Pet nursing method based on artificial intelligence and intelligent pet nursing equipment thereof

Abstract

The invention discloses a pet nursing method based on artificial intelligence and intelligent pet nursing equipment thereof, a pet AI information interaction model is applied to the intelligent pet nursing equipment, and the intelligent pet nursing equipment comprises a processor, a storage module, a communication module, a detection module, an interaction module, a massage module and a pet cleaning module. Pet information is obtained through the detection module and sent to the pet AI information interaction model, a pet is nursed according to the pet information, and the pet AI information interaction model obtains a specific pet nursing method according to the pet type, the pet body type, the pet age, the pet hair information, the pet emotional state and the pet real-time voice. The specific pet nursing method comprises the steps that the processor controls the massage module to massage the pet, and the processor controls the interaction module to simulate specific pet sound output feedback statements to communicate with the pet according to specific pet types, so that automation and intellectualization of pet nursing are achieved.

Description

Technical Field

[0001] The present invention relates to the field of pet care, and in particular to an artificial intelligence-based pet care method and intelligent pet care equipment. Background Art

[0002] Pets are human beings' most loyal friends. They will accompany you throughout your life and protect you until death. Pets can enrich human life. Veterinary experts and biologists have found through research that contact with pets can change mood, relieve symptoms and pain, and improve people's physical and mental functions. The 1993 Harvard University "Health Survey Report" shows that people who keep pets usually have lower blood pressure, more stable heart rate, and a calmer mentality.

[0003] Pet health care workers are professionals engaged in pet breeding, management, grooming and health care. They are responsible for pet breeding and disease prevention, daily health care and management of pets, scientific grooming of pets, on-site rescue, sickness care, postoperative care and rehabilitation care for injured and sick pets.

[0004] It is costly to have someone take care of your pet. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides an artificial intelligence-based pet care method and intelligent pet care device, which realize the automation and intelligence of pet care.

[0006] In order to achieve the above-mentioned purpose, the present invention provides a pet care method based on artificial intelligence, including a pet AI information interaction model, which is applied to an intelligent pet care device. The intelligent pet care device includes a processor, a storage module, a communication module, a detection module, an interaction module, a massage module and a pet cleaning module. The pet information is obtained through the detection module and sent to the pet AI information interaction model. The pet information includes pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The pet AI information interaction model obtains a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The specific pet care method includes the processor controlling the massage module to massage the pet and the processor controlling the interaction module to simulate the sound of a specific type of pet according to the specific pet type to output feedback statements to communicate with the pet. The pet AI information interaction model includes a pet AI large model or an AI search engine.

[0007] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention, the pet AI big model generates a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice analysis, and the pet AI big model obtains the specific pet care method including two steps: building a pet AI big model and calling the pet AI big model, and the building of the pet AI big model includes six steps:

[0008] Step 21101, collecting pet care data: collecting all information about pet care data presented through voice, video, text, picture or 3D model;

[0009] Step 21102, pre-processing pet care data: Process all collected information about pet care data to ensure the integrity and availability of the information, including converting information in different formats into text, and editing the text information in a certain format to facilitate subsequent training of the AI ​​large model;

[0010] Step 21103, select the AI ​​big model applicable to pets: select domestic and foreign third-party AI big models, and measure them by accuracy, response speed, and diversity indicators;

[0011] Step 21104, training the pet AI big model: organize the pet care data set through step 21102, and then fine-tune the pet care data set using the third-party AI big model. After the training is completed, a pet AI big model with all relevant data for pet care is generated;

[0012] Step 21105, verify and test the pet AI big model: perform effect detection and evaluation of the pet AI big model generated in step 21104 on a specific task. If the evaluation effect fails, continue to repeat steps 21101, 21102, 21103, and 21104, retrain until the effect evaluation passes, generate the pet AI big model and store it in the cloud platform or storage module;

[0013] Step 21106, deploy and maintain the pet AI big model: deploy the newly generated pet AI big model to the cloud platform or storage module, and perform continuous maintenance and updates, and regularly update data to ensure the timeliness and accuracy of the data; the pet information input by the user in voice, video, text, picture or 3D model format is called by the pet AI big model to generate specific care methods;

[0014] The calling of the pet AI big model specifically includes the following steps:

[0015] Step 21201, assemble the query statement;

[0016] Step 21202: The AI ​​big model performs reasoning operations: The query statement is passed to the AI ​​big model, and the AI ​​big model performs reasoning operations, which includes the following processes:

[0017] 1) Understanding input: Distributed semantic parsing first receives a text sequence and converts it into a word vector. This process is based on the distributed semantics assumption that the meaning of a word is determined by its use in context.

[0018] 2) Parameter association: context focus linkage, input these word vectors into the Transformer Encoder to generate context representation;

[0019] 3) Generate answers: Generative probability modeling. The model initializes the decoder part of the Transformer and inputs the encoder output and the current output sequence into the decoder. The decoder generates the probability distribution of the next word and selects the word with the largest probability or other set probability distribution as the output. This word will be added to the output sequence.

[0020] 4) Select the most appropriate answer: Dynamic word string evolution, repeat the above steps, adding new words to the output sequence each time until a complete output sequence is generated;

[0021] Step 21203, AI big model returns the result: the specific care method is sent to the interactive module or the user mobile terminal for confirmation after the pet AI big model reasoning operation is completed; after the pet AI big model reasoning operation is completed, relevant information about the specific care method is returned, including content provided to the user in text, picture, audio, video or 3D model format.

[0022] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention calls the AI ​​search engine to obtain a specific pet care method, including four steps:

[0023] Step 1: Apply for the API interface of the AI ​​search engine company;

[0024] Step 2: Obtain the pet AI big model through the detection module according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice, and encapsulate it into the parameters required by the API;

[0025] Step 3: Call the API interface and encapsulate the parameters required by the API to call the API interface;

[0026] Step 4: Process the API call result and extract the specific pet care method from the data returned by the API.

[0027] As an improvement of the artificial intelligence-based pet care method of the present invention, the detection module of the artificial intelligence-based pet care method of the present invention includes a visual recognition module, and the visual recognition module obtains relevant depth information, RGB color information, pet body shape and hair thickness of the pet. The processor performs three-dimensional modeling of the pet according to the relevant depth information. The processor identifies the pre-working position according to the three-dimensional modeling, the pet's body shape and hair thickness, and analyzes and calculates to generate a massage trajectory. The processor controls the massage module to adjust to a pre-positioning position according to a specific pet care method and controls the massage module to work according to the massage trajectory, wherein the pre-positioning position is the initial position of the massage module relative to the pet when it is working.

[0028] As an improvement of the artificial intelligence-based pet care method of the present invention, the visual recognition module of the artificial intelligence-based pet care method of the present invention captures the pet's position and posture in real time to analyze and calculate and generate a massage trajectory, and the processor controls the massage module in real time to work according to the massage trajectory.

[0029] As an improvement of the artificial intelligence-based pet care method of the present invention, the processor of the artificial intelligence-based pet care method of the present invention controls the interactive module to simulate the sound of a specific type of pet to output feedback statements to guide the pet to the pet care area according to the type of pet, and the processor controls the intelligent pet care device to care for the pet according to the specific pet care method, wherein the interactive module simulates the sound of a specific type of pet to output feedback statements to guide the pet, including a pet AI information interaction model that determines the characteristics and habits of a specific type of pet according to the type of pet, and generates feedback statements to guide the pet according to the characteristics and habits of the specific type of pet, and sends the feedback statements to guide the pet to the interactive module, and the processor controls the interactive module to simulate the sound of the pet to output feedback statements to guide the pet.

[0030] As an improvement of the artificial intelligence-based pet care method of the present invention, the specific pet care method also includes a cleaning method, the pet AI information interaction model analyzes the pet cleaning status according to the pet hair information and generates a specific pet cleaning method according to the pet type, pet body size and pet age, the specific pet cleaning method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet to guide the pet to walk to the pet care area according to the pet type, and the processor controls the pet cleaning module to execute the specific pet cleaning method to clean the pet.

[0031] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes a pet feeding module, and the specific pet care method also includes a pet feeding method. The pet AI information interaction model generates a specific pet feeding method according to the pet type, pet body size, pet age, pet emotional state and pet real-time voice analysis. The specific pet feeding method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet to guide the pet to walk to the pet care area according to the pet type, and the processor controls the pet feeding module to feed the pet in the pet care area according to the pet feeding method.

[0032] As an improvement of the artificial intelligence-based pet care method of the present invention, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes a defecation guidance method, and the pet AI information interaction model generates a specific defecation guidance method according to the pet type, the pet's emotional state and the pet's real-time voice analysis. The specific defecation guidance method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet according to the pet type to guide the pet's feedback statement to guide the pet to a specific defecation area for defecation.

