Social intelligent management method and device based on large model, medium and equipment

Abstract

The invention provides a socialized intelligent management method and device based on a large model, a medium and equipment. The method comprises the following steps: acquiring monitoring data of wearable equipment of a first movable object; processing the monitoring data by adopting a first large model to obtain a feature vector, and calculating the similarity between the feature vector of the first movable object and the feature vector of each second movable object in at least one community network; according to a similarity calculation result, determining a matching degree between the first movable object and the at least one community network; determining an adaptive network community of the first movable object according to the degree of matching between the first movable object and the at least one community network, and returning the adaptive network community to a host client corresponding to the first movable object, the host of the first movable object is enabled to add the first movable object to the adaptive network community on the host client. According to the invention, the community management of the movable object can be realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent management technology, and in particular to a socialized intelligent management method, device, medium and equipment based on a large model. Background Art

[0002] With the prosperity of society and the improvement of people's quality of life, mobile devices / animals, such as cars, electric cars, family pets, wild animals studied by zoologists, animals in zoos, etc., have become indispensable partners of humans. Therefore, it is necessary to provide a management solution for mobile devices / animals. Summary of the invention

[0003] In response to at least one of the above technical problems, an embodiment of the present invention provides a social intelligent management method, device, medium, and equipment based on a big model.

[0004] According to the first aspect, the social intelligent management method based on a large model provided by an embodiment of the present invention includes:

[0005] Acquiring monitoring data of a wearable device of a first movable object;

[0006] The monitoring data is processed using a first large model to obtain a feature vector of the first movable object, and similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network are calculated; wherein the first movable object and the second movable object are mobile devices or animals;

[0007] Determining, according to the similarity calculation result, the matching degree between the first movable object and the at least one community network respectively;

[0008] According to the matching degree between the first movable object and the at least one community network, an adaptation network community of the first movable object is determined, and the adaptation network community is returned to a host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

[0009] In one embodiment, the method may further include:

[0010] extracting health data from the monitoring data;

[0011] Using the second largest model to analyze the health data, and obtaining an analysis result of whether the first movable object has a health risk;

[0012] If the analysis result is that the first movable object has a health risk, a warning message is generated and the warning message is returned to a host client corresponding to the first movable object.

[0013] In one embodiment, the method may further include:

[0014] extracting behavioral data from the monitoring data;

[0015] Analyzing the behavior data by machine learning technology to obtain the behavior habits of the first movable object;

[0016] The behavior habits, health data and attribute information of the first movable object are used as social information, and the social information is published to the adaptation network community.

[0017] In one embodiment, the method may further include:

[0018] Collecting maintenance question-answer pairs of multiple movable objects; wherein the maintenance question-answer pairs are information pairs formed by questions and answers about maintenance knowledge;

[0019] The maintenance question and answer pairs are labeled; wherein the labeling information is the type of movable object involved and the type of question;

[0020] Divide the labeled multiple maintenance question-answer pairs into training sets and test sets;

[0021] Using the training set to perform model training, to obtain an intelligent question-answering model with converged training;

[0022] Using the test set to perform performance evaluation on the intelligent question answering model, and using the performance evaluation results to optimize and adjust the intelligent question answering model;

[0023] The optimized and adjusted intelligent question and answer model is deployed, and the deployed intelligent question and answer model is integrated with the host client corresponding to the first movable object, so that the host of the first movable object can raise maintenance questions on the host client and obtain maintenance answers based on the intelligent question and answer model.

[0024] In one embodiment, the method may further include:

[0025] Performing anomaly detection on the monitoring data using an anomaly detection algorithm to obtain a detection result of whether the first movable object has an anomaly;

[0026] If the detection result shows that the first movable object is abnormal, a warning message is pushed to a host client of the first movable object, and a help message is sent to host clients of movable objects adjacent to the first movable object.

[0027] In one embodiment, the method may further include:

[0028] Upon receiving a lost search request submitted by the host of the first movable object on the host client, obtaining the last known location and behavior habits of the first movable object;

[0029] predicting a movement path of the first movable object according to the behavior habit and the last known position;

[0030] A search request is broadcast to the host clients of the movable objects on the action path, and a search status fed back by the host clients of the movable objects on the action path is received.

