CN110480656A

CN110480656A - One kind is accompanied and attended to robot, accompany and attend to robot control method and device

Info

Publication number: CN110480656A
Application number: CN201910848250.3A
Authority: CN
Inventors: 张腾宇; 张静莎
Original assignee: National Research Center for Rehabilitation Technical Aids
Current assignee: National Research Center for Rehabilitation Technical Aids
Priority date: 2019-09-09
Filing date: 2019-09-09
Publication date: 2019-11-22
Anticipated expiration: 2039-09-09
Also published as: CN110480656B

Abstract

It accompanies and attends to robot, accompany and attend to robot control method and device, including physical characteristics collecting module, processor this application provides one kind；The physical characteristics collecting module, for acquiring the biological information of user, and the biological information is transmitted to the processor every preset duration；The processor for receiving the biological information of the physical characteristics collecting module transmission, and is based on the biological information, predicts the mood of the user；And the mood predicted is input in preparatory trained first branch prediction model, the determining control instruction to match with the biological information, and robot of accompanying and attending to described in control makes operation corresponding with the control instruction.By this robot of accompanying and attending to, the precision of user demand branch prediction can be improved.

Description

One kind is accompanied and attended to robot, accompany and attend to robot control method and device

Technical field

This application involves automatic control technology field, accompany and attend to robot, robot controlling party of accompanying and attending to more particularly, to one kind Method and device.

Background technique

With the aggravation of Chinese population aging process and increasing for only child, many children can not accompany constantly Old parent at one's side, this results in " Empty nest elderly " proportion in the elderly to increase day by day.The mental health of the elderly Problem has become important social concern.According to statistics, there is the symptom of feeling of lonely, especially body in 70% or more the elderly Weak, disability, solitary, advanced age, the death of one's spouse the elderly.

In order to solve the problems, such as that the elderly's emotion is accompanied and attended to and life care, in recent years it has been proposed that being replaced using robot People carries out the elderly and accompanies and attends to.Existing robot of accompanying and attending to can only generally be inputted when nursing to the elderly according to the elderly Instruction, execute corresponding operation, however required service may be not identical under different moods because of the elderly, it is this Do not consider that the mode that user emotion variation provides service only according to instruction is not able to satisfy the demand that the elderly's emotion is accompanied and attended to.

Summary of the invention

In view of this, the one kind that is designed to provide of the application is accompanied and attended to robot, accompany and attend to robot control method and device, To provide the service of reflection user's real feelings demand, and improve the precision of user demand branch prediction.

It accompanies and attends to robot in a first aspect, the embodiment of the present application provides one kind, the robot of accompanying and attending to includes biological characteristic Acquisition module, processor；

The physical characteristics collecting module, for the biological information every preset duration acquisition user, and will be described Biological information is transmitted to the processor；

The processor for receiving the biological information of the physical characteristics collecting module transmission, and is based on The biological information predicts the mood of the user；And the mood predicted is input to trained first finger in advance It enables in prediction model, the determining control instruction to match with the biological information, and robot of accompanying and attending to described in control makes Operation corresponding with the control instruction.

At least one of in a kind of possible design, the physical characteristics collecting module comprises the following modules:

Sound acquisition module, image capture module；

Include the case where sound acquisition module for the physical characteristics collecting module, the biological information includes sound Message breath；

Include the case where that image capture module, the biological information include face for the physical characteristics collecting module Portion's image information.

In a kind of possible design, the processor is being based on the biological information, is predicting the mood of the user When, it is specifically used for:

The phonetic feature in the acoustic information is extracted, the phonetic feature includes short-time energy, short-time zero-crossing rate, fundamental tone Frequency, formant feature, word speed, mel cepstrum coefficients；

The phonetic feature is input to speech recognition submodel, the phonetic feature is obtained and belongs to any one default feelings First score of thread；

The acoustic information is input to semantics recognition submodel, extracts the semantic key words in the acoustic information；

Based on the semantic key words, determine that the acoustic information belongs to the second score of any one default mood；

The facial image information is input in face recognition submodel, determines that the facial image information belongs to arbitrarily A kind of third score of default mood；

First score, second score, the third score are weighted summation according to default weight, are based on Score after summation determines the mood of the user.