[0033] As an improvement of the artificial intelligence-based pet care method of the present invention, the processor of the artificial intelligence-based pet care method of the present invention controls the interaction module to simulate the sound of a specific type of pet to output feedback sentences to communicate with the pet, the pet AI information interaction model analyzes the semantics of the pet's real-time voice according to the pet's real-time voice and generates feedback sentences, and sends the feedback sentences to the interaction module, and the processor controls the interaction module to simulate the sound of a specific type of pet to output feedback sentences according to the pet type to have a real-time conversation with the specific type of pet.

[0034] As an improvement of the artificial intelligence-based pet care method of the present invention, the feedback sentences of the artificial intelligence-based pet care method of the present invention include chat language, soothing language and command language.

[0035] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes a mobile device, the processor identifies the position of the pet through the detection module and feeds back to the processor, the processor controls the mobile device to move toward the pet according to the position of the pet, and controls the intelligent pet care device to care for the pet according to a specific pet care method.

[0036] As an improvement of the artificial intelligence-based pet care method of the present invention, the processor of the artificial intelligence-based pet care method of the present invention controls the pet cleaning module to clean the pet according to the cleaning method, including cleaning, combing, drying, and maintaining the pet's hair and adsorbing the pet's fallen hair.

[0037] As an improvement of the artificial intelligence-based pet care method of the present invention, the pet cleaning module of the artificial intelligence-based pet care method of the present invention includes an adsorption module, a combing module, a cleaning module and a drying module, and the cleaning method includes:

[0038] The processor controls the adsorption module to adsorb pet hair through a mechanical arm to adsorb pet hair and dander dropped by the pet;

[0039] The processor controls the combing module to comb the pet's hair through the robotic arm;

[0040] The processor controls the cleaning module to clean the pet through the robotic arm;

[0041] The processor controls the drying module to dry the pet's hair through the mechanical arm;

[0042] The processor controls the combing module to comb the pet's hair through the robotic arm.

[0043] As an improvement of the artificial intelligence-based pet care method of the present invention, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes motion soothing, music soothing and image soothing.

[0044] As an improvement of the artificial intelligence-based pet care method of the present invention, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes exchanging information with the pet in voice, video, picture or 3D model format through the interaction module.

[0045] As an improvement of the artificial intelligence-based pet care method of the present invention, the user of the artificial intelligence-based pet care method of the present invention inputs care task information through the interaction module, and sends the care task information to the pet AI information interaction model, and the pet AI information interaction model generates a specific pet care method based on the pet care task information analysis, wherein the interaction module also includes a touch screen, a camera, a microphone and a speaker, and the user can interact with information through the interaction module in voice, video, text, picture, code scanning recognition or 3D model format.

[0046] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes a floor cleaning module, the detection module also includes a visual module, the visual module monitors the sanitary conditions of the pet activity area, and the processor controls the floor cleaning module to clean the pet activity area.

[0047] As an improvement of the artificial intelligence-based pet care method of the present invention, the detection module of the artificial intelligence-based pet care method of the present invention also includes a force sensor. When executing a specific pet care method, the force sensor detects that the force of the massage module or the pet cleaning module is greater than a preset force and feeds back to the processor. The processor controls the massage module or the pet cleaning module to stop working.

[0048] As an improvement of the artificial intelligence-based pet care method of the present invention, the detection module of the artificial intelligence-based pet care method of the present invention also includes a safety module. When executing a specific pet care method, the safety module detects that the pet is unwell and feeds back to the processor, and the processor controls the massage module or the pet cleaning module to stop working.

[0049] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention is also provided with an induction module, which includes a heating device, an air supply device and a bait placement box. The processor controls the heating device to heat the bait placement box with bait and controls the air supply device to supply air toward the bait according to the pet information, so as to lure the pet to the pet care area. The processor controls the intelligent pet care equipment to care for the pet according to the specific pet care method.

[0050] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes an AI virtual pet image system, the detection module obtains a specific type of pet image, a specific type of pet action and a specific type of pet sound, and sends the specific type of pet image, specific type of pet action and specific type of pet sound to the AI ​​virtual pet image system, the detection module obtains the pet emotional state or the pet real-time voice and sends it to the pet AI information interaction model, the pet AI information interaction model generates a feedback statement based on the pet emotional state or the pet real-time voice analysis and sends it to the AI ​​virtual pet image system, the AI ​​virtual pet image system can analyze and train the specific pet image, specific type of pet action, specific pet sound and feedback statement to output a specific virtual pet, and send it to the interaction module, wherein the specific virtual pet includes a feedback statement output by a specific virtual specific type pet image simulating a specific virtual human action with a specific type of pet sound, and the interaction module outputs the specific virtual pet to communicate with the pet by voice.

[0051] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes an AI virtual human system, the detection module obtains the pet owner's image, pet owner's actions and pet owner's voice, and sends the pet owner's image, pet owner's actions and pet owner's voice to the AI ​​virtual human system, the detection module obtains the pet's emotional state and / or the pet's real-time voice and sends it to the pet AI information interaction model, the pet AI information interaction model generates feedback sentences based on the pet's emotional state and / or the pet's real-time voice analysis and sends it to the AI ​​virtual human system, the AI ​​virtual human system can analyze and train the pet owner's image, pet owner's actions, pet owner's voice and feedback sentences to output a specific virtual human, and send it to the interaction module, wherein the specific virtual human includes the feedback sentence output by the pet owner's image in the pet owner's voice, and the interaction module outputs the specific virtual human to have a conversation with the pet.

[0052] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes an AI agent system, the AI ​​agent system is provided with a memory module, a planning module, an execution module and a tool module, and the pet type, pet body shape, pet age, pet hair information, pet emotional state, interaction record and feeding record data are pushed to the AI ​​agent system, and the AI ​​agent system outputs a pet health management plan, wherein the AI ​​agent system outputs the pet health management plan including the following steps:

[0053] S1, the AI ​​agent system receives pet type, pet body size, pet age, pet hair information, and pet emotional state, and analyzes the pet health status according to the pet type, pet body size, pet age, pet hair information, and pet emotional state;

[0054] S2, the planning module generates a pet health management task according to the pet care knowledge base, feeding record data and interaction record of the memory module, and sends the pet health management task to the planning module, wherein the pet health management task includes pet diet management and interaction management within a specific period, and the pet health management task is decomposed into a plurality of subtasks by reasoning, wherein each of the subtasks includes feeding the pet according to the pet feeding method at any time within a period and interacting with the pet according to a specific pet care method at any time within a period, wherein the interaction includes communicating, massaging or comforting the pet according to the specific pet care method;

[0055] S3, the AI ​​agent system formulates a corresponding subtask plan according to each of the subtasks, and sends it to the interaction module for user confirmation or modification, wherein the user modifies the subtask plan and feeds back to the AI ​​agent system, and the AI ​​agent system returns to step S2 to re-plan multiple subtasks of the pet health management task until a health management pet plan is generated, wherein each of the subtask plans includes an execution date and execution time for each feeding of the pet according to the pet feeding method and an execution date and execution time for each interaction with the pet according to a specific pet care method;

[0056] The AI ​​agent system executes the health management pet plan including the following steps:

[0057] The AI ​​agent system plans the subtasks and calls the application tool through the tool module to execute the corresponding subtasks, wherein the application tool includes a calendar tool and a time tool. The tool module calls the calendar tool and the time tool to set the execution date and execution time of the interaction. The pet AI information interaction model sends the specific pet care method to the processor according to the execution date and execution time of the interaction, and the processor controls the smart pet care device to interact with the pet; the tool module calls the calendar tool and the time tool to set the execution date and execution time of feeding. The pet AI information interaction model sends the pet feeding method to the processor according to the execution date and execution time of feeding, and the processor controls the feeding module to feed the pet.

[0058] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention is further provided with a Vincent video model, and the Vincent video model generates a specific care video according to the specific pet care method, comprising the following steps:

[0059] The specific pet care method is sent to the Vincent video model, and the Vincent video model generates a specific care video according to the specific pet care method. The pet AI information interaction model converts the specific pet care method into text content, and calls the Vincent video model through an API, and provides the text content to the Vincent video model. The Vincent video model analyzes and generates the specific care video and sends the specific care video to the interaction module. When the processor executes the specific pet care method, the interaction module outputs the specific care video.

[0060] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes a Vincent music model and video synthesis software, and generating a specific care short film includes the following steps:

[0061] The pet AI information interaction model converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model, and the Vincent music model generates a specific care song according to the care lyrics;

[0062] The specific care video and the specific care song are synthesized into a specific care short film through a video synthesis system and sent to the interactive module. When the processor executes the specific pet care method, the interactive module outputs the specific care short film.

