet of Things sensors and the explanation data in the database, and based on the installed Internet of Things sensors, through an indoor positioning algorithm, the real-time and accurate positioning of the robot is achieved, and at the same time, based on the real-time and accurate positioning of the robot and the connection between the Internet of Things sensors and the explanation data, fully automatic explanation control is achieved, thereby improving the real-time and interactivity of the robot's fully automatic explanation control.

[0077] 2. The present invention performs word segmentation and part-of-speech tagging on the collected speech training data, records the word frequency of each word, selects words to construct vocabulary chains, and establishes a speech dialogue model by using a semantic similarity formula to calculate the semantic similarity between words and constructed vocabulary chains, thereby improving the interactivity of the robot's fully automatic explanation control.

[0078] 3. The present invention installs IoT sensors in various display areas of the exhibition hall and numbers them, and determines the coordinates of each IoT sensor by establishing a three-dimensional coordinate system; at the same time, by matching the explanation data stored in the database with the collections associated with the IoT sensors, the accuracy of the robot's fully automatic explanation control is improved.

[0079] 4. The present invention sets numerous sampling points through offline collection, and collects signal strength data of the Internet of Things sensor multiple times at one sampling point to preliminarily determine the positioning of the robot and construct a position fingerprint. At the same time, the real-time and precise positioning of the robot is determined online through the constructed position fingerprint, thereby improving the accuracy of the robot's fully automatic explanation and control.

[0080] 5. The present invention divides the modes of voice control and automatic explanation control, and respectively uses a voice recognition control mode and a sensor distance threshold setting mode to realize fully automatic explanation control of the robot, thereby improving the interactivity and real-time performance of the robot's fully automatic explanation control. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a schematic diagram of the flow structure of the fully automatic explanation control method of the present invention. DETAILED DESCRIPTION

[0082] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0083] Example 1

[0084] This embodiment discloses a fully automatic explanation control method, which specifically includes the following steps:

[0085] S1. Collect the explanation data of all collections in each display area of ​​the exhibition hall in real time, and save the collected explanation data to the database;

[0086] S2, collecting voice training data, extracting the collected voice training data, and establishing a voice dialogue model based on the collected explanation data;

[0087] Collecting voice training data, extracting the collected voice training data, and establishing a voice dialogue model based on the collected explanation data includes the following steps:

[0088] S21, perform word segmentation and part-of-speech tagging on the collected speech training data, record the word frequency TF of each word, and arrange them in descending order according to the number of recorded word frequencies TF;

[0089] The word frequency TF of each word in the speech training data is recorded as follows:

[0090]

[0091]

[0092] in, is the feature vector , indicating the collected The speech training data of the group, is the feature vector , indicating the collected The speech training data of the group, Indicates the word with the highest frequency. Indicates the word frequency ranking The words, Indicates the word frequency ranking The word, set the TF value to the word frequency TF number, which represents the number of times the word appears, and map the feature vector into a directed graph based on the number of times each word appears in the speech training data;

[0093] S22, select nouns and verbs in the collected speech training data , ... As a vocabulary set, where Indicates the word ranked nth in frequency, take As the initial lexical chain L;

[0094] S23, selecting a word with a TF value greater than 2 in the vocabulary chain set and calculating its semantic similarity with each word in the vocabulary chain L based on the semantic similarity formula;

[0095] The semantic similarity formula is as follows:

[0096] The semantic similarity of two words is determined by calculating the maximum cosine similarity of two feature vectors;

[0097]

[0098] in, , Expressing words and Breadth-first traversal information in directed graphs; Represents the nearest common root node of the word mapping in the directed graph; , Expressing words and The depth information in the mapped directed graph, Expressing words and The cosine similarity between ;

[0099] S24, set the semantic similarity threshold, if the semantic similarity is greater than the threshold, Inserted into the vocabulary chain L, where Indicates the word ranked jth in terms of frequency;