[0031] In one embodiment, the method may further include:

[0032] Encrypting the real-time relevant data of the first movable object; wherein the real-time relevant data includes at least one of the registration information, location update information, loss information and recovery information of the first movable object;

[0033] The encrypted real-time relevant data is stored on the blockchain, and the access and sharing of the real-time relevant data is controlled through smart contract technology.

[0034] According to the second aspect, the social intelligent management device based on the big model provided by the embodiment of the present invention includes:

[0035] A data acquisition module, used to acquire monitoring data of a wearable device of a first movable object;

[0036] a similarity calculation module, configured to process the monitoring data using a first large model to obtain a feature vector of the first movable object, and calculate similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network; wherein the first movable object and the second movable object are mobile devices or animals;

[0037] a matching determination module, configured to determine the matching degree between the first movable object and the at least one community network respectively according to the similarity calculation result;

[0038] A community determination module is used to determine an adaptation network community of the first movable object according to the matching degree between the first movable object and the at least one community network, and return the adaptation network community to a host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

[0039] According to a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method provided in the first aspect.

[0040] According to a fourth aspect, an embodiment of the present invention provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the method provided in the first aspect is implemented.

[0041] The social intelligent management method and apparatus, medium, and device based on a large model provided in the embodiment of the present invention obtain the monitoring data of the wearable device of the first movable object, and then use the first large model monitoring data for processing to obtain the feature vector of the first movable object, calculate the similarity between the feature vector of the first movable object and the feature vectors of each second movable object in at least one community network, and then determine the degree of matching between the first movable object and the at least one community network according to the similarity calculation result, so as to determine the adaptation network community of the first movable object, and return the adaptation network community to the host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client. It can be seen that the method provided in the embodiment of the present invention can realize the community management of movable objects and promote intelligent social interaction between the hosts of movable objects. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A schematic diagram of a process of a socialized intelligent management method based on a large model in one embodiment of the present invention;

[0043] Figure 2 It is a structural block diagram of a social intelligent management device based on a large model in one embodiment of the present invention. DETAILED DESCRIPTION

[0044] In the first aspect, the embodiment of the present invention provides a social intelligent management method based on a large model, see Figure 1 The method comprises the following steps S110 to S140:

[0045] S110, acquiring monitoring data of a wearable device of a first movable object;

[0046] Wearable devices can involve community detection and social interaction hardware, which uses UWB ultra-wideband technology for precise distance measurement and can establish short-range communication with surrounding movable objects, which helps to achieve rapid community construction and location awareness. UWB stands for Ultra Wide Band, a wireless carrier communication technology.

[0047] The wearable device may also include an integrated GPS module, a multi-sensor array, a low-power Bluetooth / LoRaWAN communication module, and a microprocessor. Through the built-in GPS and sensors, the activity range of the first movable object can be detected in real time, and the information of the neighboring movable objects overlapping with the current activity range of the first movable object can be obtained, so as to facilitate adding the current first movable object and the neighboring movable objects to the same community. The multi-sensor array includes a heart rate sensor, a body temperature sensor, a motion sensor, etc.

[0048] Among them, the wearable device also involves an integrated environmental perception module, which includes a highly sensitive microphone and a high-definition camera, which can accurately capture the sound and movement details of the first movable object, thereby enriching the social interaction experience between the first movable object and the host. The integrated environmental perception module can also have excellent waterproof and dustproof functions to ensure stable operation in various environments. Its built-in voice interaction system and large-size display screen make the operation more convenient and intuitive. In addition, the host can remotely monitor the actions of the first movable object through the integrated environmental perception module and keep abreast of the dynamics of the first movable object at any time. At night, the integrated environmental perception module can also be used as a lighting tool to facilitate the night search of the first movable object. The form of the integrated environmental perception module can also be customized according to whether the first movable object is an animal or a device. If it is an animal, it can be made into a module with an insect repellent function. If it is a mobile device, such as an electric car, it can be made into a driving recorder, a travel light, an alarm, etc.

[0049] It can be seen that the wearable device of the first movable object integrates activity detection, health monitoring, and wireless communication functions. It also has a fireproof, dustproof, waterproof and insect repellent design, supports remote voice, remote video, sound warning, and light warning, and is equipped with GPS, sensor array, communication and microprocessor to detect the activity range of the first movable object and conduct social interaction; monitor health data, and also facilitate the large model to analyze the behavior, health status and attributes of the first movable object.