[0063] As an improvement of the artificial intelligence-based pet care method of the present invention, the artificial intelligence-based pet care method of the present invention also includes a Vincent music model, and the Vincent music model generates a specific care song according to the specific pet care method, including the following steps:

[0064] The pet AI information interaction model converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model. The Vincent music model generates a specific care song based on the care lyrics and sends it to the interaction module. When the processor executes the specific pet care method, the interaction module outputs the specific care song.

[0065] As an improvement of the artificial intelligence-based pet care method of the present invention, the detection module of the artificial intelligence-based pet care method of the present invention includes a camera, which records the daily activities of the pet and stores them in the storage module or the cloud platform. The user's mobile terminal obtains the daily video of the pet through the communication module, and the user shares it through the user's mobile terminal.

[0066] The present invention provides an intelligent pet care device, characterized in that it includes the processor, storage module, communication module, detection module, interaction module, massage module and pet cleaning module, wherein the processor, storage module, communication module, detection module, interaction module, massage module and pet cleaning module are all electrically connected to the processor, and also includes a pet AI information interaction model arranged on the storage module or the cloud platform, wherein the pet AI information interaction model includes a pet AI big model and an AI search engine, the intelligent pet care device exchanges information with the cloud platform or the interaction module through the communication module, the detection module is used to obtain pet information and send it to the pet AI information interaction model, and the processor executes the specific pet care method described in any one of claims 1 to 10.

[0067] As an improvement of the smart pet care device of the present invention, the smart pet care device of the present invention also includes a robotic arm, a base and a mobile device, the base is arranged on the mobile device, the processor, the storage module and the communication module are arranged in the base, and the robotic arm is connected to the massage module and the pet cleaning module by grasping.

[0068] As an improvement of the intelligent pet care device of the present invention, it also includes a base and a mobile device, the base is arranged on the mobile device, the processor, the storage module and the communication module are arranged in the base, a first robotic arm is arranged corresponding to the massage module, the first robotic arm is arranged on the base, and the massage module is arranged at the end of the first robotic arm, a second robotic arm is arranged corresponding to the pet cleaning module, the second robotic arm is arranged on the base, and the pet cleaning module is arranged at the end of the second robotic arm.

[0069] As an improvement of the intelligent pet care device of the present invention, the massage module of the intelligent pet care device of the present invention further includes a vibration massage device, a knocking massage device and a kneading massage device.

[0070] As an improvement of the smart pet care device of the present invention, the pet cleaning module of the smart pet care device of the present invention includes a hair combing module, an adsorption module and a drying module.

[0071] As an improvement of the smart pet care device of the present invention, the smart pet care device of the present invention further includes a floor cleaning module, and the floor cleaning module further includes an adsorption module, a sweeping device and a garbage recycling device.

[0072] The beneficial effects of the artificial intelligence-based pet care method and intelligent pet care device of the present invention are: pet information is acquired through the detection module and the pet information is sent to the pet AI information interaction model, the pet is cared for according to the pet information, the pet AI information interaction model generates a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice analysis, the specific pet care method includes the processor controlling the massage module to massage the pet, and also includes the processor controlling the interaction module to simulate the sound of the specific pet type to output feedback statements to communicate with the pet according to the specific pet type, thereby realizing the automation and intelligence of pet care. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a work flow chart of a preferred embodiment of the pet care method based on artificial intelligence and its intelligent pet care device of the present invention;

[0074] Figure 2It is a work flow chart of another embodiment 1 of the pet care method based on artificial intelligence and the intelligent pet care device thereof of the present invention;

[0075] Figure 3 This is a workflow diagram of another embodiment 2 of the artificial intelligence-based pet care method and intelligent pet care device of the present invention. DETAILED DESCRIPTION

[0076] Below, combined with the attached Figure 1-3 As well as specific implementation methods and other embodiments, the present invention is further described. It should be noted that, under the premise of no conflict, the various technical features described below can be arbitrarily combined to form new embodiments.

[0077] In a preferred embodiment, refer to the attached Figure 1The present invention provides an artificial intelligence-based pet care method, including a pet AI information interaction model 20, which is applied to an intelligent pet care device. The intelligent pet care device includes a processor 5, a storage module 6, a communication module 4, a detection module 12, an interaction module 11, a massage module 13 and a pet cleaning module 14. The pet information is obtained through the detection module 12 and sent to the pet AI information interaction model 20. The pet information includes the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The pet AI information interaction model 20 obtains a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The specific pet care method includes the processor 5 controlling the massage module 13 to massage the pet and the processor 5 controlling the interaction module 11 to simulate the sound of the specific pet type to output feedback sentences to communicate with the pet according to the specific pet type. The pet AI information interaction model 20 includes a pet AI large model 21 or an AI search engine 27. For example, in one embodiment, the pet AI information interaction model 20 is a pet AI big model 21, and the detection module 12 is a camera. The camera is used to obtain the pet type as a Bichon Frise, and the Bichon Frise's body shape, Bichon Frise's age, Bichon Frise's hair information, Bichon Frise's emotional state and Bichon Frise's real-time voice are identified and sent to the pet AI big model. The pet AI big model determines that the Bichon Frise needs to be appeased by massage based on the Bichon Frise's emotional state and Bichon Frise's real-time voice. The pet AI big model generates a specific pet care method based on the Bichon Frise's body shape, Bichon Frise's age and Bichon Frise's hair information analysis to massage the Bichon Frise. In this embodiment, the pet AI big model 21 generates a specific pet care method based on the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice analysis. The pet AI big model 21 obtains the specific pet care method including two steps: constructing a pet AI big model 211 and calling a pet AI big model 212. The construction of the pet AI big model 211 includes six steps:

[0078] Step 21101, collecting pet care data: collecting all information about pet care data presented through voice, video, text, picture or 3D model;

[0079] Step 21102, pre-processing pet care data: Process all collected information about pet care data to ensure the integrity and availability of the information, including converting information in different formats into text, and editing the text information in a certain format to facilitate subsequent training of the AI ​​large model;

[0080] Step 21103, select the AI ​​big model applicable to pets: select domestic and foreign third-party AI big models, and measure them by accuracy, response speed, and diversity indicators;

[0081] Step 21104, training the pet AI big model: organize the pet care data set through step 21102, and then fine-tune the pet care data set using the third-party AI big model. After the training is completed, a pet AI big model with all relevant data for pet care is generated;

[0082] Step 21105, verify and test the pet AI big model: perform effect detection and evaluation of the pet AI big model generated in step 21104 on a specific task. If the evaluation effect fails, continue to repeat steps 21101, 21102, 21103, and 21104, retrain until the effect evaluation passes, generate the pet AI big model and store it in the cloud platform or storage module;

[0083] Step 21106, deploy and maintain the pet AI big model: deploy the newly generated pet AI big model to the cloud platform or storage module, and perform continuous maintenance and updates, and regularly update data to ensure the timeliness and accuracy of the data; the pet information input by the user in voice, video, text, picture or 3D model format is called by the pet AI big model to generate specific care methods;

[0084] The calling of the pet AI big model specifically includes the following steps:

[0085] Step 21201, assemble the query statement;

[0086] Step 21202: The AI ​​big model performs reasoning operations: The query statement is passed to the AI ​​big model, and the AI ​​big model performs reasoning operations, which includes the following processes:

[0087] 1) Understanding input: Distributed semantic parsing first receives a text sequence and converts it into a word vector. This process is based on the distributed semantics assumption that the meaning of a word is determined by its use in context.

[0088] 2) Parameter association: context focus linkage, input these word vectors into the Transformer Encoder to generate context representation;

[0089] 3) Generate answers: Generative probability modeling. The model initializes the decoder part of the Transformer and inputs the encoder output and the current output sequence into the decoder. The decoder generates the probability distribution of the next word and selects the word with the largest probability or other set probability distribution as the output. This word will be added to the output sequence.

[0090] 4) Select the most appropriate answer: Dynamic word string evolution, repeat the above steps, adding new words to the output sequence each time until a complete output sequence is generated;

[0091] Step 21203, AI big model returns the result: the specific care method is sent to the interactive module or the user mobile terminal for confirmation after the pet AI big model reasoning operation is completed; after the pet AI big model reasoning operation is completed, relevant information about the specific care method is returned, including content provided to the user in text, picture, audio, video or 3D model format.

[0092] In other embodiments, the pet AI information interaction model 20 is an AI search engine 27, and calling the AI ​​search engine 27 to obtain a specific pet care method includes four steps:

[0093] Step 1: Apply for the API interface of an AI search engine company; the API interface includes OpenAI's AI search engine, Google's AI search engine, Secret Tower's AI search engine, Perplexity's AI search engine, You's AI search engine, or Tiangong AI's search engine;

[0094] Step 2: Obtain the pet AI big model through the detection module according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice, and encapsulate it into the parameters required by the API;

[0095] Step 3: Call the API interface and encapsulate the parameters required by the API to call the API interface;

[0096] Step 4: Process the API call result and extract the specific pet care method from the data returned by the API.

