CN113223526A - Voice control method and system for water dispenser - Google Patents

Voice control method and system for water dispenser Download PDF

Info

Publication number
CN113223526A
CN113223526A CN202110491761.1A CN202110491761A CN113223526A CN 113223526 A CN113223526 A CN 113223526A CN 202110491761 A CN202110491761 A CN 202110491761A CN 113223526 A CN113223526 A CN 113223526A
Authority
CN
China
Prior art keywords
evaluation
preset
perception
complexity
voice control
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
CN202110491761.1A
Other languages
Chinese (zh)
Inventor
陈芒
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qlife Tech Co ltd
Original Assignee
Qlife Tech Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Qlife Tech Co ltd filed Critical Qlife Tech Co ltd
Priority to CN202110491761.1A priority Critical patent/CN113223526A/en
Publication of CN113223526A publication Critical patent/CN113223526A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
    • A—HUMAN NECESSITIES
    • A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
    • A47J—KITCHEN EQUIPMENT; COFFEE MILLS; SPICE MILLS; APPARATUS FOR MAKING BEVERAGES
    • A47J31/00—Apparatus for making beverages
    • A47J31/44—Parts or details or accessories of beverage-making apparatus
    • A47J31/52—Alarm-clock-controlled mechanisms for coffee- or tea-making apparatus ; Timers for coffee- or tea-making apparatus; Electronic control devices for coffee- or tea-making apparatus
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
    • G10L2015/223—Execution procedure of a spoken command

Landscapes

  • Engineering & Computer Science (AREA)
  • Food Science & Technology (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Devices For Dispensing Beverages (AREA)

Abstract

The invention provides a voice control method and a voice control system for a water dispenser, wherein the method comprises the following steps: acquiring a first voice control instruction input by a user; analyzing the first voice control instruction, and if the first voice control instruction is a wake-up instruction, acquiring a second voice control instruction input by a user; analyzing the second voice control instruction, and determining the type of the beverage which the user wants to brew; determining a reply voice and a target water temperature corresponding to the type of the beverage; and playing the reply voice, and simultaneously controlling the water dispenser to adjust the water temperature to the target water temperature. According to the voice control method and the voice control system for the water dispenser, when a user uses the water dispenser to brew different types of drinks, only the awakening instruction and the type of the drink to be brewed need to be spoken, and the water dispenser automatically prepares water with corresponding appropriate temperature for the user to use, so that the voice control method and the voice control system for the water dispenser are very convenient and fast.