[0100] S25, if If the semantic similarity with each word in the chain is less than the threshold, a new vocabulary chain will be created. insert;

[0101] S26, repeating steps S23 to S25 until all words are calculated;

[0102] S27, calculating the semantic similarity between the created vocabulary chain and the collected explanation data based on the semantic similarity formula, and matching and combining the created vocabulary chain and the explanation data according to the set semantic similarity threshold, and setting each created vocabulary chain after the summary matching and combination as a group of voice dialogue models;

[0103] S3, by installing IoT sensors in various display areas of the exhibition hall, and establishing a connection between the installed IoT sensors and the explanation data in the database;

[0104] The following steps are involved in installing IoT sensors in various display areas of the exhibition hall and establishing a connection between the installed IoT sensors and the explanation data in the database:

[0105] S31. Based on the installation of IoT sensors, number q IoT sensors q1, q2, ...q q , set each IoT sensor to be associated with a collection;

[0106] S32, taking the IoT sensor q1 as the coordinate origin, establishing a three-dimensional coordinate system according to the coordinates of q1, and determining the coordinates of other IoT sensors based on the IoT sensor q1;

[0107] S33, saving the coordinates of each IoT sensor in a database of the system;

[0108] S34, matching the explanation data stored in the database with the collections associated with the IoT sensor, and merging the matched explanation data with the coordinates of each IoT sensor, represented by G(V,E), where V represents the explanation data, E represents the coordinates of the IoT sensor, and G represents the combined data;

[0109] S4, based on the installed IoT sensors, the robot can be accurately positioned in real time through indoor positioning algorithms;

[0110] Based on the installed IoT sensors, the real-time and accurate positioning of the robot is achieved through the indoor positioning algorithm, which includes the following steps:

[0111] S41, performing preliminary positioning of the robot through offline acquisition;

[0112] S411, dividing numerous sampling points to fully cover the area to be detected or positioned, and collecting signal strength data of the IoT sensor at one sampling point multiple times;

[0113] S412, performing preliminary positioning of the robot based on the collected signal strength data of the IoT sensor;

[0114] It is assumed that the signal strength of IoT sensors will decay evenly as the distance increases in space;

[0115] Signal strength in space propagation loss function The formula is as follows:

[0116] ;

[0117] in, Represents the signal strength propagation loss function in space, Indicates the signal output frequency of the IoT sensor, Indicates the distance between the robot and the IoT sensor;

[0118] The output signal of the IoT sensor is set to be constant, and the distance between the sampling point and each IoT sensor is determined based on the signal strength of each IoT sensor collected at each sampling point;

[0119] S413, integrating the signal strength data of the IoT sensors collected at all sampling points and the distances between the IoT sensors to construct a location fingerprint and save it to a database;

[0120] After receiving the IoT sensor signal, the sampling point stores a set of triplet data including the location coordinates of the collection point, the IoT sensor signal strength and the IoT sensor number in the database;

[0121] A set of triplet data is set as a location fingerprint;

[0122] S42, accurately positioning the robot through online positioning;

[0123] In the online positioning stage, when the robot receives the IoT sensor signal, it compares the received IoT sensor signal strength and the IoT sensor number with the location fingerprint in the database to find the same or similar location fingerprints, and uses the location coordinates of the collection points in the same or similar location fingerprints as the robot's location coordinates;

[0124] S5. Based on the real-time precise positioning of the robot and the connection between the IoT sensor and the explanation data, fully automatic explanation control is realized;

[0125] Based on the real-time precise positioning of the robot and the connection between the IoT sensor and the explanation data, the following steps are involved to achieve fully automatic explanation control:

[0126] S51, voice control;

[0127] Collect user voice data in real time, and perform word segmentation and part-of-speech tagging on the collected user voice data to obtain user voice words collected in real time;

[0128] Furthermore, the similarity between the user's voice words collected in real time and the established voice dialogue model is calculated through a semantic similarity formula;

[0129] Further, a group of voice dialogue models with the highest similarity is selected, and explanation data corresponding to the selected voice dialogue model is played;

[0130] S52, automatic explanation and control;

[0131] Determine the robot's position coordinates in real time and set the distance threshold between the robot and the IoT sensor. When the distance between the robot's position and the IoT sensor is less than the set distance threshold, play the explanation data associated with the IoT sensor.