[0097] In a preferred embodiment, the detection module 12 of the artificial intelligence-based pet care method of the present invention includes a visual recognition module, the visual recognition module obtains the relevant depth information, RGB color information, pet body shape and hair thickness of the pet 3, the processor 5 performs three-dimensional modeling on the pet according to the relevant depth information, the processor 5 identifies the pre-working position according to the three-dimensional modeling, the pet body shape and hair thickness, and analyzes and calculates to generate a massage trajectory, the processor 5 controls the massage module 13 to adjust to the pre-positioned position according to the specific pet care method and controls the massage module 13 to work according to the massage trajectory, wherein the pre-positioned position is the initial position of the massage module 13 relative to the pet 3 when working. In a specific embodiment, the relevant depth information of the Bichon Frise is obtained through a 3D camera, the processor 5 performs three-dimensional modeling on the Bichon Frise according to the relevant depth information to identify the head position of the Bichon Frise as the pre-positioned position, and calculates and generates a massage trajectory, and the massage trajectory is from the pet head area to the body area.

[0098] The visual recognition module captures the position and posture of the pet 3 in real time to analyze and calculate to generate a massage trajectory, and the processor 5 controls the massage module 13 to work according to the massage trajectory in real time. The processor controls the massage module 13 to work according to the massage trajectory in real time.

[0099] In a preferred embodiment, the processor 5 of the artificial intelligence-based pet care method of the present invention controls the interaction module 11 to simulate the sound of a specific type of pet to output feedback statements to guide the pet to the pet care area according to the type of pet, and the processor 5 controls the intelligent pet care device to care for the pet according to the specific pet care method, wherein the interaction module 11 simulates the sound of a specific type of pet to output feedback statements to guide the pet, including a pet AI information interaction model 20 that determines the characteristics and habits of a specific type of pet according to the type of pet, and generates feedback statements to guide the pet according to the characteristics and habits of the specific type of pet, and sends the feedback statements to the interaction module 11, and the processor 5 controls the interaction module 11 to simulate the sound of the pet to output feedback statements to guide the pet. For example, the pet AI information interaction model 20 is the pet AI large model 21, and the interaction module 11 includes a speech synthesis model. The pet AI large model 21 determines that the pet type is a Bichon Frise, and generates a feedback sentence calling the Bichon Frise based on the characteristics and habits of the Bichon Frise. The processor controls the speech synthesis model to simulate the Bichon Frise sound output to call the Bichon Frise based on the content of the feedback sentence, and guides the Bichon Frise to the care area. The processor 5 controls the intelligent pet care device to care for the Bichon Frise according to a specific pet care method.

[0100] In a preferred embodiment, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes a cleaning method, the pet AI information interaction model 20 analyzes the pet cleaning state according to the pet hair information and generates a specific pet cleaning method according to the pet type, pet body size and pet age, the specific pet cleaning method includes the processor controlling the interaction to simulate the specific type of pet sound output to guide the pet to walk to the pet care area according to the pet type, and the processor controlling the pet cleaning module to execute the specific pet cleaning method to clean the pet. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the pet AI large model 21 analyzes the Bichon Frise cleaning state according to the Bichon Frise hair information and generates a specific pet cleaning method according to the Bichon Frise body size and age, and the specific pet cleaning method includes using a shampoo designed specifically for Bichon Frise, avoiding the use of human shampoo, which may not be suitable for the skin of Bichon Frise; when bathing, start from the lower part of the body, gradually clean upwards, and finally wash the head; rinse thoroughly after bathing, dry with a towel, and then blow dry with a hair dryer.

[0101] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention also includes a pet feeding module 15, and the specific pet care method also includes a pet feeding method. The pet AI information interaction model 20 generates a specific pet feeding method based on the pet type, pet body size, pet age, pet emotional state and pet real-time voice analysis. The specific pet feeding method includes the processor 5 controlling the interaction module 11 to simulate the sound output of the specific type of pet to guide the pet to walk to the pet care area according to the pet type, and the processor 5 controls the pet feeding module 15 to feed the pet in the pet care area according to the pet feeding method. For example, the pet AI information interaction model 20 is a pet AI big model 21. The pet AI big model 21 analyzes the Bichon Frise's emotional state and real-time voice and finds that the Bichon Frise is hungry and needs to eat. The pet AI big model 21 generates a specific pet feeding method based on the Bichon Frise's hunger state information and the Bichon Frise's body shape and puppy. The specific pet feeding method includes feeding dog food with goat milk powder in a quantitative manner to avoid feeding too much at one time to avoid indigestion. The pet AI big model 21 determines that the pet type is a Bichon Frise based on the pet type, and generates a feedback statement calling the Bichon Frise based on the characteristics and habits of the Bichon Frise. The processor controls the speech synthesis model to simulate the Bichon Frise's voice to output the feedback statement calling the Bichon Frise based on the content of the feedback statement, and guides the Bichon Frise to the pet care area for feeding.

[0102] In a preferred embodiment, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes a defecation guidance method, and the pet AI information interaction model 20 generates a specific defecation guidance method according to the pet type, the pet emotional state and the pet real-time voice analysis, and the specific defecation guidance method includes the processor 5 controlling the interaction module 11 to simulate the voice of the specific type of pet according to the pet type to output a feedback statement to guide the pet to walk to a specific defecation area for defecation. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the pet AI large model 21 analyzes the Bichon Frise's emotional state and the Bichon Frise's real-time voice to determine that the Bichon Frise needs to defecate. The pet AI large model 21 determines that the pet type is a Bichon Frise, and generates a feedback statement calling the Bichon Frise according to the characteristics and habits of the Bichon Frise. The processor controls the speech synthesis model to simulate the Bichon Frise voice output according to the content of the feedback statement to call the Bichon Frise, and guides the Bichon Frise to walk to a specific defecation area for defecation.

[0103] In a preferred embodiment, the processor 5 of the pet care method based on artificial intelligence of the present invention controls the interactive module 11 to simulate the voice of a specific type of pet to output feedback statements to communicate with the pet, the pet AI information interaction model 20 analyzes the semantics of the pet's real-time voice according to the pet's real-time voice and generates feedback statements, and sends the feedback statements to the interactive module 11, and the processor 5 controls the interactive module 11 to simulate the voice of a specific type of pet according to the pet's type to output feedback statements and conduct real-time dialogue with the specific type of pet. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the pet AI large model 21 analyzes its semantics and generates feedback statements based on the real-time voice of the Bichon Frise and based on the characteristics and habits of the Bichon Frise, and the processor controls the speech synthesis model to simulate the voice of the Bichon Frise to output feedback statements and conduct real-time dialogue with the Bichon Frise. Among them, the feedback statements include chat language, soothing language and command language.

[0104] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention also includes a mobile device 16, and the processor 5 identifies the position of the pet through the detection module 12 and feeds back to the processor 5. The processor 5 controls the mobile device 16 to move toward the pet according to the position of the pet, and controls the intelligent pet care device to care for the pet according to a specific pet care method. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the detection module 12 is a camera. The camera detects that the Bichon Frise has hair loss and feeds back to the processor and the pet AI large model 21. The pet AI large model 21 generates a method for absorbing Bichon Frise hair based on the pet's body shape, pet age and pet hair information analysis. The processor controls the mobile device 16 to move toward the pet and controls the adsorption module to absorb the pet's fallen hair and dandruff according to the Bichon Frise hair adsorption method.

[0105] In a preferred embodiment, the processor 5 of the artificial intelligence-based pet care method of the present invention controls the pet cleaning module 14 to clean the pet according to the cleaning method, including cleaning, combing, drying, and maintaining the pet's hair, and absorbing the pet's fallen hair. Specifically, the pet cleaning module 14 includes an absorption module, a combing module, a cleaning module, and a drying module, and the cleaning method includes:

[0106] The processor 5 controls the adsorption module to adsorb pet hair through the robot arm 8 to adsorb pet hair and dander dropped by the pet;

[0107] The processor 5 controls the combing module to comb the pet's hair through the robot arm 8;

[0108] The processor 5 controls the cleaning module to clean the pet through the robot arm 8;

[0109] The processor 5 controls the drying module to dry the pet's hair through the mechanical arm 8;

[0110] The processor 5 controls the combing module to comb the pet's hair through the robot arm 8.

[0111] In a preferred embodiment, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes motion soothing, music soothing and image soothing. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the detection module is a camera, through which the information of the Bichon Frise is obtained. The pet AI large model analyzes the Bichon Frise in a manic state based on the real-time video data of the Bichon Frise and determines that the Bichon Frise needs to be soothed by motion. The pet AI large model generates a method of soothing the Bichon Frise by motion based on the body shape, age and hair information of the Bichon Frise. The processor controls the intelligent pet care device to soothe the Bichon Frise according to the method of soothing the Bichon Frise by motion.