Description

Voice control method and system for water dispenser
Technical Field
The invention relates to the technical field of voice control, in particular to a voice control method and system for a water dispenser.
Background
At present, when a user uses a water dispenser to brew different types of drinks (coffee, tea, milk powder and the like), water with corresponding appropriate temperature is needed to brew, and the user needs to set the water temperature by operating a physical key on the water dispenser, which is complicated.
Disclosure of Invention
One of the purposes of the invention is to provide a voice control method and system for a water dispenser, when a user uses the water dispenser to brew different types of drinks, only a wake-up instruction and the type of the drink to be brewed need to be spoken, and the water dispenser automatically prepares water with corresponding appropriate temperature for the user to use, so that the voice control method and system are very convenient.
The embodiment of the invention provides a voice control method of a water dispenser, which comprises the following steps:
acquiring a first voice control instruction input by a user;
analyzing the first voice control instruction, and if the first voice control instruction is a wake-up instruction, acquiring a second voice control instruction input by a user;
analyzing the second voice control instruction, and determining the type of the beverage which the user wants to brew;
determining a reply voice and a target water temperature corresponding to the type of the beverage;
and playing the reply voice, and simultaneously controlling the water dispenser to adjust the water temperature to the target water temperature.
Preferably, the acquiring the first voice control instruction input by the user includes:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000021
Figure BDA0003052606450000022
wherein, theta1As a first evaluation index, mu1And mu2Is a preset weight value, alpha1,iRecording the ith first evaluation value n output after evaluation by the first evaluation model for voice control1For recording the total number of first evaluation values, alpha, output after evaluation by the first evaluation model for speech control1,0Is a preset first evaluation value threshold value, and e is a natural constant；
Acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000023
Figure BDA0003052606450000024
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for voice control2For recording the total number of second evaluation values, alpha, output after evaluation by the second evaluation model for speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first judgment index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure BDA0003052606450000025
wherein, γ1Is a first decision index, σ1And σ2Is a preset weight value, theta1Is a first evaluation index, θ2Is a second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring a first voice control instruction.
Preferably, the determining of the response voice corresponding to the beverage type and the target water temperature includes:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the type of the drink and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second partition.
Preferably, the first partition and the second partition are filled before the answer voice corresponding to the beverage type and the target water temperature are determined;
the filling steps of the first partition and the second partition are as follows:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first associated data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after a first perception model perceives first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure BDA0003052606450000031
Figure BDA0003052606450000032
wherein, P1Is a first perception index, J1And J2Is a preset weight value, q1,tA t-th first perception value, z, output after the first associated data are perceived by the first perception model1A total number q of first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second perception model perceives the second associated data;
calculating a second perception index based on the second perception value, the calculation formula being as follows:
Figure BDA0003052606450000033
Figure BDA0003052606450000034
wherein, P2Is a second perception index, J1And J2Is a preset weight value, q2,tA t second perception value, z, output after the second associated data is perceived by the second perception model2The total number of second perception values q output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure BDA0003052606450000035
wherein, γ2Is a second determination index, ω1And ω2Is a preset weight value, P1Is a first perceptual index, P2Is a second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
Preferably, the voice control method of the water dispenser further comprises the following steps:
acquiring first time for analyzing a first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
inquiring a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on first complexity according to the interval formulation rule, and formulating a second complexity interval based on second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity falling within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein the calculation formula is as follows:
Figure BDA0003052606450000041
wherein ε is a demand index, ArIs the fourth complexity, X1Is the total number of the fourth complexity, BrIs the fifth complexity, X2Is the total number of the fifth complexity, O1To a first complexity, O2A second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
The embodiment of the invention provides a voice control system of a water dispenser, which comprises:
the acquisition module is used for acquiring a first voice control instruction input by a user;
the first analysis module is used for analyzing the first voice control instruction, and acquiring a second voice control instruction input by a user if the first voice control instruction is a wake-up instruction;
the second analysis module is used for analyzing the second voice control instruction and determining the type of the beverage which the user wants to brew;
the determining module is used for determining the answer voice and the target water temperature corresponding to the beverage type;
and the control module is used for playing the reply voice and controlling the water dispenser to adjust the water temperature to the target water temperature.
Preferably, the obtaining module performs operations including:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000051
Figure BDA0003052606450000052
wherein, theta1As a first evaluation index, mu1And mu2Is a preset weight value, alpha1,iRecording the ith first evaluation value n output after evaluation by the first evaluation model for voice control1For recording the total number of first evaluation values, alpha, output after evaluation by the first evaluation model for speech control1,0A preset first evaluation value threshold value is set, and e is a natural constant;
acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000053
Figure BDA0003052606450000054
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for voice control2For recording the total number of second evaluation values, alpha, output after evaluation by the second evaluation model for speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first judgment index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure BDA0003052606450000055
wherein, γ1Is a first decision index, σ1And σ2Is a preset weight value, theta1Is a first evaluation index, θ2Is a second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring a first voice control instruction.
Preferably, the determining module performs operations including:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the type of the drink and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second partition.
Preferably, the voice control system of the water dispenser further comprises:
the filling module is used for filling the first partition and the second partition before the determining module determines the answer voice and the target water temperature corresponding to the beverage type;
the filling module executes the following operations:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first associated data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after a first perception model perceives first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure BDA0003052606450000061
Figure BDA0003052606450000062
wherein, P1Is a first perception index, J1And J2Is a preset weight value, q1,tA t-th first perception value, z, output after the first associated data are perceived by the first perception model1A total number q of first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second perception model perceives the second associated data;
calculating a second perception index based on the second perception value, the calculation formula being as follows:
Figure BDA0003052606450000063