In a kind of possible design, the robot of accompanying and attending to further include: receiving module and memory module；

The receiving module is transmitted to the processor for receiving the instruction of user, and by described instruction；

The processor, is also used to: when receiving the described instruction of the receiving module transmission, it is special to control the biology Sign acquisition module acquires the biological information of the user；The behavior letter of the user is determined based on the biological information Breath, and by described instruction, the time of reception described instruction and the behavioural information, it is transmitted to the memory module；The use The behavioural information at family is used to indicate the active state of the user；

The memory module, for storing described instruction, the time for receiving described instruction and the behavioural information.

In a kind of possible design, the processor is also used to train the first branch prediction mould by the following method Type:

Obtain at least one history emotional prediction of the user as a result, and each history emotional prediction result it is associated Instruction；

The history emotional prediction result is input in branch prediction model to be trained, it is pre- to obtain the history mood Survey the corresponding prediction instruction of result；

Based on prediction instruction and instruction associated with the history emotional prediction result, to described to be trained Branch prediction model carries out epicycle training；

By more wheels training to described instruction prediction model, the first branch prediction model is obtained.

In a kind of possible design, the processor is also used to after receiving the biological information:

Based on the biological information, the behavioural information of the user is determined；

By the behavioural information of the user and the time of the reception biological information, it is input to and trains in advance The second branch prediction model in, the determining control instruction to match with the biological information, and machine of accompanying and attending to described in controlling Device people makes operation corresponding with the control instruction.

In a kind of possible design, the processor trains the second branch prediction model in accordance with the following methods:

The instruction of memory module storage is obtained, the time of described instruction is received and described specifies corresponding behavior Information；

By the time for receiving described instruction and described corresponding behavioural information is specified to be input to prediction to be trained In model, output obtains prediction instruction；

It is instructed based on the instruction obtained from the memory module and the prediction, epicycle is carried out to the prediction model Training；

By more wheels training to the prediction model, the second branch prediction model is obtained.

In a kind of possible design, the robot of accompanying and attending to further include: voice synthetic module；

The voice synthetic module is issued for extracting the audio frequency characteristics of template voice, and receiving the processor Voice play instruction when, based on the audio frequency characteristics issue voice.

In a kind of possible design, the robot of accompanying and attending to further include: alarm module；

The alarm module, for carrying out the behavioural information with abnormal behaviour information in the memory module is stored in Comparison sends warning message to the equipment bound in advance when comparing successfully.

In a kind of possible design, the robot of accompanying and attending to, further includes: apart from detection module, mobile module；

It is described apart from detection module, for detecting accompany and attend to the distance between the robot and the user, and will described in Distance is sent to the mobile module；

The mobile module, for court, robot of when detecting that the distance is greater than pre-determined distance, accompanying and attending to described in control It is moved to the position where the user.

In a kind of possible design, the robot of accompanying and attending to further include: clean module；

The cleaning module, for controlling and the machine of accompanying and attending to when receiving the cleaning instruction that the processor is sent The sweeping robot of device people connection cleans.

In a kind of possible design, infrared positioning is both provided on accompany and attend to robot and the sweeping machine human agent Device；

The sweeping robot complete clean after, based on the infrared positioning apparatus again with the robot of accompanying and attending to Establish connection.

Second aspect, the embodiment of the present application also provide one kind and accompany and attend to robot control method, comprising:

Receive the biological information of user；

Based on the biological information, the mood of the user is predicted；

The mood predicted is input in preparatory trained first branch prediction model, the determining and biological characteristic The control instruction that information matches；

Robot of accompanying and attending to described in control makes operation corresponding with the control instruction.