[0132] Example 2

[0133] This embodiment discloses a fully automatic explanation control system, including: a database, an Internet of Things sensor, a data acquisition module, a robot, and a positioning module;

[0134] The data collection module is used to collect the explanation data of all collections in real time;

[0135] The database is used to store the collected data in real time;

[0136] The Internet of Things sensor includes a signal output module;

[0137] The signal output module is used to constantly output wireless signals;

[0138] The robot includes a signal receiving module, a voice receiving module and a voice control module;

[0139] The signal receiving module is used to receive the wireless signal output by the signal output module in real time;

[0140] The voice receiving module is used to receive voice data for input in real time;

[0141] The voice control module is used to identify, extract and control the received voice data;

[0142] The positioning module is used to perform real-time positioning of the robot according to the collected signal strength.

[0143] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A fully automatic explanation control method, It is characterized in that The following steps are involved: S1. Collect the explanation data of all collections in each display area of ​​the exhibition hall in real time, and save the collected explanation data to the database; S2, collecting voice training data, extracting the collected voice training data, and establishing a voice dialogue model based on the collected explanation data; S3, by installing IoT sensors in various display areas of the exhibition hall, and establishing a connection between the installed IoT sensors and the explanation data in the database; S4, based on the installed IoT sensors, the robot can be accurately positioned in real time through indoor positioning algorithms; S5. Based on the real-time precise positioning of the robot and the connection between the IoT sensor and the explanation data, fully automatic explanation control is realized; The collecting of speech training data, extracting the collected speech training data, and establishing a speech dialogue model based on the collected explanation data comprises the following steps: S21, perform word segmentation and part-of-speech tagging on the collected speech training data, record the word frequency TF of each word, and arrange them in descending order according to the number of recorded word frequencies TF; The word frequency TF of each word in the speech training data is recorded as follows: in, is the feature vector , indicating the collected The speech training data of the group, is the feature vector , indicating the collected The speech training data of the group, Indicates the word with the highest frequency. Indicates the word frequency ranking The words, Indicates the word frequency ranking The word, set the TF value to the word frequency TF number, which represents the number of times the word appears, and map the feature vector into a directed graph based on the number of times each word appears in the speech training data; S22, select nouns and verbs in the collected speech training data , ... As a vocabulary set, where Indicates the word ranked nth in frequency, take As the initial lexical chain L; S23, selecting a word with a TF value greater than 2 in the vocabulary chain set and calculating its semantic similarity with each word in the vocabulary chain L based on the semantic similarity formula; S24, set the semantic similarity threshold, if the semantic similarity is greater than the threshold, Inserted into the vocabulary chain L, where Indicates the word ranked jth in terms of frequency; S25, if If the semantic similarity with each word in the chain is less than the threshold, a new vocabulary chain will be created. insert; S26, repeating steps S23 to S25 until all words are calculated; S27, calculating the semantic similarity between the created vocabulary chain and the collected explanation data based on the semantic similarity formula, and matching and combining the created vocabulary chain and the explanation data according to the set semantic similarity threshold, and setting each created vocabulary chain after the summary matching and combination as a group of voice dialogue models; The steps of installing IoT sensors in various display areas of the exhibition hall and establishing a connection between the installed IoT sensors and the explanation data in the database include: S31. Based on the installation of IoT sensors, number q IoT sensors 1 ,q 2 ,...q q , set each IoT sensor to be associated with a collection; S32, IoT Sensors 1 is the coordinate origin, according to q 1 Coordinates establish a three-dimensional coordinate system and based on the IoT sensor q 1 Determine the coordinates of various other IoT sensors; S33, saving the coordinates of each IoT sensor in a database of the system; S34, matching the explanation data stored in the database with the collections associated with the IoT sensor, and merging the matched explanation