[0112] In a preferred embodiment, the specific pet care method of the artificial intelligence-based pet care method of the present invention also includes information interaction with the pet in voice, video, picture or 3D model format through the interaction module 11.

[0113] In a preferred embodiment, the user of the artificial intelligence-based pet care method of the present invention inputs care task information through the interactive module 11, and sends the care task information to the pet AI information interaction model 20, and the pet AI information interaction model 20 generates a specific pet care method based on the pet care task information analysis, wherein the interactive module 11 also includes a touch screen, a camera, a microphone and a speaker, and the user can interact with information in the form of voice, video, text, picture, code scanning recognition or 3D model through the interactive module 11. For example, the pet AI information interaction model 20 is a pet AI large model 21, and the user inputs "feeding a 10 kg adult Bichon Frise" through the touch screen, and the pet AI large model 21 "feeding a 10 kg adult Bichon Frise" generates a specific pet feeding method, and the specific pet feeding method includes selecting low-salt, high-quality dog ​​food, and appropriately adding meat, vegetables and other foods according to the preferences and health status of the Bichon Frise.

[0114] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention also includes a floor cleaning module 17, the detection module 12 also includes a visual module, the visual module monitors the sanitary conditions of the pet activity area, and the processor 5 controls the floor cleaning module 17 to clean the pet activity area.

[0115] In a preferred embodiment, the detection module 12 of the artificial intelligence-based pet care method of the present invention further includes a force sensor. When executing a specific pet care method, the force sensor detects that the force of the massage module 13 or the pet cleaning module 14 is greater than a preset force and feeds back to the processor 5, and the processor 5 controls the massage module 13 or the pet cleaning module 14 to stop working. For example, when cleaning and combing the hair of a Bichon Frise according to a specific pet care method, the preset force is in the range of 0.1 to 0.5 Newtons, and the cleaning module 14 is obstructed. When the combing force of the cleaning module 14 exceeds 0.5 Newtons, it still cannot comb according to the predetermined trajectory, and the pet cleaning module 14 stops working.

[0116] In a preferred embodiment, the detection module 12 of the artificial intelligence-based pet care method of the present invention further includes a safety module. When executing a specific pet care method, the safety module detects that the pet is uncomfortable and feeds back to the processor 5, and the processor 5 controls the massage module 13 or the pet cleaning module 14 to stop working. For example, when massaging a Bichon Frise according to a specific pet care method, when the safety module detects that the Bichon Frise makes an uncomfortable sound, the processor 5 controls the massage module 13 to stop working.

[0117] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention is also provided with an induction module 18, and the induction module 18 includes a heating device, an air supply device and a bait placement box. The processor 5 controls the heating device to heat the bait placement box containing bait and controls the air supply device to supply air toward the bait according to the pet information, so as to lure the pet to the pet care area. The processor 5 controls the intelligent pet care equipment to care for the pet according to the specific pet care method.

[0118] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention also includes an AI virtual animal image system 26, the detection module 12 obtains a specific type of pet image, a specific type of pet action and a specific type of pet sound, and sends the specific type of pet image, specific type of pet action and specific type of pet sound to the AI ​​virtual animal image system 26, the detection module 12 obtains the pet's emotional state or the pet's real-time voice and sends it to the pet AI information interaction model 20, the pet AI information interaction model 20 generates a feedback statement based on the pet's emotional state or the pet's real-time voice analysis and sends it to the AI ​​virtual animal image system 26, the AI ​​virtual animal image system 26 can analyze and train the specific pet image, specific type of pet action, specific pet sound and feedback statement to output a specific virtual pet, and send it to the interaction module 11, wherein the specific virtual pet includes a specific virtual specific type of pet image simulating a specific virtual human action and outputting a feedback statement using a specific type of pet sound, and the interaction module 11 outputs the specific virtual pet to communicate with the pet by voice.

[0119] In a preferred embodiment, the pet AI information interaction model 20 is a pet AI big model 21, and also includes an AI virtual human system 22. The detection module obtains the pet owner's image, pet owner's actions and pet owner's voice, and sends the pet owner's image, pet owner's actions and pet owner's voice to the AI ​​virtual human system 22. The detection module 12 obtains the pet's emotional state and / or the pet's real-time voice and sends it to the pet AI big model 21. The pet AI big model 21 generates feedback sentences based on the pet's emotional state and / or the pet's real-time voice analysis and sends it to the AI ​​virtual human system 22. The AI ​​virtual human system 22 can analyze and train the pet owner's image, pet owner's actions, pet owner's voice and feedback sentences to output a specific virtual human, and send it to the interaction module 11, wherein the specific virtual human includes the feedback sentence output by the pet owner's image in the pet owner's voice, and the interaction module 11 outputs the specific virtual human to have a conversation with the pet. For example, the detection module 12 is a camera, through which the information of the Bichon Frise is obtained. The pet AI large model analyzes the real-time video data of the Bichon Frise and finds that the Bichon Frise is in a relatively low mood and determines that the Bichon Frise needs the company of its owner. The pet AI large model 21 generates a feedback statement based on the emotional state analysis of the Bichon Frise and sends it to the AI ​​virtual human system 22. The AI ​​virtual human system 22 can analyze and train to output a specific virtual human based on the image, actions, voice and feedback statements of the pet owner. The interaction module 11 outputs the specific virtual human to communicate with the pet to comfort the Bichon Frise.

[0120] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention further includes an AI agent system 23, wherein the AI ​​agent system 23 is provided with a memory module, a planning module, an execution module, and a tool module, and the pet type, pet body shape, pet age, pet hair information, pet emotional state, interaction record, and feeding record data are pushed to the AI ​​agent system 23, and the AI ​​agent system 23 outputs a pet health management plan, wherein the AI ​​agent system 23 outputs a pet health management plan including the following steps:

[0121] S1, the AI ​​agent system 23 receives the pet type, pet body shape, pet age, pet hair information, and pet emotional state, and analyzes the pet's health status according to the pet type, pet body shape, pet age, pet hair information, and pet emotional state; in this embodiment, the AI ​​agent system itself may have the function of analyzing the pet's health status. In other embodiments, the AI ​​big model may also analyze the pet's health status according to the pet's information, such as Alibaba's M6, Zhiyuan Research Institute's Wudao 2.0, Tencent's Hunyuan, Baidu's Wenxin, Institute of Automation's Zidong Taichu, Zhipu Huazhang's ChatGLM3, Baichuan of Baichuan Intelligence, Shusheng of Shanghai Artificial Intelligence Experiment, iFlytek's Spark, SenseTime's InternLM, Chuntian Zhiyun (Tik Tok)'s Doubao, Alibaba Cloud's Tongyi Qianwen, and Huawei's Pangu, etc.

[0122] S2, the planning module generates a pet health management task according to the pet care knowledge base, feeding record data and interaction record of the memory module, and sends the pet health management task to the planning module, wherein the pet health management task includes pet diet management and interaction management within a specific period, and the pet health management task is decomposed into a plurality of subtasks by reasoning, wherein each of the subtasks includes feeding the pet according to the pet feeding method at any time within a period and interacting with the pet according to a specific pet care method at any time within a period, wherein the interaction includes communicating, massaging or comforting the pet according to the specific pet care method; in other embodiments, the planning module may also be trained by an AI large model, such as Alibaba's M6, Zhiyuan Research Institute's Wudao 2.0, Tencent's Hunyuan, Baidu's Wenxin, Zidong Taichu of the Institute of Automation, ChatGLM3 of Zhipu Huazhang, Baichuan of Baichuan Intelligence, Shusheng of Shanghai Artificial Intelligence Experiment, Spark of iFlytek, InternLM of SenseTime, Doubao of Chuntian Zhiyun (Tik Tok), Tongyi Qianwen of Alibaba Cloud, and Pangu of Huawei, etc.;

[0123] S3, the AI ​​agent system 23 formulates a corresponding subtask plan according to each of the subtasks, and sends it to the interaction module 11 for user confirmation or modification, wherein the user modifies the subtask plan and feeds back to the AIagent system 23, and the AI ​​agent system 23 returns to step S2 to re-plan multiple subtasks of the pet health management task until a health management pet plan is generated, each of the subtask plans including the execution date and execution time of each feeding of the pet according to the pet feeding method and the execution date and execution time of each interaction with the pet according to a specific pet care method;

[0124] The AI ​​agent system executes the health management pet plan including the following steps:

[0125] The AI ​​agent system 23 allocates the subtask plan and calls the application tool through the tool module to execute the corresponding subtask, wherein the application tool includes a calendar tool and a time tool, and the tool module calls the calendar tool and the time tool to set the execution date and execution time of the interaction, and the pet AI information interaction model 20 sends the specific pet care method to the processor 5 according to the execution date and execution time of the interaction, and the processor 5 controls the smart pet care device to interact with the pet; the tool module calls the calendar tool and the time tool to set the execution date and execution time of feeding, and the pet AI information interaction model 20 sends the pet feeding method to the processor 5 according to the execution date and execution time of feeding, and the processor 5 controls the feeding module to feed the pet.