[0050] S120, using a first large model to process the monitoring data to obtain a feature vector of the first movable object, and calculating similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network; wherein the first movable object and the second movable object are mobile devices or animals;

[0051] S130, determining the matching degree between the first movable object and the at least one community network respectively according to the similarity calculation result;

[0052] S140: Determine an adaptation network community for the first movable object according to the degree of matching between the first movable object and the at least one community network, and return the adaptation network community to a host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

[0053] It can be seen that the monitoring data of the wearable device from the first movable object is received; the first large model is used to process and analyze the first movable object to generate a feature vector of the first movable object; the similarity between the feature vector of the first movable object and the feature vectors of each second movable object in each existing community network is calculated, so as to know the similarity between the first movable object and each second movable object in each existing community network, so as to determine the matching degree between the first movable object and each existing community network, and then determine the adaptation network community of the first movable object, return the adaptation network community to the host client corresponding to the first movable object, and the host of the first movable object can add the first movable object to the adaptation network community on the host client.

[0054] In one embodiment, the method may further include:

[0055] 11. Extracting health data from the monitoring data;

[0056] 12. Analyze the health data using the second largest model to obtain an analysis result of whether the first movable object poses a health risk;

[0057] 13. If the analysis result shows that the first movable object has a health risk, a warning message is generated and the warning message is returned to a host client corresponding to the first movable object.

[0058] Among them, health data refers to data in the monitoring data that can reflect the health status of the first movable object. For mobile devices, the health status reflects whether the mobile device can operate normally. For animals, the health status reflects whether the animals can live normally.

[0059] It can be seen that by analyzing the health data using the second largest model, it is possible to determine whether the first movable object has a health risk. If there is a health risk, an early warning message is generated and sent to the host client, which helps the host to take corresponding measures in time to reduce the occurrence of health risks.

[0060] In one embodiment, the method may further include the following steps 21 to 23:

[0061] 21. Extracting behavioral data from the monitoring data;

[0062] For mobile devices, behavioral data may include driving trajectory, car brand, configuration, oil usage, refueling location, music frequently played on the device, current wear and tear of the device, etc.

[0063] For animals, behavioral data may include running, playing, barking, eating, defecating, etc.

[0064] 22. Analyze the behavior data by machine learning technology to obtain the behavior habits of the first movable object;

[0065] Among them, behavioral habits refer to information with certain behavioral patterns that can be summarized based on behavioral data.

[0066] 23. Use the behavior habits, health data and attribute information of the first movable object as social information, and publish the social information to the adaptation network community.

[0067] For mobile devices, the attributes refer to some basic information of the mobile device, such as name, model, size, etc. For animals, the attributes may be basic information of the animal, such as breed, name, age, personality, etc.

[0068] It can be seen that the behavior habits, health data and attribute information of the first movable object are published as social information to the adaptation network community so that other hosts can understand and interact with it.

[0069] In one embodiment, the method may further include the following steps 31 to 36:

[0070] 31. Collect maintenance question-answer pairs for multiple movable objects; wherein the maintenance question-answer pairs are information pairs formed by questions and answers about maintenance knowledge;

[0071] Specifically, common questions and answers related to the maintenance of movable objects can be collected and organized from multiple sources, covering the matters and data involved in the process from the creation to the extinction of movable objects, and the collection should be as comprehensive and diverse as possible.

[0072] 32. Annotate the maintenance question and answer pair; wherein the annotation information includes the type of movable object involved and the type of question;

[0073] Specifically, the collected maintenance question-answer pairs are carefully annotated to ensure that the intelligent question-answering model can understand and accurately generate answers. Annotation information is additional label information, including the type of mobile device / animal (e.g., car, bicycle, cat, dog, bird, wild animal, zoo animal, etc.) and question type (e.g., health, diet, behavior, exercise, brand, location, etc.). This label information helps the model classify and process questions in different fields and topics.

[0074] 33. Divide the labeled multiple maintenance question-answer pairs into training sets and test sets;

[0075] 34. Perform model training using the training set to obtain an intelligent question-answering model with converged training;

[0076] Specifically, as a starting point, you can choose a pre-trained ChatGPT model. This model already has certain language understanding and generation capabilities, providing a good starting point for model training. Then use the prepared training set to fine-tune the model so that it can generate more accurate answers for movable objects in the maintenance field. During the fine-tuning process, the model will learn and adapt to the specific context and expression of movable objects in the maintenance field.