The third aspect is accompanied and attended to robot controller the embodiment of the present application also provides one kind, comprising:

Receiving module, for receiving the biological information of user；

Prediction module predicts the mood of the user for being based on the biological information；

Determining module is determined for the mood predicted to be input in preparatory trained first branch prediction model The control instruction to match with the biological information；

Control module makes operation corresponding with the control instruction for controlling the robot of accompanying and attending to.

Fourth aspect, the embodiment of the present application also provide a kind of electronic equipment, comprising: processor, memory and bus, it is described Memory is stored with the executable machine readable instructions of the processor, when electronic equipment operation, the processor with it is described By bus communication between memory, the machine readable instructions are executed when being executed by the processor in above-mentioned second aspect Step.

5th aspect, the embodiment of the present application also provide a kind of computer readable storage medium, the computer-readable storage medium Computer program is stored in matter, which executes the step in above-mentioned second aspect when being run by processor.

Robot provided by the embodiments of the present application of accompanying and attending to, accompany and attend to robot control method and device, are accompanied and attended to by being deployed in The biological information of physical characteristics collecting module acquisition user in robot, is then based on the processing being deployed in robot Device carries out emotional prediction according to biological information, and the mood at prediction is input to trained first instruction in advance in advance It surveys in model, the determining control instruction to match with biological information, this control instruction determines that method combines user and works as Preceding mood, therefore the instruction determined is more accurate, when control robot makes operation corresponding with control instruction, The operation made also more meets user demand.

To enable the above objects, features, and advantages of the application to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate Appended attached drawing, is described in detail below.

Detailed description of the invention

Technical solution in ord to more clearly illustrate embodiments of the present application, below will be to needed in the embodiment attached Figure is briefly described, it should be understood that the following drawings illustrates only some embodiments of the application, therefore is not construed as pair The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this A little attached drawings obtain other relevant attached drawings.

Fig. 1 shows a kind of configuration diagram for robot of accompanying and attending to provided by the embodiment of the present application；

Fig. 2 shows processors provided by the embodiment of the present application in the biology for receiving physical characteristics collecting module transmission Processing flow schematic diagram after characteristic information；

Fig. 3 shows a kind of flow diagram of the method for the mood for predicting user provided by the embodiment of the present application；

Fig. 4 shows the flow diagram of the first branch prediction model training method provided by the embodiment of the present application；

Fig. 5 shows the configuration diagram of the robot of accompanying and attending to of another possibility provided by the embodiment of the present application；

Fig. 6 shows the flow diagram of the training method of the second branch prediction model provided by the embodiment of the present application；

Fig. 7 shows a kind of flow diagram for robot control method of accompanying and attending to provided by the embodiment of the present application；

Fig. 8 shows a kind of configuration diagram for robot controller of accompanying and attending to provided by the embodiment of the present application；

Fig. 9 shows the structural schematic diagram of electronic equipment 900 provided by the embodiment of the present application.

Specific embodiment

To keep the purposes, technical schemes and advantages of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application Middle attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only It is some embodiments of the present application, instead of all the embodiments.The application being usually described and illustrated herein in the accompanying drawings is real The component for applying example can be arranged and be designed with a variety of different configurations.Therefore, below to the application's provided in the accompanying drawings The detailed description of embodiment is not intended to limit claimed scope of the present application, but is merely representative of the selected reality of the application Apply example.Based on embodiments herein, those skilled in the art institute obtained without making creative work There are other embodiments, shall fall in the protection scope of this application.

For convenient for understanding the present embodiment, the robot of accompanying and attending to of one kind disclosed in the embodiment of the present application is carried out first It is discussed in detail.

It is shown in Figure 1, it is a kind of configuration diagram for robot of accompanying and attending to provided by the embodiment of the present application, including biology Collection apparatus module and processor.

Specifically, physical characteristics collecting module is used to acquire the biological information of user every preset duration, then will The biological information of acquisition is transmitted to processor.

Wherein, physical characteristics collecting module may include at least one of sound acquisition module, image capture module.