[0126] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention is further provided with a Vincent video model 25, and the Vincent video model 25 generates a specific care video according to the specific pet care method, including the following steps:

[0127] The specific pet care method is sent to the Vincent video model 25, and the Vincent video model 25 generates a specific care video according to the specific pet care method. The pet AI information interaction model 20 converts the specific pet care method into text content, calls the Vincent video model 25 through the API, and provides the text content to the Vincent video model 25. The Vincent video model 25 analyzes and generates the specific care video and sends the specific care video to the interaction module 11. When the processor 5 executes the specific pet care method, the interaction module 11 outputs the specific care video. For example, the pet AI information interaction model 20 is a pet AI big model 21, and the detection module 12 is a camera. The camera is used to obtain the information of the Bichon Frise. The pet AI big model 21 analyzes the real-time video data of the Bichon Frise and determines that the Bichon Frise is in a manic state and needs to be soothed by action. The pet AI big model generates a method for soothing the Bichon Frise by action based on the body shape, age and hair information of the Bichon Frise, and converts the method for soothing the Bichon Frise by action into text content. The Vincent video model 25 is called through the API, and the text content is provided to the Vincent video model 25. The Vincent video model 25 analyzes the text content to generate the specific care video and sends it to the interaction module 11. The interaction module 11 displays the content of the specific care video, including the processor 5 controlling the intelligent pet care device to soothe the Bichon Frise according to the method for soothing the Bichon Frise by action. In this embodiment, the Vincent video model is Keling. In other embodiments, the Vincent video model can be sora, invideo AI, Outfit Anyone, Runway, Etna, Boximator, WALT, MagicVideo-V2, Stable Video Diffusion, Flash-AI Digital Human, KreadoAI, Star Group, Opera, Tongyi Dance King, Mo Xiaoxian, Wenxin Yige, Phenaki, Make-a-Video, Mobius, Yifang Miaochuang, and Ju Ri. AI, Yiying AI, D-Human digital human, Xinyi video large model, Doubao, Magic Words, Blue Label Clone, Liveportrait, Meitu AI digital human DreamAvatar, Pika or Gen-2.

[0128] It also includes Vincent Music Model 24 and video synthesis software, and the steps of generating a specific nursing video include:

[0129] The pet AI information interaction model 20 converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model 24, and the Vincent music model 24 generates a specific care song based on the care lyrics;

[0130] The specific care video and the specific care song are synthesized into a specific care short film through a video synthesis system and sent to the interactive module 11. When the processor 5 executes the specific pet care method, the interactive module 11 outputs the specific care short film.

[0131] In a preferred embodiment, the artificial intelligence-based pet care method of the present invention further includes a Vincent music model 24, and the Vincent music model 24 generates a specific care song according to the specific pet care method, including the following steps:

[0132] The pet AI information interaction model 20 converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model 24. The Vincent music model 24 generates a specific care song based on the care lyrics and sends it to the interaction module 11. When the processor 5 executes the specific pet care method, the interaction module 11 outputs the specific care song. For example, the pet AI information interaction model 20 is a pet AI model 21, the detection module 12 is a camera, and the camera is used to obtain the information of the Bichon Frise. The pet AI model 21 analyzes the Bichon Frise in a manic state based on the real-time video data of the Bichon Frise and determines that the Bichon Frise needs to be comforted by music. The pet AI model 21 generates lyrics to comfort the Bichon Frise based on the emotional state information of the Bichon Frise and sends them to the Vincent music model 24. The Vincent music model 24 generates specific soothing music based on the lyrics to comfort the Bichon Frise and sends it to the interaction module 11. The interaction module 11 outputs specific soothing music to comfort the Bichon Frise. In this embodiment, the Vincent music model 24 is Suno. In other embodiments, the Vincent music model 24 is Udio, Sonauto, Soundation, stable audio, Music AISandbox, NetEase Tianyin or Tiangong AI. In addition, the specific nursing video and the specific soothing music can be synthesized into a specific nursing short film through a video synthesis system and sent to the interactive module 11, and the interactive module 11 outputs the specific nursing short film to please the Bichon Frise.

[0133] In a preferred embodiment, the detection module 12 of the artificial intelligence-based pet care method of the present invention includes a camera, which records the daily activities of the pet and stores them in the storage module 6 or the cloud platform 3. The user mobile terminal 1 obtains the daily video of the pet 3 through the communication module 4, and the user shares it through the user mobile terminal 1, such as the user shares the daily video of the pet to the social platform through the user mobile terminal 1.

[0134] like Figure 1As shown, in the preferred embodiment, the pet AI information interaction model 20, the Vincent video model 25, the AI ​​virtual human system 22, the AI ​​virtual animal image system 26, the AI ​​agent system 23 and the Vincent music model 24 are all arranged in the storage module 6; in other embodiments, as Figure 2 As shown, the pet AI information interaction model 20, the Vincent video model 25, the AI ​​virtual human system 22, the AI ​​virtual animal image system 26, the AI ​​agent system 23 and the Vincent music model 24 are all set on the cloud platform 2.

[0135] The present invention provides an intelligent pet care device, characterized in that it includes a processor 5, a storage module 6, a communication module 4, a detection module 12, an interaction module 11, a massage module 13 and a pet cleaning module 14, wherein the processor 5, the storage module 6, the communication module 4, the detection module 12, the interaction module 11, the massage module 13 and the pet cleaning module 14 are all electrically connected to the processor 5, and also includes a pet AI information interaction model 20 arranged on the storage module 6 or the cloud platform 3, wherein the pet AI information interaction model 20 includes a pet AI large model 21 and an AI search engine 27, the intelligent pet care device performs information interaction with the cloud platform 3 or the interaction module 11 through the communication module 4, the detection module 12 is used to obtain pet information and send it to the pet AI information interaction model 20, and the processor 5 executes the specific pet care method described in any one of claims 1 to 10.

[0136] In a preferred embodiment, an intelligent pet care device of the present invention also includes a robotic arm 8, a base 7 and a mobile device 16, the base 7 is arranged on the mobile device 16, the processor 5, the storage module 6 and the communication module 4 are arranged in the base 7, and the robotic arm 8 is connected to the massage module 13 and the pet cleaning module 14 by grasping.

[0137] In another embodiment three, Figure 3 As shown, an intelligent pet care device of the present invention also includes a base 7 and a mobile device 16, wherein the base 7 is arranged on the mobile device 16, the processor 5, the storage module 6 and the communication module 4 are arranged in the base 7, a first robotic arm 31 is provided corresponding to the massage module 13, the first robotic arm 31 is provided on the base 7, and the massage module 13 is provided at the end of the first robotic arm 31, and a second robotic arm 32 is provided corresponding to the pet cleaning module 14, the second robotic arm 32 is provided on the base 7, and the pet cleaning module 14 is provided at the end of the second robotic arm 32.

[0138] In a preferred embodiment, the massage module 13 of the intelligent pet care device of the present invention further includes a vibration massage device, a percussion massage device and a kneading massage device.

[0139] In a preferred embodiment, the pet cleaning module 14 of the intelligent pet care device of the present invention includes a hair combing module, an adsorption module and a drying module.

[0140] In a preferred embodiment, the intelligent pet care device of the present invention further includes a floor cleaning module 17, and the floor cleaning module 17 further includes an adsorption module, a sweeping device and a garbage recycling device.

[0141] The beneficial effects of the artificial intelligence-based pet care method and intelligent pet care device of the present invention are: pet information is obtained through the detection module and the pet information is sent to the pet AI information interaction model, the pet is cared for according to the pet information, the pet AI information interaction model generates a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice analysis, the specific pet care method includes the processor controlling the massage module to massage the pet, and also includes the processor controlling the interaction module to simulate the sound of the specific pet type to output feedback statements to communicate with the pet according to the specific pet type, thereby realizing the automation and intelligence of pet care.

[0142] The present invention has been described in detail above. The above description is only a preferred embodiment of the present invention and should not limit the scope of implementation of the present invention. That is, all equivalent changes and modifications made within the scope of this application should still fall within the scope of the present invention.

Claims

1. A pet care method based on artificial intelligence, characterized in that: It includes a pet AI information interaction model, which is applied to an intelligent pet care device. The intelligent pet care device includes a processor, a storage module, a communication module, a detection module, an interaction module, a massage module and a pet cleaning module. The pet information is obtained through the detection module and sent to the pet AI information interaction model. The pet information includes the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The pet AI information interaction model obtains a specific pet care method according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice. The specific pet care method includes the processor controlling the massage module to massage the pet and the processor controlling the interaction module to simulate the sound of a specific type of pet according to the specific pet type to output feedback statements to communicate with the pet. The pet AI information interaction model includes a pet AI large model or an AI search engine.