Figure BDA0003052606450000071
wherein, P2Is a second perception index, J1And J2Is a preset weight value, q2,tA t second perception value, z, output after the second associated data is perceived by the second perception model2The total number of second perception values q output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure BDA0003052606450000072
wherein, γ2Is a second determination index, ω1And ω2Is a preset weight value, P1Is a first perceptual index, P2Is a second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
Preferably, the voice control system of the water dispenser further comprises:
a demand determination module;
the requirement determining module executes the following operations:
acquiring first time for analyzing a first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
inquiring a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on first complexity according to the interval formulation rule, and formulating a second complexity interval based on second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity falling within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein the calculation formula is as follows:
Figure BDA0003052606450000073
wherein ε is a demand index, ArIs the fourth complexity, X1Is the total number of the fourth complexity, BrIs the fifth complexity, X2Is the total number of the fifth complexity, O1To a first complexity, O2A second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of a voice control method of a water dispenser in an embodiment of the invention;
fig. 2 is a schematic diagram of a voice control system of a water dispenser in an embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
The embodiment of the invention provides a voice control method of a water dispenser, which comprises the following steps as shown in figure 1:
s1, acquiring a first voice control instruction input by a user;
s2, analyzing the first voice control instruction, and if the first voice control instruction is a wake-up instruction, acquiring a second voice control instruction input by a user;
s3, analyzing the second voice control instruction, and determining the type of the beverage which the user wants to brew;
s4, determining the answer voice and the target water temperature corresponding to the type of the beverage;
and S5, playing the reply voice and controlling the water dispenser to adjust the water temperature to the target water temperature.
The working principle and the beneficial effects of the technical scheme are as follows:
the first voice control command input by the user for the first time is analyzed, when the command is a wake-up command (for example: 'hello | drinking fountain'), a second voice control command input by the user again is obtained, analyzing the instruction, determining the type of the beverage which the user wants to brew, determining the response voice and the target water temperature corresponding to the type of the beverage (for example, if the second voice control instruction is ' I want to brew a cup of green tea ', the type of the beverage is green tea, the response voice corresponding to the green tea is ' good ', the proper temperature for brewing the green tea is 85 ℃, heating is started ', and the target water temperature corresponding to the green tea is 85 ℃), playing the response voice to reply the user, meanwhile, the water dispenser is controlled to adjust the water temperature to the target water temperature, and after the water temperature is adjusted, a notification voice of successful adjustment can be played (for example, the green tea brewing water is ready for use).
According to the embodiment of the invention, when a user uses the water dispenser to brew different types of beverages, only the awakening instruction and the type of the beverage to be brewed need to be spoken, and the water dispenser automatically prepares water with a corresponding appropriate temperature for the user to use, so that the water dispenser is very convenient and fast.
The embodiment of the invention provides a voice control method for a water dispenser, which is used for acquiring a first voice control instruction input by a user and comprises the following steps:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000091
Figure BDA0003052606450000092
wherein, theta1As a first evaluation index, mu1And mu2Is a preset weight value, alpha1,iRecording the ith first evaluation value n output after evaluation by the first evaluation model for voice control1For recording voice controlEstimating the total number of the first estimation values, alpha, output after the model estimation1,0A preset first evaluation value threshold value is set, and e is a natural constant;
acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000093
Figure BDA0003052606450000094
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for voice control2For recording the total number of second evaluation values, alpha, output after evaluation by the second evaluation model for speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first judgment index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure BDA0003052606450000101
wherein, γ1Is a first decision index, σ1And σ2Is a preset weight value, theta1Is a first evaluation index, θ2Is a second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring a first voice control instruction.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset voice control record specifically comprises the following steps: after the water dispenser is put into use, the received voice control instruction and the record of executing the corresponding operation are recorded; the preset first evaluation model specifically includes: the evaluation model can evaluate whether the phenomenon of child disorganization (such as imitating that an adult frequently speaks a wake-up instruction and/or a brewing beverage type instruction and the like) occurs in the last short time (such as 5 minutes) based on the voice control record, and automatically output a first evaluation value after evaluation, wherein the smaller the first evaluation value is, the greater the possibility of child disorganization is; the preset second evaluation model specifically comprises: a model generated after learning a large amount of records of the adult voice control disorderly operation by using a machine learning algorithm, wherein the evaluation model can evaluate whether the phenomenon of the adult disorderly operation (such as repeatedly speaking the same brewing beverage type instruction and the like) occurs in the last short time (such as 5 minutes) based on the voice control records, and after evaluation, a second evaluation value is automatically output, and the smaller the second evaluation value is, the more possibility of the adult disorderly operation is shown to occur; the purpose of evaluating the entire speech control recording is to evaluate whether the features (e.g., voiceprints, etc.) characterized by the models in conjunction with the most recently occurring instructions have previously appeared in combination, for example: if the operation happens, the possibility of operation disorder is high; in order to avoid the problem of accidental errors, the first evaluation model and the second evaluation model carry out multiple evaluations and output multiple evaluation values; calculating a first evaluation index based on each first evaluation value, wherein the smaller the first evaluation index is, the greater the overall possibility of occurrence of child disorientation is; calculating a second evaluation index based on each second evaluation value, the smaller the second evaluation index is, the greater the overall possibility that the adult dishonest operation occurs is indicated to be; calculating a first judgment index based on the first evaluation index and the second evaluation index, and acquiring a first voice control instruction when the first judgment index is greater than or equal to a preset first judgment index threshold (for example: 85); the preset first evaluation value threshold is specifically: for example, 85; the preset second evaluation value threshold is specifically: such as 86.
According to the embodiment of the invention, the first evaluation model and the second evaluation model are used for evaluating the voice control record, so that child random operation and/or adult random operation are intelligently ignored, the use power consumption is reduced, and the user experience is improved.