[0077] 35. Using the test set to perform performance evaluation on the intelligent question answering model, and using the performance evaluation results to optimize and adjust the intelligent question answering model;

[0078] Specifically, the performance of the model is evaluated through the test set, and optimization and adjustment are performed based on the evaluation results. This step is to ensure that the model can show good performance in practical applications.

[0079] 36. Deploy the optimized and adjusted intelligent question and answer model, and integrate the deployed intelligent question and answer model with the host client corresponding to the first movable object, so that the host of the first movable object can raise maintenance questions on the host client and obtain maintenance answers based on the intelligent question and answer model.

[0080] It can be seen that the trained model is deployed in the actual question-answering system to provide users with intelligent question-answering services related to movable object maintenance. The question-answering system is integrated with the front-end host client so that users can easily ask questions and get answers through the front-end interface. The design of the front-end interface should be concise and clear for easy user operation.

[0081] It can be seen that the implementation of the intelligent question-answering model is based on ChatGPT, natural language processing, and machine learning technologies. The host asks questions through the front-end interface, which can be questions about the health, diet, behavior, exercise, brand, location, etc. of the movable object. The intelligent question-answering model analyzes the questions, retrieves relevant information from the database based on the analysis results and label information, and generates accurate and useful answers to the host in combination with its own language generation capabilities. Further interactions can be carried out based on the host's feedback. The host can evaluate the answers or ask further questions, and the intelligent question-answering model will make corresponding adjustments and optimizations based on the user's feedback.

[0082] Of course, we can also continue to collect user feedback and data to evaluate and improve the performance of the model. User feedback and data are important bases for optimizing the model. Based on demand and feedback, we continuously update and optimize the database and intelligent question-answering model, including adding new questions and answers, optimizing the parameters and algorithms of the intelligent question-answering model, etc., to improve the accuracy and practicality of the intelligent question-answering model.

[0083] In one embodiment, the method may further include S41 to S42:

[0084] 41. Perform anomaly detection on the monitoring data using an anomaly detection algorithm to obtain a detection result of whether the first movable object is abnormal;

[0085] 42. If the detection result shows that the first movable object is abnormal, push warning information to a host client of the first movable object, and send help information to host clients of movable objects adjacent to the first movable object.

[0086] Specifically, the anomaly detection algorithm One-Class SVM can be used to detect anomalies in the monitoring data. If an anomaly is found after detection, an early warning is immediately pushed through the host client. In an emergency, a help message can be automatically sent to the host of the nearby movable object, and the help path can be optimized using multi-hop network theory, thereby realizing the early warning and help functions.

[0087] In one embodiment, the method may further include the following steps 51 to 53:

[0088] 51. upon receiving a lost search application submitted by the host of the first movable object on the host client, obtaining the last known location and behavior habits of the first movable object;

[0089] 52. Predicting a movement path of the first movable object based on the behavior habit and the last known position;

[0090] 53. Broadcast a search request to host clients of the movable objects on the action path, and receive search status fed back by the host clients of the movable objects on the action path.

[0091] In actual scenarios, the host client can periodically update the last location of the movable object. Specifically, this can be achieved by calling the updateLastLocation method. When a movable object is lost, the host reports the loss through the host client, the backend records the query record, and triggers a big data analysis task. The big data analysis task uses a machine learning prediction model to predict the current possible location of the movable object based on the last known location and behavioral habits of the movable object, and returns the prediction result to the host client. The host can initiate a retrieval action based on the predicted location. Of course, in order to search together with other hosts and increase the recall probability, the action path of the first movable object can be predicted, and then the search request can be broadcast to the host client of the movable object on the action path. In this way, each host on the action path will help find it, and then feedback the search situation, which can greatly improve the recall probability.

[0092] In one embodiment, the method may further include the following steps 61-62:

[0093] 61. Encrypting the real-time relevant data of the first movable object; wherein the real-time relevant data includes at least one of the registration information, location update information, loss information and recovery information of the first movable object;

[0094] 62. Store the encrypted real-time relevant data on the blockchain, and control access to and sharing of the real-time relevant data through smart contract technology.