When physical characteristics collecting module includes sound acquisition module, biological information includes acoustic information, works as biology When collection apparatus module includes image capture module, biological information includes facial image information.

Processor can execute after the biological information for receiving physical characteristics collecting module transmission such as Fig. 2 institute The treatment process stated, including the following steps:

Step 201 is based on biological information, predicts the mood of user.

In a kind of possible embodiment, processor is being based on biological information, can be with when predicting the mood of user Pass through method as shown in Figure 3, comprising:

Phonetic feature in step 301, extraction acoustic information.

Wherein, the phonetic feature in acoustic information include short-time energy, short-time zero-crossing rate, fundamental frequency, formant feature, Word speed, mel cepstrum coefficients, the specific method for extracting the phonetic feature in acoustic information will not describe herein in expansion.

Phonetic feature is input to speech recognition submodel by step 302, is obtained phonetic feature and is belonged to any one default feelings First score of thread.

In a kind of possible embodiment, default mood may include: it is happy, sad, irritated, angry etc., by voice After feature is input in speech recognition submodel, the available phonetic feature belongs to the first of any one default mood Point.

Illustratively, after one section of phonetic feature is input in speech recognition submodel, it is likely to be obtained voice spy It is 80 that sign, which belongs to the first happy score, and belonging to the first sad score is 30, and belonging to the first irritated score is 20, belongs to anger First score of anger is 10.

Acoustic information is input to semantics recognition submodel by step 303, extracts the semantic key words in acoustic information.

It is semantic after acoustic information is input to semantics recognition submodel in a kind of possible embodiment of the application The acoustic information first can be converted to corresponding text information by identification submodel, then utilize the N in semantics recognition submodel Meta-model, to text information carry out cutting, obtain set of words corresponding to text information, then by the set of words with Pre-stored keyword compares in database, will compare successful word and is determined as semantic key words.

Step 304 is based on semantic key words, determines that acoustic information belongs to the second score of any one default mood.

Facial image information is input in face recognition submodel by step 305, determines that facial image information belongs to arbitrarily A kind of third score of default mood.

Specifically, face-image can also be extracted before facial image information is input in face recognition submodel Characteristics of image of information, such as geometrical characteristic, deformation characteristics, the motion feature at each position of face etc., then by characteristics of image It is input in face recognition submodel, determines that facial image information belongs to the third score of any one default mood.

Wherein, it should be noted that the execution of step 301- step 302, step 303- step 304 and step 305 is regardless of elder generation Sequence afterwards.

First score, the second score, third score are weighted summation according to default weight by step 306, based on summation Score afterwards determines the mood of user.

Wherein, based on the score after summation, when determining the mood of user, the mood of highest scoring can be determined as using The mood at family.

In a kind of possible application scenarios, robot of accompanying and attending to may only collect the acoustic information of user or only adopt The facial image information of user is collected, then when based on the mood of biological information prediction user, will can not collect Information corresponding to score be determined as 0.

Illustratively, if only having collected the acoustic information of user, third score is determined as 0, is determining user's feelings When the score of thread, the first score and the second score are weighted summation according to default weight.

In one example of the application, face recognition submodel, semantics recognition submodel and speech recognition submodel be can be One of support vector machines, convolutional neural networks, k nearest neighbor disaggregated model.

The mood predicted is input in preparatory trained first branch prediction model, determining and biology by step 202 The control instruction that characteristic information matches.

Step 203, control robot make operation corresponding with control instruction.

In a kind of possible embodiment, processor, can be according to such as Fig. 4 institute in the first branch prediction model of training The method shown is trained, including the following steps:

Step 401 obtains at least one history emotional prediction of user as a result, and each history emotional prediction result phase Associated instruction.

In a kind of possible design, robot of accompanying and attending to further includes memory module, and history mood is stored in memory module Prediction result and each associated instruction of history emotional prediction result, processor, can in training instruction prediction model To obtain history emotional prediction from memory module as a result, and each associated instruction of history emotional prediction result.