The embodiment of the invention provides a voice control method of a water dispenser, which is used for determining a reply voice and a target water temperature corresponding to a beverage type and comprises the following steps:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the type of the drink and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second partition.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset pairing database specifically comprises: a plurality of pairing items, each pairing item comprising: beverage type, answer voice and target water temperature; the pairing database is divided into two partitions, namely a first partition and a second partition; the matching database contains suitable water temperatures corresponding to various beverage types (such as green tea, jasmine tea, white tea, oolong tea, black tea, Pu' er tea, yellow tea, oolong tea, milk powder, coffee, milk tea, cocoa powder, fruit juice, coconut powder, matcha powder, ginger tea, etc.); when the answer voice and the target water temperature corresponding to the beverage type are determined, the answer voice and the target water temperature corresponding to the beverage type are preferentially determined from the first partition, and when the determination fails, the answer voice and the target water temperature corresponding to the beverage type are determined from the second partition.
The pairing database of the embodiment of the invention contains the appropriate water temperatures corresponding to various beverage types, so that the requirements of users can be met to the greatest extent, and the user experience is improved.
The embodiment of the invention provides a voice control method of a water dispenser, which is characterized in that a first partition and a second partition are filled before a reply voice corresponding to a beverage type and a target water temperature are determined;
the filling steps of the first partition and the second partition are as follows:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first associated data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after a first perception model perceives first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure BDA0003052606450000111
Figure BDA0003052606450000112
wherein, P1Is a first perception index, J1And J2Is a preset weight value, q1,tA t-th first perception value, z, output after the first associated data are perceived by the first perception model1A total number q of first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second perception model perceives the second associated data;
calculating a second perception index based on the second perception value, the calculation formula being as follows:
Figure BDA0003052606450000121
Figure BDA0003052606450000122
wherein, P2Is a second perception index, J1And J2Is a preset weight value, q2,tA t second perception value, z, output after the second associated data is perceived by the second perception model2The total number of second perception values q output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure BDA0003052606450000123
wherein, γ2Is a second determination index, ω1And ω2Is a preset weight value, P1Is a first perceptual index, P2Is a second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset pairing item list specifically includes: a plurality of pairing items, each pairing item comprising: beverage type, answer voice and target water temperature; the preset first obtaining path specifically includes: the system is connected with a local beverage type selection record database (a database for recording the beverage type record selected by the user) and used for acquiring local beverage type selection record data; the first associated data is data associated with the selected pairing item in the local beverage type selection record data; the preset first perception model specifically includes: a model generated by learning a large amount of local beverage type selection record data and manual judgment preference records by using a machine learning algorithm, wherein the model can beSensing the first associated data, sensing the preference of the user for selecting the pairing item, and outputting a first sensing value, wherein the greater the first sensing value is, the more the user prefers the pairing item; the preset second obtaining path specifically includes: the beverage type selection recording data of each user can be obtained through the path, namely big data; the second associated data is data associated with the selected pairing item in the big data; the preset second perception model specifically includes: the model is generated after a large amount of beverage type selection record data and manual evaluation preference records are learned by utilizing a machine learning algorithm, the model can perform data elimination and other operations based on the credibility of a certain data source in big data, second associated data can be sensed, the preference degree of different users for selecting the pairing item is sensed, a second sensing value is output, and the greater the second sensing value is, the more most users prefer the pairing item; calculating a first perception index based on each first perception value, wherein the larger the first perception index is, the higher the overall degree of preference of the user of the water dispenser for the pairing item is; calculating a second perception index based on each second perception value, wherein the larger the second perception index is, the higher the overall degree of preference of most users for the paired items is; calculating a second judgment index based on the first perception index and the second perception index, wherein when the second judgment index is greater than or equal to a preset second judgment index threshold (for example: 90), the pair is high in possibility of being selected and should be listed in the first partition for preferential determination; in calculating the second decision index threshold, ω is the value of ω because of the emphasis on recording data in conjunction with the user's local selection1Should be greater than omega2(ii) a The preset first perception value threshold specifically includes: for example, 88; the preset second perception value threshold specifically includes: such as 89.
The embodiment of the invention can intelligently decide whether the selected matching item is filled into the first partition or the second partition based on the first correlation data and the second correlation data, can quickly list the selected matching item with high possibility into the first partition for priority determination, greatly reduces the determination time when determining the response voice and the target water temperature corresponding to the beverage type, improves the working efficiency of the system, reduces the delay of the system for replying the user and adopting temperature regulation control, and improves the user experience.
The embodiment of the invention provides a voice control method of a water dispenser, which further comprises the following steps:
acquiring first time for analyzing a first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
inquiring a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on first complexity according to the interval formulation rule, and formulating a second complexity interval based on second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity falling within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein the calculation formula is as follows:
Figure BDA0003052606450000131
wherein ε is a demand index, ArIs the fourth complexity, X1Is the total number of the fourth complexity, BrIs the fifth complexity, X2Is the total number of the fifth complexity, O1To a first complexity, O2A second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset complexity comparison table specifically comprises: a plurality of control items, each control item comprising: time interval and complexity; when the complexity comparison table is inquired, the corresponding complexity can be determined when the time of use falls within a certain time interval; the longer the analysis time is, the higher the complexity is; the preset interval formulation rule is specifically as follows: the numerical value is increased by a certain value to be used as an upper limit of the interval, and the numerical value is decreased by a certain value to be used as a lower limit of the interval; the preset complexity record database specifically comprises: historically, according to a record of the complexity determined by the resolution user; the preset time period is as follows: the first 60 minutes; generally, when a voice control instruction input by a user is analyzed, a local semantic recognition database (a database with a large amount of data for semantic recognition stored in the database) is called to recognize the voice control instruction, however, when the water dispenser is applied to some special scenes (for example, the user Mandarin is not very standard and the like), the system analysis duration is increased, the requirement index is calculated based on each complexity, the greater the requirement index is, the greater the analysis difficulty is, and when the requirement index is greater than or equal to a preset requirement index threshold (for example, 95), the control system updates the local semantic recognition database to improve the semantic recognition capability of the water dispenser; and when the voice recognition model is analyzed again, the networking recognition model is called, and the model can be combined with the network semantic recognition data to recognize some special voices (such as dialects).