[0095] It can be seen that the main function of the above steps is to achieve data storage and protection. Specifically, blockchain technology is used to ensure the immutability and traceability of data. Real-time related data is encrypted and stored on the blockchain, and access control and sharing control of data are achieved through smart contract technology. Real-time related data includes: registration information, location update information, loss information, location prediction, retrieval information, etc. For example, the host submits the relevant information of the movable object through the front-end host client, and the back-end service calls the registration method of the smart contract to record the information of the movable object on the blockchain to achieve the storage of registration information.

[0096] It is understandable that data storage and protection use blockchain and cloud technology to ensure the security and traceability of the data of the movable object. Moreover, privacy protection algorithms can be used to ensure a balance between big data model analysis and blockchain publicity, protecting the privacy of the host of the movable object while realizing the social and activity range detection functions of the movable object.

[0097] Among them, IPFS, the InterPlanetary File System, can be combined with blockchain technology to store data of movable objects to ensure that the data cannot be tampered with and is stored in a decentralized manner.

[0098] Among them, zero-knowledge proof technology can be used to verify the authenticity of movable object information without revealing specific details. Manage data sharing rules through smart contract technology to ensure compliant access. To manage data sharing through smart contract technology, you must first deploy smart contracts: deploy smart contracts that manage data access rights on the Ethereum network. Secondly, complete it through your own development language: connect to the Ethereum node. Then load the smart contract instance, and use the user address to call the smart contract's checkAccess method to determine whether the user has permission to access the data. According to the result returned by the smart contract, decide whether to continue the data access operation. The above business logic simplifies many details, such as account management, transaction signatures, etc. More comprehensive security measures and error handling are required in actual applications. In addition, according to the specific application scenario, the design of the smart contract and the functions of the client of the project itself can be further refined.

[0099] The main components of the host client mentioned above are: Movable Object Planet, Movable Object Square, Movable Object Home, Movable Object Remote, Movable Object Smart Brain, Movable Object Portrait, and My Movable Object.

[0100] Among them, the movable object planet. Through this part, the host can see the movable objects nearby. Active social method: actively click on the avatar of the movable object, enter the chat box, and actively communicate. It can also automatically send social signals or recommend friends based on the similarity score. It can also have different intelligent social interactions such as love bell, friend bell, play together, chat room, private chat, etc.

[0101] Among them, the movable object square. This part dynamically recommends excellent movable objects to the movable object host square based on the label, attribute / character, and distance calculation of the movable object. The host can post the dynamics of the movable object in the square to increase the exposure of the movable object, increase the attractiveness of their own movable object, and meet the social needs of the hosts.

[0102] Among them, the movable object returns home. This part can realize the registration, location update, loss report, etc. of the movable object, which needs to be actively completed by the host.

[0103] Among them, the movable object can be remotely controlled. The host can remotely perform voice, video, sound, and light warning operations.

[0104] Among them, the movable object intelligent brain is integrated with the intelligent question-answering model, which can receive questions input by the host and return maintenance knowledge.

[0105] Among them, the movable object portrait mainly displays the basic information, exercise, health, animal diet diary, animal sleep, monitoring and early warning of the movable object, and comprehensively records the relevant data of the movable object from multiple aspects. Among them, monitoring and early warning include health status, abnormal behavior, etc. When the movable object has health problems or abnormal behavior, an early warning will be issued to the host client, so that the host can deal with it in time. When the movable object is not with the host, the host can issue an early warning and ask for help, which can be seen by the hosts of other movable objects within 3 kilometers, and remind the surrounding hosts to avoid the same thing. By real-time monitoring and analysis of the behavioral data of movable objects, possible risk events such as illness, loss, etc. can be predicted, and early warning signals can be issued in advance.

[0106] Among them, My Movable Objects can display the basic information of the movable objects.

[0107] It can be seen that the embodiment of the present invention uniquely integrates big models and blockchain technology. Its main goal is to significantly improve the management level of the host of movable objects through a series of tightly designed functions, including intelligent social matching, comprehensive health monitoring, advanced data protection, and lost retrieval. The general process includes: real-time collection of various data of movable objects, after the data is governed, the use of AI big models and blockchain technology to realize functions such as intelligent social matching, health monitoring, danger warning, data protection, and lost relocation and retrieval. The host client can provide an interactive interface for social matching, comprehensive health monitoring, danger warning, and lost retrieval services for movable objects. Based on the embodiment of the present invention, the host can easily establish contact with other hosts, share maintenance experience, and can also monitor the health status of movable objects in real time, and quickly take corresponding measures when the movable objects are in danger or lost. The embodiment of the present invention can also provide a unified management interface for smart life solutions through a smart life management website, on which functions such as portrait, maintenance intelligent question and answer, analysis and warning, social, data storage and protection are realized.