The history emotional prediction that memory module is stored is as a result, and each associated finger of history emotional prediction result Order can be, and physical characteristics collecting module is then communicated to processor every the biological information of preset duration acquisition user, Processor is predicted based on mood of the biological information to user, is receiving physical characteristics collecting module transmission again Before biological information, if receiving the instruction of user's input, the feelings that the instruction that user inputs is determined as and is predicted The associated instruction of thread, the instruction storage for then inputting the mood predicted and user to memory module.

History emotional prediction result is input in branch prediction model to be trained by step 402, and it is pre- to obtain history mood Survey the corresponding prediction instruction of result.

Step 403 is based on prediction instruction and instruction associated with history emotional prediction result, treats trained instruction Prediction model carries out epicycle training.

In specific implementation, this can be determined based on prediction instruction and instruction associated with history emotional prediction result Intersection entropy loss during wheel training is then based on the mould for intersecting and instructing prediction model in entropy loss adjustment epicycle training process Shape parameter.

Step 404 is trained by more wheels to branch prediction model, obtains the first branch prediction model.

In above process, the sample data in the training process of the first branch prediction model derives from the finger of user's input It enables and during instruction input each time, the user emotion predicted, therefore, first trained by this method refers to It enables prediction model meet the behavioural characteristic of active user, personalized service can be provided for user.

In a kind of possible application scenarios, if user includes more people, processor can be known according to biological information Not multiple users, and the emotional prediction result of different user and the associated instruction of each emotional prediction result are stored Into in different memory modules.Processor can be deposited out in module according to difference and be stored in the first branch prediction model of training Emotional prediction result and each associated instruction of emotional prediction result, training obtain the first different branch predictions Model.When processor receives biological information, can based on biological information carry out user's identification, then determine with The corresponding first branch prediction model of the user, and control instruction is predicted based on the first branch prediction model.

In scheme provided in this embodiment, physical characteristics collecting module acquires the biology of a user every preset duration Characteristic information, in the embodiment of another possibility, physical characteristics collecting module can also be received in robot of accompanying and attending to After the instruction of user's input, the biological information of user is acquired.

It is shown in Figure 5, it is the framework signal of the robot of accompanying and attending to of another possibility provided by the embodiment of the present application Figure, which can also include receiving module and memory module；

Receiving module is transmitted to processor for receiving the instruction of user, and by received instruction；

Processor can be also used for when receiving the instruction of receiving module transmission, and control physical characteristics collecting module is adopted Collect the biological information of user, and determine the behavioural information of user based on biological information, received instruction, reception are referred to The time of order and the behavioural information determined are all transmitted to memory module；

Memory module, the received instruction transmitted for storage processor, receive instruction time and determine Behavioural information.

Wherein, the behavioural information of user is used to indicate the active state of user, if behavioural information may include sitting quietly, sleeping Feel, get up, walking, tumble etc..

In a kind of possible embodiment, processor is also based on biology after receiving biological information Characteristic information determines the behavioural information of user, then by the behavioural information of user and receive biological information time, It is input in preparatory trained second branch prediction model, the determining control instruction to match with biological information, and controls It makes robot of accompanying and attending to and makes operation corresponding with control instruction.

In a kind of possible application scenarios, if user includes more people, processor can be known according to biological information Not multiple users, and different storages is sent by the time of the behavioural information of different user and reception biological information In module.Second branch prediction model can also be biological according to the behavioural information for the user that different memory modules are stored, reception The time of characteristic information and dependent instruction train the second different branch prediction models, when receiving biological information, root Corresponding second branch prediction model is determined according to biological information, and then control instruction is predicted.

In a kind of possible design, robot of accompanying and attending to can also include alarm module, and alarm module can believe behavior It ceases and is compared with the abnormal behaviour information of storage in a storage module, when comparing successfully, sent to the equipment bound in advance Warning message.

Wherein, the training method of the second branch prediction model is referred to method shown in fig. 6, including following step It is rapid:

Step 601, the instruction for obtaining memory module storage receive the time instructed and specify corresponding behavioural information.