The embodiment of the invention can automatically determine whether the current capability of the local semantic recognition database is enough, and automatically update the local semantic recognition database or call the networking recognition model to analyze the voice control instruction when the current capability of the local semantic recognition database is not enough, so that the networking is not needed, the power consumption of equipment is reduced, and meanwhile, the invention is more intelligent.
The embodiment of the invention provides a voice control system of a water dispenser, as shown in figure 2, comprising:
the system comprises an acquisition module 1, a processing module and a control module, wherein the acquisition module is used for acquiring a first voice control instruction input by a user;
the first analysis module 2 is used for analyzing the first voice control instruction, and acquiring a second voice control instruction input by a user if the first voice control instruction is a wake-up instruction;
the second analysis module 3 is used for analyzing the second voice control instruction and determining the type of the beverage which the user wants to brew;
the determining module 4 is used for determining the answer voice and the target water temperature corresponding to the beverage type;
and the control module 5 is used for playing the reply voice and controlling the water dispenser to adjust the water temperature to the target water temperature.
The working principle and the beneficial effects of the technical scheme are as follows:
the first voice control command input by the user for the first time is analyzed, when the command is a wake-up command (for example: 'hello | drinking fountain'), a second voice control command input by the user again is obtained, analyzing the instruction, determining the type of the beverage which the user wants to brew, determining the response voice and the target water temperature corresponding to the type of the beverage (for example, if the second voice control instruction is ' I want to brew a cup of green tea ', the type of the beverage is green tea, the response voice corresponding to the green tea is ' good ', the proper temperature for brewing the green tea is 85 ℃, heating is started ', and the target water temperature corresponding to the green tea is 85 ℃), playing the response voice to reply the user, meanwhile, the water dispenser is controlled to adjust the water temperature to the target water temperature, and after the water temperature is adjusted, a notification voice of successful adjustment can be played (for example, the green tea brewing water is ready for use).
According to the embodiment of the invention, when a user uses the water dispenser to brew different types of beverages, only the awakening instruction and the type of the beverage to be brewed need to be spoken, and the water dispenser automatically prepares water with a corresponding appropriate temperature for the user to use, so that the water dispenser is very convenient and fast.
The embodiment of the invention provides a voice control system of a water dispenser, wherein an acquisition module 1 executes the following operations:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000151
Figure BDA0003052606450000152
wherein, theta1As a first evaluation index, mu1And mu2Is a preset weight value, alpha1，iRecording the ith first evaluation value n output after evaluation by the first evaluation model for voice control1To record for speech control the total number of first evaluation values output after evaluation by the first evaluation model, a1,0A preset first evaluation value threshold value is set, and e is a natural constant;
acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure BDA0003052606450000153
Figure BDA0003052606450000154
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for voice control2For recording the total number of second evaluation values, alpha, output after evaluation by the second evaluation model for speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first judgment index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure BDA0003052606450000155
wherein, γ1Is a first decision index, σ1And σ2Is a preset weight value, theta1Is a first evaluation index, θ2Is a second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring a first voice control instruction.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset voice control record specifically comprises the following steps: after the water dispenser is put into use, the received voice control instruction and the record of executing the corresponding operation are recorded; the preset first evaluation model specifically includes: the evaluation model can evaluate whether the phenomenon of child disorganization (such as imitating that an adult frequently speaks a wake-up instruction and/or a brewing beverage type instruction and the like) occurs in the last short time (such as 5 minutes) based on the voice control record, and automatically output a first evaluation value after evaluation, wherein the smaller the first evaluation value is, the greater the possibility of child disorganization is; the preset second evaluation model specifically comprises: a model generated after learning a large amount of records of the adult voice control disorderly operation by using a machine learning algorithm, wherein the evaluation model can evaluate whether the phenomenon of the adult disorderly operation (such as repeatedly speaking the same brewing beverage type instruction and the like) occurs in the last short time (such as 5 minutes) based on the voice control records, and after evaluation, a second evaluation value is automatically output, and the smaller the second evaluation value is, the more possibility of the adult disorderly operation is shown to occur; the purpose of evaluating the entire speech control recording is to evaluate whether the features (e.g., voiceprints, etc.) characterized by the models in conjunction with the most recently occurring instructions have previously appeared in combination, for example: if the operation happens, the possibility of operation disorder is high; in order to avoid the problem of accidental errors, the first evaluation model and the second evaluation model carry out multiple evaluations and output multiple evaluation values; calculating a first evaluation index based on each first evaluation value, wherein the smaller the first evaluation index is, the greater the overall possibility of occurrence of child disorientation is; calculating a second evaluation index based on each second evaluation value, the smaller the second evaluation index is, the greater the overall possibility that the adult dishonest operation occurs is indicated to be; calculating a first judgment index based on the first evaluation index and the second evaluation index, and acquiring a first voice control instruction when the first judgment index is greater than or equal to a preset first judgment index threshold (for example: 85); the preset first evaluation value threshold is specifically: for example, 85; the preset second evaluation value threshold is specifically: such as 86.
According to the embodiment of the invention, the first evaluation model and the second evaluation model are used for evaluating the voice control record, so that child random operation and/or adult random operation are intelligently ignored, the use power consumption is reduced, and the user experience is improved.
The embodiment of the invention provides a voice control system of a water dispenser, and a determination module 4 executes the following operations:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the type of the drink and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second partition.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset pairing database specifically comprises: a plurality of pairing items, each pairing item comprising: beverage type, answer voice and target water temperature; the pairing database is divided into two partitions, namely a first partition and a second partition; the matching database contains suitable water temperatures corresponding to various beverage types (such as green tea, jasmine tea, white tea, oolong tea, black tea, Pu' er tea, yellow tea, oolong tea, milk powder, coffee, milk tea, cocoa powder, fruit juice, coconut powder, matcha powder, ginger tea, etc.); when the answer voice and the target water temperature corresponding to the beverage type are determined, the answer voice and the target water temperature corresponding to the beverage type are preferentially determined from the first partition, and when the determination fails, the answer voice and the target water temperature corresponding to the beverage type are determined from the second partition.