2. The artificial intelligence-based pet care method according to claim 1, characterized in that: The pet AI big model generates a specific pet care method based on the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice analysis. The pet AI big model obtains the specific pet care method including two steps: building a pet AI big model and calling the pet AI big model. The building of the pet AI big model includes six steps: Step 21101, collecting pet care data: collecting all information about pet care data presented through voice, video, text, picture or 3D model; Step 21102, pre-processing pet care data: Process all collected information about pet care data to ensure the integrity and availability of the information, including converting information in different formats into text, and editing the text information in a certain format to facilitate subsequent training of the AI ​​large model; Step 21103, select the AI ​​big model applicable to pets: select domestic and foreign third-party AI big models, and measure them by accuracy, response speed, and diversity indicators; Step 21104, training the pet AI big model: organize the pet care data set through step 21102, and then fine-tune the pet care data set using the third-party AI big model. After the training is completed, a pet AI big model with all relevant data for pet care is generated; Step 21105, verify and test the pet AI big model: perform effect detection and evaluation of the pet AI big model generated in step 21104 on a specific task. If the evaluation effect fails, continue to repeat steps 21101, 21102, 21103, and 21104, retrain until the effect evaluation passes, generate the pet AI big model and store it in the cloud platform or storage module; Step 21106, deploy and maintain the pet AI big model: deploy the newly generated pet AI big model to the cloud platform or storage module, and perform continuous maintenance and updates, and regularly update data to ensure the timeliness and accuracy of the data; the pet information input by the user in voice, video, text, picture or 3D model format is called by the pet AI big model to generate specific care methods; The calling of the pet AI big model specifically includes the following steps: Step 21201, assemble the query statement; Step 21202: The AI ​​big model performs reasoning operations: The query statement is passed to the AI ​​big model, and the AI ​​big model performs reasoning operations, which includes the following processes: 1) Understanding input: Distributed semantic parsing first receives a text sequence and converts it into a word vector. This process is based on the distributed semantics assumption that the meaning of a word is determined by its use in context. 2) Parameter association: context focus linkage, input these word vectors into the Transformer Encoder to generate context representation; 3) Generate answers: Generative probability modeling. The model initializes the decoder part of the Transformer and inputs the encoder output and the current output sequence into the decoder. The decoder generates the probability distribution of the next word and selects the word with the largest probability or other set probability distribution as the output. This word will be added to the output sequence. 4) Select the most appropriate answer: Dynamic word string evolution, repeat the above steps, adding new words to the output sequence each time until a complete output sequence is generated; Step 21203, AI big model returns the result: the specific care method is sent to the interactive module or the user mobile terminal for confirmation after the pet AI big model reasoning operation is completed; after the pet AI big model reasoning operation is completed, relevant information about the specific care method is returned, including content provided to the user in text, picture, audio, video or 3D model format.

3. The artificial intelligence-based pet care method according to claim 1, characterized in that: Calling the AI ​​search engine to obtain specific pet care methods includes four steps: Step 1: Apply for the API interface of the AI ​​search engine company; Step 2: Obtain the pet AI big model through the detection module according to the pet type, pet body shape, pet age, pet hair information, pet emotional state and pet real-time voice, and encapsulate it into the parameters required by the API; Step 3: Call the API interface and encapsulate the parameters required by the API to call the API interface; Step 4: Process the API call result and extract the specific pet care method from the data returned by the API.

4. The artificial intelligence-based pet care method according to claim 1, characterized in that: The detection module includes a visual recognition module, which obtains relevant depth information, RGB color information, pet's body shape and hair thickness of the pet. The processor performs three-dimensional modeling of the pet according to the relevant depth information. The processor identifies the pre-working position according to the three-dimensional modeling, the pet's body shape and hair thickness, and analyzes and calculates to generate a massage trajectory. The processor controls the massage module to adjust to a pre-positioning position according to a specific pet care method and controls the massage module to work according to the massage trajectory, wherein the pre-positioning position is the initial position of the massage module relative to the pet when it is working.

5. The artificial intelligence-based pet care method according to claim 4, characterized in that: The visual recognition module captures the position and posture of the pet in real time to analyze and calculate to generate a massage trajectory, and the processor controls the massage module in real time to work according to the massage trajectory.

6. The artificial intelligence-based pet care method according to claim 1, characterized in that: The processor controls the interaction module to simulate the sound of a specific type of pet to output feedback statements to guide the pet to the pet care area according to the type of pet, and the processor controls the intelligent pet care device to care for the pet according to the specific pet care method, wherein the interaction module simulates the sound of a specific type of pet to output feedback statements to guide the pet, including a pet AI information interaction model that determines the characteristics and habits of a specific type of pet according to the type of pet, and generates feedback statements to guide the pet according to the characteristics and habits of the specific type of pet, and sends the feedback statements to guide the pet to the interaction module, and the processor controls the interaction module to simulate the sound of the pet to output feedback statements to guide the pet.

7. The artificial intelligence-based pet care method according to claim 6, characterized in that: The specific pet care method also includes a cleaning method. The pet AI information interaction model analyzes the pet cleaning status according to the pet hair information and generates a specific pet cleaning method according to the pet type, pet body size and pet age. The specific pet cleaning method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet to guide the pet to walk to the pet care area according to the pet type. The processor controls the pet cleaning module to execute the specific pet cleaning method to clean the pet.

8. The artificial intelligence-based pet care method according to claim 6, characterized in that: It also includes a pet feeding module, and the specific pet care method also includes a pet feeding method. The pet AI information interaction model generates a specific pet feeding method according to the pet type, pet body size, pet age, pet emotional state and pet real-time voice analysis. The specific pet feeding method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet to guide the pet to walk to the pet care area according to the pet type, and the processor controls the pet feeding module to feed the pet in the pet care area according to the pet feeding method.

9. The artificial intelligence-based pet care method according to claim 6, characterized in that: The specific pet care method also includes a defecation guidance method, wherein the pet AI information interaction model generates a specific defecation guidance method according to the pet type, the pet's emotional state and the pet's real-time voice analysis, and the specific defecation guidance method includes the processor controlling the interaction module to simulate the sound output of the specific type of pet according to the pet type to guide the pet's feedback statement to walk to a specific defecation area for defecation.

10. The artificial intelligence-based pet care method according to claim 1, characterized in that: The processor controls the interaction module to simulate the sound of a specific type of pet to output feedback statements to communicate with the pet. The pet AI information interaction model analyzes the semantics of the pet's real-time voice according to the pet's real-time voice and generates feedback statements, and sends the feedback statements to the interaction module. The processor controls the interaction module to simulate the sound of a specific type of pet to output feedback statements according to the pet type to have a real-time conversation with the specific type of pet.

11. The artificial intelligence-based pet care method according to claim 10, characterized in that: The feedback sentences include chat language, soothing language and command language.

12. The artificial intelligence-based pet care method according to claim 1, characterized in that: It also includes a mobile device, the processor identifies the location of the pet through the detection module and feeds back to the processor, the processor controls the mobile device to move toward the pet according to the location of the pet, and controls the smart pet care device to care for the pet according to a specific pet care method.

13. The artificial intelligence-based pet care method according to claim 7, characterized in that: The processor controls the pet cleaning module to clean the pet according to the cleaning method, including cleaning, combing, drying, and maintaining the pet's hair, and absorbing the pet's fallen hair.

14. The artificial intelligence-based pet care method according to claim 13, characterized in that: The pet cleaning module includes an adsorption module, a combing module, a cleaning module and a drying module, and the cleaning method includes: The processor controls the adsorption module to adsorb pet hair through a mechanical arm to adsorb pet hair and dander dropped by the pet; The processor controls the combing module to comb the pet's hair through the robotic arm; The processor controls the cleaning module to clean the pet through the robotic arm; The processor controls the drying module to dry the pet's hair through the mechanical arm; The processor controls the combing module to comb the pet's hair through the robotic arm.

15. The artificial intelligence-based pet care method according to claim 1, characterized in that: The specific pet care methods also include motion soothing, music soothing and image soothing.

16. The artificial intelligence-based pet care method according to claim 1, characterized in that: The specific pet care method also includes exchanging information with the pet in voice, video, picture or 3D model format through the interaction module.

17. The artificial intelligence-based pet care method according to claim 1, characterized in that: The user inputs care task information through the interaction module and sends the care task information to the pet AI information interaction model, and the pet AI information interaction model generates a specific pet care method based on the pet care task information analysis. The interaction module also includes a touch screen, a camera, a microphone and a speaker. The user can use the interaction module to interact with information in voice, video, text, picture, code scanning recognition or 3D model format.