Step 602 will receive the time instructed and corresponding behavioural information specified to be input to prediction model to be trained In, output obtains prediction instruction.

Step 603 is instructed based on the instruction and prediction obtained from memory module, carries out epicycle training to prediction model.

Wherein, based on instruction prediction instruction and obtained from memory module, epicycle training is carried out to prediction model When, it can be based on instruction prediction instruction and obtained from memory module, determine the cross entropy in epicycle training process, so Model parameter afterwards based on the prediction model in cross entropy adjustment epicycle training process.

Step 604 is trained by more wheels to prediction model, obtains the second branch prediction model.

In a kind of possible design, robot of accompanying and attending to can also include voice synthetic module, and voice synthetic module can be with For extracting the audio frequency characteristics of template voice, and when the voice for receiving processor sending plays instruction, the mould based on extraction The audio frequency characteristics of plate voice issue voice.

Wherein, template voice can be through what is manually inputted, be also possible to connect by the external of robot of accompanying and attending to Mouth is imported from external equipment.

In a kind of possible design, robot of accompanying and attending to can also include apart from detection module, mobile module；Distance detection Module can be used for detecting accompany and attend to the distance between robot and user, and send mobile module for distance, and mobile module is used In when the distance for detecting that detection module is sent is greater than pre-determined distance, control accompany and attend to the position where robot towards user into Row movement.

Wherein, mobile module detection accompany and attend to robot between user at a distance from when, can first pass through to be mounted on and accompany and attend to Video camera shooting in robot includes the image of user, is then based on image segmentation and determines the position of user in the picture, Again based on the infrared facility accompanied and attended in robot is mounted on, is detected and accompanied and attended between robot and user by way of infrared distance measurement Distance.

In a kind of possible design, robot of accompanying and attending to can also include cleaning module, clean module and receive for working as When the cleaning instruction that processor is sent, the sweeping robot for the robot connection that controls and accompany and attend to is cleaned.

Wherein, sweeping robot can be connect by way of embedded with robot of accompanying and attending to.It accompanies and attends to robot and scavenging machine It can be both provided with infrared positioning apparatus in the main body of device people, after sweeping robot is completed to clean, filled based on infrared positioning It sets and establishes connection with robot of accompanying and attending to again.

In a kind of possible design, robot of accompanying and attending to can also be attached with external equipment, and pass through external device Heart rate, the blood pressure etc. of user are detected, when detecting the appearance exception such as heart rate, blood pressure, by alarm module to preparatory The equipment of binding sends warning message.

Based on identical design, accompany and attend to robot control method present invention also provides one kind, it is shown in Figure 7, for this Apply for a kind of flow diagram for robot control method of accompanying and attending to provided by embodiment, including the following steps:

Step 701, the biological information for receiving user.

Step 702 is based on biological information, predicts the mood of user.

The mood predicted is input in preparatory trained first branch prediction model, determining and biology by step 703 The control instruction that characteristic information matches.

Step 704, control robot of accompanying and attending to make operation corresponding with control instruction.

It accompanies and attends to robot controller present invention also provides one kind, it is shown in Figure 8, it is provided by the embodiment of the present application A kind of robot controller of accompanying and attending to configuration diagram, including receiving module 801, prediction module 802, determining module 803, Control module 804, specific:

Receiving module 801, for receiving the biological information of user；

Prediction module 802 predicts the mood of the user for being based on the biological information；

Determining module 803, for the mood predicted to be input in preparatory trained first branch prediction model, really The fixed control instruction to match with the biological information；

Control module 804 makes operation corresponding with the control instruction for controlling the robot of accompanying and attending to.