The pairing database of the embodiment of the invention contains the appropriate water temperatures corresponding to various beverage types, so that the requirements of users can be met to the greatest extent, and the user experience is improved.
The embodiment of the invention provides a voice control system of a water dispenser, which further comprises:
the filling module is used for filling the first partition and the second partition before the determining module 4 determines the answer voice and the target water temperature corresponding to the beverage type;
the filling module executes the following operations:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first associated data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after a first perception model perceives first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure BDA0003052606450000171
Figure BDA0003052606450000172
wherein, P1Is a first perception index, J1And J2Is a preset weight value, q1,tThe tth first associated data is output after being sensed by the first sensing modelPerception value, z1A total number q of first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second perception model perceives the second associated data;
calculating a second perception index based on the second perception value, the calculation formula being as follows:
Figure BDA0003052606450000181
Figure BDA0003052606450000182
wherein, P2Is a second perception index, J1And J2Is a preset weight value, q2,tA t second perception value, z, output after the second associated data is perceived by the second perception model2The total number of second perception values q output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure BDA0003052606450000183
wherein, γ2Is a second determination index, ω1And ω2Is a preset weight value, P1Is a first perceptual index, P2Is a second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset pairing item list specifically includes: a plurality of pairing items, each pairing item comprising: beverage type, answer voice and target water temperature; the preset first obtaining path specifically includes: the system is connected with a local beverage type selection record database (a database for recording the beverage type record selected by the user) and used for acquiring local beverage type selection record data; the first associated data is data associated with the selected pairing item in the local beverage type selection record data; the preset first perception model specifically includes: the model is generated after a large amount of local beverage type selection record data and manual judgment preference records are learned by utilizing a machine learning algorithm, the model can sense first associated data, the preference degree of a user for selecting the pairing item is sensed, a first sensing value is output, and the larger the first sensing value is, the more the user prefers the pairing item; the preset second obtaining path specifically includes: the beverage type selection recording data of each user can be obtained through the path, namely big data; the second associated data is data associated with the selected pairing item in the big data; the preset second perception model specifically includes: the model is generated after a large amount of beverage type selection record data and manual evaluation preference records are learned by utilizing a machine learning algorithm, the model can perform data elimination and other operations based on the credibility of a certain data source in big data, second associated data can be sensed, the preference degree of different users for selecting the pairing item is sensed, a second sensing value is output, and the greater the second sensing value is, the more most users prefer the pairing item; calculating a first perception index based on each first perception value, wherein the larger the first perception index is, the higher the overall degree of preference of the user of the water dispenser for the pairing item is; calculating a second perception index based on each second perception value, wherein the larger the second perception index is, the higher the overall degree of preference of most users for the paired items is; calculating a first perception index based on the first perception index and the second perception indexA second decision index, which indicates that the pair is highly likely to be selected when the second decision index is greater than or equal to a preset second decision index threshold (e.g., 90), and should be listed in the first partition for priority determination; in calculating the second decision index threshold, ω is the value of ω because of the emphasis on recording data in conjunction with the user's local selection1Should be greater than omega2(ii) a The preset first perception value threshold specifically includes: for example, 88; the preset second perception value threshold specifically includes: such as 89.
The embodiment of the invention can intelligently decide whether the selected matching item is filled into the first partition or the second partition based on the first correlation data and the second correlation data, can quickly list the selected matching item with high possibility into the first partition for priority determination, greatly reduces the determination time when determining the response voice and the target water temperature corresponding to the beverage type, improves the working efficiency of the system, reduces the delay of the system for replying the user and adopting temperature regulation control, and improves the user experience.
The embodiment of the invention provides a voice control system of a water dispenser, which further comprises:
a demand determination module 4;
the requirement determining module 4 performs operations including:
acquiring first time for analyzing a first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
inquiring a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on first complexity according to the interval formulation rule, and formulating a second complexity interval based on second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity falling within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein the calculation formula is as follows:
Figure BDA0003052606450000191
wherein ε is a demand index, ArIs the fourth complexity, X1Is the total number of the fourth complexity, BrIs the fifth complexity, X2Is the total number of the fifth complexity, O1To a first complexity, O2A second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
The working principle and the beneficial effects of the technical scheme are as follows:
the preset complexity comparison table specifically comprises: a plurality of control items, each control item comprising: time interval and complexity; when the complexity comparison table is inquired, the corresponding complexity can be determined when the time of use falls within a certain time interval; the longer the analysis time is, the higher the complexity is; the preset interval formulation rule is specifically as follows: the numerical value is increased by a certain value to be used as an upper limit of the interval, and the numerical value is decreased by a certain value to be used as a lower limit of the interval; the preset complexity record database specifically comprises: historically, according to a record of the complexity determined by the resolution user; the preset time period is as follows: the first 60 minutes; generally, when a voice control instruction input by a user is analyzed, a local semantic recognition database (a database with a large amount of data for semantic recognition stored in the database) is called to recognize the voice control instruction, however, when the water dispenser is applied to some special scenes (for example, the user Mandarin is not very standard and the like), the system analysis duration is increased, the requirement index is calculated based on each complexity, the greater the requirement index is, the greater the analysis difficulty is, and when the requirement index is greater than or equal to a preset requirement index threshold (for example, 95), the control system updates the local semantic recognition database to improve the semantic recognition capability of the water dispenser; and when the voice recognition model is analyzed again, the networking recognition model is called, and the model can be combined with the network semantic recognition data to recognize some special voices (such as dialects).
The embodiment of the invention can automatically determine whether the current capability of the local semantic recognition database is enough, and automatically update the local semantic recognition database or call the networking recognition model to analyze the voice control instruction when the current capability of the local semantic recognition database is not enough, so that the networking is not needed, the power consumption of equipment is reduced, and meanwhile, the invention is more intelligent.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims (10)