18. The artificial intelligence-based pet care method according to claim 1, characterized in that: It also includes a floor cleaning module, and the detection module also includes a visual module. The visual module monitors the sanitary conditions of the pet activity area, and the processor controls the floor cleaning module to clean the pet activity area.

19. The artificial intelligence-based pet care method according to claim 1, characterized in that: The detection module also includes a force sensor. When executing a specific pet care method, the force sensor detects that the force of the massage module or the pet cleaning module is greater than a preset force and feeds back to the processor, and the processor controls the massage module or the pet cleaning module to stop working.

20. The artificial intelligence-based pet care method according to claim 1, characterized in that: The detection module also includes a safety module. When executing a specific pet care method, the safety module detects that the pet is unwell and feeds back to the processor, and the processor controls the massage module or the pet cleaning module to stop working.

21. The artificial intelligence-based pet care method according to claim 1, characterized in that: An induction module is also provided, which includes a heating device, an air supply device and a bait placement box. The processor controls the heating device to heat the bait placement box containing bait and controls the air supply device to supply air toward the bait according to pet information, so as to lure the pet to the pet care area. The processor controls the intelligent pet care device to care for the pet according to a specific pet care method.

22. The artificial intelligence-based pet care method according to claim 1, characterized in that: It also includes an AI virtual animal image system, the detection module obtains a specific type of pet image, a specific type of pet action and a specific type of pet sound, and sends the specific type of pet image, specific type of pet action and specific type of pet sound to the AI ​​virtual animal image system, the detection module obtains the pet's emotional state or the pet's real-time voice and sends it to the pet AI information interaction model, the pet AI information interaction model generates a feedback statement based on the pet's emotional state or the pet's real-time voice analysis and sends it to the AI ​​virtual animal image system, the AI ​​virtual animal image system can analyze and train the specific pet image, specific type of pet action, specific pet sound and feedback statement to output a specific virtual pet, and send it to the interaction module, wherein the specific virtual pet includes a feedback statement output by a specific virtual specific type of pet image simulating a specific virtual human action with a specific type of pet sound, and the interaction module outputs the specific virtual pet to communicate with the pet by voice.

23. The artificial intelligence-based pet care method according to claim 1, characterized in that: It also includes an AI virtual human system, wherein the detection module obtains the pet owner's image, pet owner's actions and pet owner's voice, and sends the pet owner's image, pet owner's actions and pet owner's voice to the AI ​​virtual human system, the detection module obtains the pet's emotional state and / or the pet's real-time voice and sends it to the pet AI information interaction model, the pet AI information interaction model generates feedback statements based on the pet's emotional state and / or the pet's real-time voice analysis and sends it to the AI ​​virtual human system, the AI ​​virtual human system can analyze and train the pet owner's image, pet owner's actions, pet owner's voice and feedback statements to output a specific virtual human, and send it to the interaction module, wherein the specific virtual human includes the feedback statement output by the pet owner's image in the pet owner's voice, and the interaction module outputs the specific virtual human to have a conversation with the pet.

24. The artificial intelligence-based pet care method according to claim 8, characterized in that: It also includes an AI agent system, wherein the AI ​​agent system is provided with a memory module, a planning module, an execution module and a tool module, and the pet type, pet body shape, pet age, pet hair information, pet emotional state, interaction record and feeding record data are pushed to the AI ​​agent system, and the AI ​​agent system outputs a pet health management plan, wherein the AI ​​agent system outputs a pet health management plan including the following steps: S1, the AI ​​agent system receives pet type, pet body size, pet age, pet hair information, and pet emotional state, and analyzes the pet health status according to the pet type, pet body size, pet age, pet hair information, and pet emotional state; S2, the planning module generates a pet health management task according to the pet care knowledge base, feeding record data and interaction record of the memory module, and sends the pet health management task to the planning module, wherein the pet health management task includes pet diet management and interaction management within a specific period, and the pet health management task is decomposed into a plurality of subtasks by reasoning, wherein each of the subtasks includes feeding the pet according to the pet feeding method at any time within a period and interacting with the pet according to a specific pet care method at any time within a period, wherein the interaction includes communicating, massaging or soothing the pet according to the specific pet care method; S3, the AI ​​agent system formulates a corresponding subtask plan according to each of the subtasks, and sends it to the interaction module for user confirmation or modification, wherein the user modifies the subtask plan and feeds back to the AI ​​agent system, and the AI ​​agent system returns to step S2 to re-plan multiple subtasks of the pet health management task until a health management pet plan is generated, wherein each of the subtask plans includes an execution date and execution time for each feeding of the pet according to the pet feeding method and an execution date and execution time for each interaction with the pet according to a specific pet care method; The AI ​​agent system executes the health management pet plan including the following steps: The AI ​​agent system plans the subtasks and calls the application tool through the tool module to execute the corresponding subtasks, wherein the application tool includes a calendar tool and a time tool. The tool module calls the calendar tool and the time tool to set the execution date and execution time of the interaction. The pet AI information interaction model sends the specific pet care method to the processor according to the execution date and execution time of the interaction, and the processor controls the smart pet care device to interact with the pet; the tool module calls the calendar tool and the time tool to set the execution date and execution time of feeding. The pet AI information interaction model sends the pet feeding method to the processor according to the execution date and execution time of feeding, and the processor controls the feeding module to feed the pet.

25. The artificial intelligence-based pet care method according to claim 1, characterized in that: A Vincent video model is also provided, and the Vincent video model generates a specific care video according to the specific pet care method, including the following steps: The specific pet care method is sent to the Vincent video model, and the Vincent video model generates a specific care video according to the specific pet care method. The pet AI information interaction model converts the specific pet care method into text content, and calls the Vincent video model through an API, and provides the text content to the Vincent video model. The Vincent video model analyzes and generates the specific care video and sends the specific care video to the interaction module. When the processor executes the specific pet care method, the interaction module outputs the specific care video.

26. The artificial intelligence-based pet care method according to claim 23, characterized in that: It also includes a Vincent music model and video synthesis software. The steps to generate a specific nursing video include: The pet AI information interaction model converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model, and the Vincent music model generates a specific care song according to the care lyrics; The specific care video and the specific care song are synthesized into a specific care short film through a video synthesis system and sent to the interactive module. When the processor executes the specific pet care method, the interactive module outputs the specific care short film.

27. The artificial intelligence-based pet care method according to claim 1, characterized in that: Also included is a Vincent music model, which generates a specific pet care song according to the specific pet care method, including the following steps: The pet AI information interaction model converts the specific pet care method into care lyrics and sends the care lyrics to the Vincent music model. The Vincent music model generates a specific care song based on the care lyrics and sends it to the interaction module. When the processor executes the specific pet care method, the interaction module outputs the specific care song.

28. The artificial intelligence-based pet care method according to claim 1, characterized in that: The detection module includes a camera, which records the daily activities of the pet and stores them in the storage module or the cloud platform. The user's mobile terminal obtains the pet's daily video through the communication module, and the user shares it through the user's mobile terminal.

29. An intelligent pet care device, characterized in that: The system comprises a processor, a storage module, a communication module, a detection module, an interaction module, a massage module and a pet cleaning module, wherein the processor, the storage module, the communication module, the detection module, the interaction module, the massage module and the pet cleaning module are all electrically connected to the processor, and further comprises a pet AI information interaction model arranged on the storage module or the cloud platform, wherein the pet AI information interaction model comprises a pet AI large model and an AI search engine, the intelligent pet care device performs information interaction with the cloud platform or the interaction module through the communication module, the detection module is used to obtain pet information and send it to the pet AI information interaction model, and the processor executes the specific pet care method described in any one of claims 1 to 10.

30. The intelligent pet care device according to claim 28, characterized in that: It also includes a mechanical arm, a base and a mobile device, the base is arranged on the mobile device, the processor, the storage module and the communication module are arranged in the base, and the mechanical arm is connected to the massage module and the pet cleaning module by grasping.

31. The intelligent pet care device according to claim 28, characterized in that: It also includes a base and a mobile device, the base is arranged on the mobile device, the processor, the storage module and the communication module are arranged in the base, a first robotic arm is arranged corresponding to the massage module, the first robotic arm is arranged on the base, and the massage module is arranged at the end of the first robotic arm, and a second robotic arm is arranged corresponding to the pet cleaning module, the second robotic arm is arranged on the base, and the pet cleaning module is arranged at the end of the second robotic arm.

32. The intelligent pet care device according to claim 29 or 30, characterized in that: The massage module also includes a vibration massage device, a knocking massage device and a kneading massage device.

33. The intelligent pet care device according to claim 29 or 30, characterized in that: The pet cleaning module includes a hair combing module, an adsorption module and a drying module.

34. The intelligent pet care device according to claim 28, characterized in that: It also includes a floor cleaning module, which includes an adsorption module, a floor sweeping device and a garbage collection device.

Images