Based on same technical concept, the embodiment of the present application also provides a kind of electronic equipment.It is this Shen referring to shown in Fig. 9 Please the structural schematic diagram of electronic equipment 900 that provides of embodiment, including processor 901, memory 902 and bus 903.Wherein, Memory 902 is executed instruction for storing, including memory 9021 and external memory 9022；Here memory 9021 is also referred to as memory Reservoir, for temporarily storing the operational data in processor 901, and the data exchanged with external memories 9022 such as hard disks, Processor 901 carries out data exchange by memory 9021 and external memory 9022, when electronic equipment 900 is run, processor It is communicated between 901 and memory 902 by bus 903, so that processor 901 is being executed to give an order:

Receive the biological information of user；

Based on the biological information, the mood of the user is predicted；

The embodiment of the present application also provides a kind of computer readable storage medium, is stored on the computer readable storage medium Computer program, the computer program execute any of the above-described robot as described in the examples of accompanying and attending to and control when being run by processor The step of method.

Specifically, which can be general storage medium, such as mobile disk, hard disk, on the storage medium Computer program when being run, be able to carry out it is above-mentioned accompany and attend to robot control method the step of, needed to improve and improve user Seek the precision of branch prediction.

Carry out accompanying and attending to provided by the embodiment of the present application the computer program product of robot control method, including stores The computer readable storage medium of the executable non-volatile program code of processor, the instruction that said program code includes are available In executing previous methods method as described in the examples, specific implementation can be found in embodiment of the method, and details are not described herein.

It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description, The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.

In several embodiments provided herein, it should be understood that disclosed systems, devices and methods, it can be with It realizes by another way.The apparatus embodiments described above are merely exemplary, for example, the division of the unit, Only a kind of logical function partition, there may be another division manner in actual implementation, in another example, multiple units or components can To combine or be desirably integrated into another system, or some features can be ignored or not executed.Another point, it is shown or beg for The mutual coupling, direct-coupling or communication connection of opinion can be through some communication interfaces, device or unit it is indirect Coupling or communication connection can be electrical property, mechanical or other forms.

The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme 's.

It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit It is that each unit physically exists alone, can also be integrated in one unit with two or more units.

It, can be with if the function is realized in the form of SFU software functional unit and when sold or used as an independent product It is stored in the executable non-volatile computer-readable storage medium of a processor.Based on this understanding, the application Technical solution substantially the part of the part that contributes to existing technology or the technical solution can be with software in other words The form of product embodies, which is stored in a storage medium, including some instructions use so that One computer equipment (can be personal computer, server or the network equipment etc.) executes each embodiment institute of the application State all or part of the steps of method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. is various to deposit Store up the medium of program code.

Finally, it should be noted that embodiment described above, the only specific embodiment of the application, to illustrate the application Technical solution, rather than its limitations, the protection scope of the application is not limited thereto, although with reference to the foregoing embodiments to this Shen It please be described in detail, those skilled in the art should understand that: anyone skilled in the art Within the technical scope of the present application, it can still modify to technical solution documented by previous embodiment or can be light It is readily conceivable that variation or equivalent replacement of some of the technical features；And these modifications, variation or replacement, do not make The essence of corresponding technical solution is detached from the spirit and scope of the embodiment of the present application technical solution, should all cover the protection in the application Within the scope of.Therefore, the protection scope of the application shall be subject to the protection scope of the claim.

Claims

The robot 1. one kind is accompanied and attended to, which is characterized in that the robot of accompanying and attending to includes physical characteristics collecting module, processor；

The physical characteristics collecting module, for every preset duration, acquiring the biological information of user, and by the biology Characteristic information is transmitted to the processor；

The processor, for receiving the biological information of the physical characteristics collecting module transmission, and based on described Biological information predicts the mood of the user；And the mood predicted is input to trained first instruction in advance in advance It surveys in model, the determining control instruction to match with the biological information, and robot of accompanying and attending to described in control makes and institute State the corresponding operation of control instruction.
2. robot according to claim 1 of accompanying and attending to, the physical characteristics collecting module comprise the following modules at least It is a kind of:

Sound acquisition module, image capture module；

Include the case where sound acquisition module for the physical characteristics collecting module, the biological information includes sound letter Breath；