1. A voice control method of a water dispenser is characterized by comprising the following steps:
acquiring a first voice control instruction input by a user;
analyzing the first voice control instruction, and if the first voice control instruction is a wake-up instruction, acquiring a second voice control instruction input by the user;
analyzing the second voice control instruction, and determining the type of the beverage which the user wants to brew;
determining a reply voice and a target water temperature corresponding to the beverage type;
and playing the reply voice, and simultaneously controlling the water dispenser to adjust the water temperature to the target water temperature.
2. The voice control method of the water dispenser as claimed in claim 1, wherein the obtaining of the first voice control instruction input by the user comprises:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, wherein the calculation formula is as follows:
Figure FDA0003052606440000011
Figure FDA0003052606440000012
wherein, theta1Is the first evaluation index, mu1And mu2Is a preset weight value, alpha1，iRecording the ith first evaluation value n output after evaluation by the first evaluation model for the voice control1Recording the total number of the first evaluation values, alpha, output after evaluation by the first evaluation model for the speech control1，0A preset first evaluation value threshold value is set, and e is a natural constant;
acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure FDA0003052606440000013
Figure FDA0003052606440000021
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for the voice control2Recording the total number of the second evaluation values, alpha, output after evaluation by the second evaluation model for the speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first decision index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure FDA0003052606440000022
wherein, γ1Is the first decision index, σ1And σ2Is a preset weight value, theta1Is the first evaluation index, θ2Is the second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring the first voice control instruction.
3. The voice control method for the water dispenser as claimed in claim 1, wherein the step of determining the reply voice and the target water temperature corresponding to the type of the beverage comprises the following steps:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the beverage type and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second subarea.
4. The voice control method for the water dispenser as claimed in claim 3, characterized in that the first partition and the second partition are filled before the answer voice corresponding to the type of the drink and the target water temperature are determined;
the filling steps of the first partition and the second partition are as follows:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first association data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after the first perception model perceives the first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure FDA0003052606440000023
Figure FDA0003052606440000024
wherein, P1Is the first perception index, J1And J2Is a preset weight value, q1,tThe t-th first perception value, z, output after the first associated data is perceived by the first perception model1The total number q of the first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second associated data are perceived by the second perception model;
calculating a second perception index based on the second perception value, wherein the calculation formula is as follows:
Figure FDA0003052606440000031
Figure FDA0003052606440000032
wherein, P2Is the second perception index, J1And J2Is a preset weight value, q2,tThe t-th second perception value, z, output after the second associated data is perceived by the second perception model2The total number q of the second perception values output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure FDA0003052606440000033
wherein, γ2Is the second determination index, ω1And ω2Is a preset weight value, P1Is the first perception index, P2Is the second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
5. The voice control method of the water dispenser as claimed in claim 1, further comprising:
acquiring first time for analyzing the first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
querying a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on the first complexity according to the interval formulation rule, and formulating a second complexity interval based on the second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity that falls within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein a calculation formula is as follows:
Figure FDA0003052606440000041
wherein ε is the demand index, ArFor the fourth complexity, X, of the r1Is the total number of the fourth complexity, BrIs the fifth complexity, X, of the r2Is the total number of the fifth complexity, O1To the first complexity, O2The second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
6. A voice control system of a water dispenser is characterized by comprising:
the acquisition module is used for acquiring a first voice control instruction input by a user;
the first analysis module is used for analyzing the first voice control instruction, and acquiring a second voice control instruction input by the user if the first voice control instruction is a wake-up instruction;
the second analysis module is used for analyzing the second voice control instruction and determining the type of the beverage which the user wants to brew;
the determining module is used for determining the answer voice and the target water temperature corresponding to the beverage type;
and the control module is used for playing the reply voice and controlling the water dispenser to adjust the water temperature to the target water temperature.
7. The voice control system of the water dispenser of claim 6, wherein the obtaining module performs operations comprising:
acquiring a preset voice control record;
acquiring a preset first evaluation model, and evaluating the voice control record by adopting the first evaluation model;
acquiring a plurality of first evaluation values output after the first evaluation model evaluates the voice control record;
calculating a first evaluation index based on the first evaluation value, wherein the calculation formula is as follows:
Figure FDA0003052606440000042
Figure FDA0003052606440000051
wherein, theta1Is the first evaluation index, mu1And mu2Is a preset weight value, alpha1,iRecord for the voice control via the secondAn ith first evaluation value n output after evaluation by an evaluation model1Recording the total number of the first evaluation values, alpha, output after evaluation by the first evaluation model for the speech control1,0A preset first evaluation value threshold value is set, and e is a natural constant;
acquiring a preset second evaluation model, and evaluating the voice control record by adopting the second evaluation model;
acquiring a plurality of second evaluation values output after the second evaluation model evaluates the voice control record;
calculating a second evaluation index based on the second evaluation value, the calculation formula being as follows:
Figure FDA0003052606440000052
Figure FDA0003052606440000053
wherein, theta2For the second evaluation index, μ1And mu2Is a preset weight value, alpha2,iRecording the ith second evaluation value n output after evaluation by the second evaluation model for the voice control2Recording the total number of the second evaluation values, alpha, output after evaluation by the second evaluation model for the speech control2,0E is a natural constant, and is a preset second evaluation value threshold;
calculating a first decision index based on the first evaluation index and the second evaluation index, the calculation formula being as follows:
Figure FDA0003052606440000054
wherein, γ1Is the first decision index, σ1And σ2Is a preset weight value, theta1Is the first evaluation index，θ2Is the second evaluation index;
and when the first judgment index is greater than or equal to a preset first judgment index threshold value, acquiring the first voice control instruction.
8. The voice control system of the water dispenser of claim 6, wherein the determining module performs operations comprising:
acquiring a preset pairing database, wherein the pairing database comprises: a first partition and a second partition;
determining a reply voice corresponding to the beverage type and a target water temperature from the first partition;
and if the determination fails, determining a reply voice corresponding to the type of the drink and the target water temperature from the second subarea.
9. The voice control system of the water dispenser of claim 8, further comprising:
the filling module is used for filling the first partition and the second partition before the determining module determines the answer voice and the target water temperature corresponding to the beverage type;
the filling module executes the following operations:
acquiring a preset pairing item list, and selecting any pairing item from the pairing item list;
acquiring first association data associated with the pairing item through a preset first acquisition path;
acquiring a preset first perception model, and perceiving the first associated data by adopting the first perception model;
acquiring a plurality of first perception values output after the first perception model perceives the first associated data;
calculating a first perception index based on the first perception value, wherein the calculation formula is as follows:
Figure FDA0003052606440000061
Figure FDA0003052606440000062
wherein, P1Is the first perception index, J1And J2Is a preset weight value, q1,tThe t-th first perception value, z, output after the first associated data is perceived by the first perception model1The total number q of the first perception values output after the first associated data are perceived by the first perception model1,0Is a preset first perception value threshold value, and e is a natural constant;
acquiring second associated data associated with the pairing item through a preset second acquisition path;
acquiring a preset second perception model, and perceiving the second associated data by adopting the second perception model;
acquiring a plurality of second perception values output after the second associated data are perceived by the second perception model;
calculating a second perception index based on the second perception value, wherein the calculation formula is as follows:
Figure FDA0003052606440000063
Figure FDA0003052606440000064
wherein, P2Is the second perception index, J1And J2Is a preset weight value, q2,tThe t-th second perception value, z, output after the second associated data is perceived by the second perception model2The total number q of the second perception values output after the second associated data are perceived by the second perception model2,0The value is a preset second perception value threshold value, and e is a natural constant;
calculating a second decision index based on the first perception index and the second perception index, the calculation formula being as follows:
Figure FDA0003052606440000065
wherein, γ2Is the second determination index, ω1And ω2Is a preset weight value, P1Is the first perception index, P2Is the second perception index;
and when the second judgment index is larger than or equal to a preset second judgment index threshold value, filling the pairing item into the first partition, otherwise, filling the pairing item into the second partition.
10. The voice control system of the water dispenser of claim 6, further comprising:
a demand determination module;
the requirement determining module performs operations comprising:
acquiring first time for analyzing the first voice control instruction;
acquiring second time for analyzing the second voice control instruction;
querying a preset complexity comparison table, and respectively determining a first complexity corresponding to the first time and a second complexity corresponding to the second time;
acquiring a preset interval formulation rule, formulating a first complexity interval based on the first complexity according to the interval formulation rule, and formulating a second complexity interval based on the second complexity according to the interval formulation rule;
acquiring a preset complexity record database, and determining a plurality of third complexities corresponding to a preset time period from the complexity record database;
taking the third complexity falling within the first complexity interval as a fourth complexity;
taking the third complexity that falls within the second complexity interval as a fifth complexity;
calculating a demand index based on the first complexity, the second complexity, the fourth complexity and the fifth complexity, wherein a calculation formula is as follows:
Figure FDA0003052606440000071
wherein ε is the demand index, ArFor the fourth complexity, X, of the r1Is the total number of the fourth complexity, BrIs the fifth complexity, X, of the r2Is the total number of the fifth complexity, O1To the first complexity, O2The second complexity;
and when the demand index is larger than or equal to a preset demand index threshold value, updating a preset local semantic recognition database, and/or calling a preset networking recognition model to analyze the first voice control instruction or the second voice control instruction when the first voice control instruction or the second voice control instruction is analyzed next time.
CN202110491761.1A 2021-05-06 2021-05-06 Voice control method and system for water dispenser Withdrawn CN113223526A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110491761.1A CN113223526A (en) 2021-05-06 2021-05-06 Voice control method and system for water dispenser