[0108] It can be seen that the algorithm model implemented by the embodiment of the present invention based on big data, AI model, and blockchain technology has greatly promoted the intelligent social interaction between movable objects, improved the accuracy of health management of movable objects, enhanced data security and privacy protection, and at the same time, provided an efficient solution for the recovery of lost movable objects. It can be seen that the embodiment of the present invention not only improves the equipment security of movable objects or the quality of life of animals, but also brings peace of mind and convenience to the host.

[0109] In a second aspect, an embodiment of the present invention provides a social intelligent management device based on a large model, see Figure 2 , the device 100 comprises:

[0110] The data acquisition module 110 is used to acquire monitoring data of a wearable device of a first movable object;

[0111] A similarity calculation module 120 is used to process the monitoring data using a first large model to obtain a feature vector of the first movable object, and calculate similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network; wherein the first movable object and the second movable object are mobile devices or animals;

[0112] A matching determination module 130, configured to determine the matching degree between the first movable object and the at least one community network respectively according to the similarity calculation result;

[0113] The community determination module 140 is used to determine the adaptation network community of the first movable object according to the matching degree between the first movable object and the at least one community network, and return the adaptation network community to the host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

[0114] In one embodiment, the device may further include:

[0115] A health detection module is used to extract health data from the monitoring data; use the second largest model to analyze the health data to obtain an analysis result of whether the first movable object has a health risk; if the analysis result is that the first movable object has a health risk, generate warning information, and return the warning information to a host client corresponding to the first movable object.

[0116] In one embodiment, the device may further include:

[0117] A social publishing module is used to extract behavioral data from the monitoring data; analyze the behavioral data through machine learning technology to obtain the behavioral habits of the first movable object; use the behavioral habits, health data and attribute information of the first movable object as social information, and publish the social information to the adaptation network community.

[0118] In one embodiment, the device may further include:

[0119] A model training module is used to collect maintenance question and answer pairs of multiple movable objects; wherein the maintenance question and answer pairs are information pairs formed by questions and answers about maintenance knowledge; the maintenance question and answer pairs are labeled; wherein the labeled information is the type of movable objects involved and the type of questions; the labeled multiple maintenance question and answer pairs are divided into a training set and a test set; the training set is used to perform model training to obtain an intelligent question and answer model with converged training; the test set is used to perform performance evaluation on the intelligent question and answer model, and the performance evaluation results are used to optimize and adjust the intelligent question and answer model; the optimized and adjusted intelligent question and answer model is deployed, and the deployed intelligent question and answer model is integrated with the host client corresponding to the first movable object, so that the host of the first movable object can raise maintenance questions on the host client and obtain maintenance answers based on the intelligent question and answer model.

[0120] In one embodiment, the device may further include:

[0121] An anomaly detection module is used to perform anomaly detection on the monitoring data using an anomaly detection algorithm to obtain a detection result of whether the first movable object has an anomaly; if the detection result is that the first movable object has an anomaly, a warning message is pushed to a host client of the first movable object, and a help message is sent to a host client of a neighboring movable object of the first movable object.

[0122] In one embodiment, the device may further include:

[0123] A lost recall module is used to obtain the last known position and behavior habits of the first movable object upon receiving a lost search application submitted by the host of the first movable object on the host client; predict the movement path of the first movable object based on the behavior habits and the last known position; broadcast a search request to the host client of the movable object on the movement path, and receive the search status feedback from the host client of the movable object on the movement path.

[0124] In one embodiment, the device may further include:

[0125] A data storage module is used to encrypt the real-time relevant data of the first movable object; wherein the real-time relevant data includes at least one of the registration information, location update information, loss information and recovery information of the first movable object; the encrypted real-time relevant data is stored on the blockchain, and the access and sharing of the real-time relevant data are controlled through smart contract technology.

[0126] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0127] In a third aspect, an embodiment of the present invention provides a computer-readable medium, on which computer instructions are stored. When the computer instructions are executed by a processor, the processor executes the method provided in the first aspect.