data with the coordinates of each IoT sensor, represented by G(V,E), where V represents the explanation data, E represents the coordinates of the IoT sensor, and G represents the combined data; The real-time and accurate positioning of the robot based on the installed IoT sensor through the indoor positioning algorithm includes the following steps: S41, performing preliminary positioning of the robot through offline acquisition; S42, accurately positioning the robot through online positioning; The preliminary positioning of the robot by offline acquisition includes the following steps: S411, dividing numerous sampling points to fully cover the area to be detected or positioned, and collecting signal strength data of the IoT sensor at one sampling point multiple times; S412, performing preliminary positioning of the robot based on the collected signal strength data of the IoT sensor; It is assumed that the signal strength of IoT sensors will decay evenly as the distance increases in space; Signal strength in space propagation loss function The formula is as follows: ; in, Represents the signal strength propagation loss function in space, Indicates the signal output frequency of the IoT sensor, Indicates the distance between the robot and the IoT sensor; The output signal of the IoT sensor is set to be constant, and the distance between the sampling point and each IoT sensor is determined based on the signal strength of each IoT sensor collected at each sampling point; S413, integrating the signal strength data of the IoT sensors collected at all sampling points and the distances between the IoT sensors to construct a location fingerprint and save it to a database; After receiving the IoT sensor signal, the sampling point stores a set of triplet data including the location coordinates of the collection point, the IoT sensor signal strength and the IoT sensor number in the database; A set of triplet data is set as a location fingerprint; The precise positioning of the robot by online positioning includes the following steps: In the online positioning stage, when the robot receives the IoT sensor signal, it compares the received IoT sensor signal strength and the IoT sensor number with the location fingerprint in the database to find the same or similar location fingerprints, and uses the location coordinates of the collection points in the same or similar location fingerprints as the robot's location coordinates; The method of realizing fully automatic explanation control based on the real-time accurate positioning of the robot and the connection between the IoT sensor and the explanation data includes the following steps: S51, voice control; Collect user voice data in real time, and perform word segmentation and part-of-speech tagging on the collected user voice data to obtain user voice words collected in real time; The similarity between the real-time collected user voice words and the established voice dialogue model is calculated through the semantic similarity formula; Select a group of voice dialogue models with the highest similarity, and play the explanation data corresponding to the selected voice dialogue model; S52, automatic explanation and control; The robot's position coordinates are determined in real time, and a distance threshold between the robot and the IoT sensor is set. When the distance between the robot's position and the IoT sensor is less than the set distance threshold, the explanation data associated with the IoT sensor is played.

2. A fully automatic explanation control method according to claim 1, It is characterized in that The semantic similarity formula is as follows: The semantic similarity of two words is determined by calculating the maximum cosine similarity of two feature vectors; in, , Expressing words and Breadth-first traversal information in directed graphs; Represents the nearest common root node of the word mapping in the directed graph; , Expressing words and The depth information in the mapped directed graph, Expressing words and The cosine similarity between .

3. A system for realizing the fully automatic explanation and control method according to claim 1, It is characterized in that include: Database, IoT sensors, data acquisition modules, robots, and positioning modules; The data collection module is used to collect the explanation data of all collections in real time; The database is used to store the collected data in real time; The Internet of Things sensor includes a signal output module; The signal output module is used to constantly output wireless signals; The robot includes a signal receiving module, a voice receiving module and a voice control module; The signal receiving module is used to receive the wireless signal output by the signal output module in real time; The voice receiving module is used to receive voice data for input in real time; The voice control module is used to identify, extract and control the received voice data; The positioning module is used to perform real-time positioning of the robot according to the signal strength collected from the IoT sensor.