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110491761.1A CN113223526A (en) 2021-05-06 2021-05-06 Voice control method and system for water dispenser

Publications (1)

Publication Number Publication Date
CN113223526A true CN113223526A (en) 2021-08-06

Family

ID=77091043

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110491761.1A Withdrawn CN113223526A (en) 2021-05-06 2021-05-06 Voice control method and system for water dispenser

Country Status (1)

Country Link
CN (1) CN113223526A (en)

Similar Documents

Publication Publication Date Title
CN105520457B (en) Control Intelligent tea kettle brews method, device and the Intelligent tea kettle of tea
CN105117007B (en) Show control method, device and the intelligent pad of equipment
CN107863138B (en) Menu generating device and method
CN105550223B (en) User recommendation method and device
CN108826824B (en) Refrigerator control method and device, storage medium and refrigerator
CN106485052A (en) A kind of intelligent cooking system and method
CN105534209B (en) Control the method, apparatus and Intelligent tea kettle of Intelligent tea kettle blunt tea water
CN107809667A (en) Television voice exchange method, interactive voice control device and storage medium
CN105959749A (en) Intelligent terminal, remote controller and recommending method and system
CN104983225B (en) Method of adjustment, device and the terminal of pad
CN107886348A (en) A kind of artificial intelligence method of ordering and system
WO2015051462A1 (en) System and method for classifying food products
CN110045638A (en) A kind of recommended method of cooking information, device and storage medium
CN106033199A (en) Liquid supply device and control method thereof
CN111667261B (en) Intelligent voice dialogue household wine cabinet
CN114517136A (en) Control method and device for beer preparation equipment and beer preparation equipment
CN105536264A (en) User-interaction toy and interaction method of the toy
CN113223526A (en) Voice control method and system for water dispenser
JP6344784B1 (en) Recipe information providing apparatus, recipe information providing method, and recipe information providing program
CN111435229A (en) Method and device for controlling cooking mode and cooking appliance
KR20220006837A (en) Method and apparatus for recommending indivudual beverage based on face recognition
KR20150016046A (en) method and system for meal information entry
CN113223525A (en) Voice control method and system for intelligent water dispenser under multiple instructions
CN113223523A (en) Voice recognition method and system for intelligent water dispenser
JP2019087225A (en) Recipe information provision device, recipe information provision method and recipe information provision program

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication
WW01 Invention patent application withdrawn after publication

Application publication date: 20210806