[0128] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0129] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0130] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0131] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0132] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0133] It is understandable that the explanation, specific implementation methods, beneficial effects, examples, etc. of the relevant contents in the computer-readable medium provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0134] In a fourth aspect, an embodiment of the present specification provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements the method in any embodiment of the specification.

[0135] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the computing device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0136] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0137] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, widgets, or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0138] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A social intelligent management method based on a large model, characterized in that: include: Acquiring monitoring data of a wearable device of a first movable object; The monitoring data is processed using a first large model to obtain a feature vector of the first movable object, and similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network are calculated; wherein the first movable object and the second movable object are mobile devices or animals; Determining, according to the similarity calculation result, the matching degree between the first movable object and the at least one community network respectively; According to the matching degree between the first movable object and the at least one community network, an adaptation network community of the first movable object is determined, and the adaptation network community is returned to a host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

2. The method according to claim 1, characterized in that Also includes: extracting health data from the monitoring data; Using the second largest model to analyze the health data, and obtaining an analysis result of whether the first movable object has a health risk; If the analysis result is that the first movable object has a health risk, a warning message is generated and the warning message is returned to a host client corresponding to the first movable object.

3. The method according to claim 1, characterized in that Also includes: extracting behavioral data from the monitoring data; Analyzing the behavior data by machine learning technology to obtain the behavior habits of the first movable object; The behavior habits, health data and attribute information of the first movable object are used as social information, and the social information is published to the adaptation network community.

4. The method according to claim 1, characterized in that: Also includes: Collecting maintenance question-answer pairs of multiple movable objects; wherein the maintenance question-answer pairs are information pairs formed by questions and answers about maintenance knowledge; The maintenance question and answer pairs are labeled; wherein the labeling information is the type of movable object involved and the type of question; Divide the labeled multiple maintenance question-answer pairs into training sets and test sets; Using the training set to perform model training, to obtain an intelligent question-answering model with converged training; Using the test set to perform performance evaluation on the intelligent question answering model, and using the performance evaluation results to optimize and adjust the intelligent question answering model; The optimized and adjusted intelligent question and answer model is deployed, and the deployed intelligent question and answer model is integrated with the host client corresponding to the first movable object, so that the host of the first movable object can raise maintenance questions on the host client and obtain maintenance answers based on the intelligent question and answer model.

5. The method according to claim 1, characterized in that Also includes: Performing anomaly detection on the monitoring data using an anomaly detection algorithm to obtain a detection result of whether the first movable object has an anomaly; If the detection result shows that the first movable object is abnormal, a warning message is pushed to a host client of the first movable object, and a help message is sent to host clients of movable objects adjacent to the first movable object.

6. The method according to claim 3, characterized in that Also includes: Upon receiving a lost search request submitted by the host of the first movable object on the host client, obtaining the last known location and behavior habits of the first movable object; predicting a movement path of the first movable object according to the behavior habit and the last known position; A search request is broadcast to host clients of movable objects on the action path, and search conditions fed back by host clients of movable objects on the action path are received.

7. The method according to claim 1, characterized in that Also includes: Encrypting the real-time relevant data of the first movable object; wherein the real-time relevant data includes at least one of the registration information, location update information, loss information and recovery information of the first movable object; The encrypted real-time relevant data is stored on the blockchain, and the access and sharing of the real-time relevant data is controlled through smart contract technology.

8. A social intelligent management device based on a large model, characterized in that: include: A data acquisition module, used to acquire monitoring data of a wearable device of a first movable object; a similarity calculation module, configured to process the monitoring data using a first large model to obtain a feature vector of the first movable object, and calculate similarities between the feature vector of the first movable object and feature vectors of each second movable object in at least one community network; wherein the first movable object and the second movable object are mobile devices or animals; a matching determination module, configured to determine the matching degree between the first movable object and the at least one community network respectively according to the similarity calculation result; A community determination module is used to determine an adaptation network community of the first movable object according to the matching degree between the first movable object and the at least one community network, and return the adaptation network community to a host client corresponding to the first movable object, so that the host of the first movable object adds the first movable object to the adaptation network community on the host client.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

10. A computing device, characterized in that The method comprises a memory and a processor, wherein the memory stores executable codes, and when the processor executes the executable codes, the method according to any one of claims 1 to 7 is implemented.

Images