Include the case where that image capture module, the biological information include face figure for the physical characteristics collecting module As information.
3. robot according to claim 2 of accompanying and attending to, the processor are being based on the biological information, are predicting institute When stating the mood of user, it is specifically used for:

The phonetic feature in the acoustic information is extracted, the phonetic feature includes short-time energy, short-time zero-crossing rate, fundamental tone frequency Rate, formant feature, word speed, mel cepstrum coefficients；

The phonetic feature is input to speech recognition submodel, the phonetic feature is obtained and belongs to any one default mood First score；

The acoustic information is input to semantics recognition submodel, extracts the semantic key words in the acoustic information；

Based on the semantic key words, determine that the acoustic information belongs to the second score of any one default mood；

The facial image information is input in face recognition submodel, determines that the facial image information belongs to any one The third score of default mood；

First score, second score, the third score are weighted summation according to default weight, based on summation Score afterwards determines the mood of the user.
4. robot according to claim 1 of accompanying and attending to, which is characterized in that the robot of accompanying and attending to further include: receiving module, And memory module；

The receiving module is transmitted to the processor for receiving the instruction of user, and by described instruction；

The processor, is also used to: when receiving the described instruction of the receiving module transmission, controlling the biological characteristic and adopts Collection module acquires the biological information of the user；The behavioural information of the user is determined based on the biological information, And by described instruction, the time of reception described instruction and the behavioural information, it is transmitted to the memory module；The user Behavioural information be used to indicate the active state of the user；

The memory module, for storing described instruction, the time for receiving described instruction and the behavioural information.
5. robot according to claim 1 of accompanying and attending to, which is characterized in that the processor is also used to by the following method Training the first branch prediction model:

At least one history emotional prediction of the user is obtained as a result, and each associated finger of history emotional prediction result It enables；

The history emotional prediction result is input in branch prediction model to be trained, obtains the history emotional prediction knot The corresponding prediction instruction of fruit；

Based on prediction instruction and instruction associated with the history emotional prediction result, to the instruction to be trained Prediction model carries out epicycle training；

By more wheels training to described instruction prediction model, the first branch prediction model is obtained.
6. robot according to claim 4 of accompanying and attending to, which is characterized in that the processor, receiving, the biology is special After reference breath, it is also used to:

Based on the biological information, the behavioural information of the user is determined；

By the behavioural information of the user and time of the biological information is received, is input to trained in advance the In two branch prediction models, the determining control instruction to match with the biological information, and robot of accompanying and attending to described in control Make operation corresponding with the control instruction.
7. robot according to claim 6 of accompanying and attending to, which is characterized in that the processor trains institute in accordance with the following methods State the second branch prediction model:

The instruction of memory module storage is obtained, the time of described instruction is received and described specifies corresponding behavioural information；

By the time for receiving described instruction and described corresponding behavioural information is specified to be input to prediction model to be trained In, output obtains prediction instruction；

It is instructed based on the instruction obtained from the memory module and the prediction, epicycle training is carried out to the prediction model；

By more wheels training to the prediction model, the second branch prediction model is obtained.
8. robot according to claim 6 of accompanying and attending to, which is characterized in that the robot of accompanying and attending to further include: alarm module；

The alarm module, for by the behavioural information and being stored in abnormal behaviour information in the memory module and carrying out pair Than sending warning message to the equipment bound in advance when comparing successfully.
The robot control method 9. one kind is accompanied and attended to characterized by comprising

Receive the biological information of user；

Based on the biological information, the mood of the user is predicted；

The mood predicted is input in preparatory trained first branch prediction model, the determining and biological information The control instruction to match；

Robot of accompanying and attending to described in control makes operation corresponding with the control instruction.
The robot controller 10. one kind is accompanied and attended to characterized by comprising

Receiving module, for receiving the biological information of user；

Prediction module predicts the mood of the user for being based on the biological information；

Determining module, for the mood predicted to be input in preparatory trained first branch prediction model, determining and institute State the control instruction that biological information matches；

Control module makes operation corresponding with the control instruction for controlling the robot of accompanying and attending to.