Small household electrical appliance diagnosis system based on AI customer service and AR imaging

Abstract

The invention provides a small household appliance diagnosis system based on AI customer service and AR imaging, and relates to the field of information systems and management. The management main system comprises an AI intelligent interaction module, an AR visual presentation module, a multi-source data fusion module, an intelligent diagnosis decision module, a fault prediction and prevention module, a user feedback learning module, an equipment linkage coordination module and a system core regulation and control unit. According to the small household electrical appliance diagnosis system based on the AI customer service and the AR imaging, the multi-source data fusion module is utilized, and user description data acquired by the AI intelligent interaction module and household electrical appliance operation parameters acquired by the sensor are accurately integrated. By means of a Kalman filtering algorithm and a principal component analysis algorithm, data noise is effectively eliminated, key features are extracted, a data foundation is built for the intelligent diagnosis decision module, in the intelligent diagnosis decision module, a support vector machine and a Bayesian classification algorithm cooperatively operate, and the fault diagnosis accuracy is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of information systems and management, and specifically to a small household appliance diagnosis system based on AI customer service and AR imaging. Background Art

[0002] With the rise of smart home concepts, small household appliances are becoming increasingly prevalent in our daily lives, playing a crucial role in improving our quality of life. From electric kettles and microwave ovens for cooking to vacuum cleaners and robot vacuums for cleaning, small household appliances have greatly facilitated our lives. However, they also frequently malfunction over time. According to market research, nearly 50% of users experience varying degrees of malfunction with their small appliances after 2-3 years of use. These malfunctions not only diminish the user experience, but some unaddressed faults can even pose safety risks, such as fires caused by motor short circuits. Therefore, developing an efficient and accurate small appliance fault diagnosis and repair guidance system is crucial for ensuring user quality of life and improving home safety.

[0003] At present, there are obvious deficiencies in the way small household appliances are handled:

[0004] Traditional after-sales repair: After a user discovers a problem, they contact after-sales service. After-sales service will initially diagnose the problem based on the user's description and arrange a home visit or provide guidance for repairs. However, users' descriptions of the problem are often inaccurate, making it difficult for after-sales service personnel to quickly identify the problem, increasing repair costs and time. According to statistics, approximately 40% of on-site repair visits require a second visit due to initial misdiagnosis, which not only wastes manpower and resources but also prolongs user waiting time.

[0005] Self-resolved issues: Some users consult product manuals or search online for solutions. However, manuals are often brief and inadequate for complex issues. Online information is also highly variable, requiring users to sift through it, and the accuracy of the information is uncertain. Surveys show that the success rate of self-resolved issues is only around 30%, and improper operation can easily lead to secondary damage.

[0006] Simple detection tools are available on the market to assist with small household appliance detection tools, such as simple multimeters for detecting circuit continuity. However, these tools have limited functions and can only detect the basic parameters of specific components. They cannot perform comprehensive diagnosis of overall household appliance failures and require users to have certain professional knowledge. The threshold for ordinary users to use them is relatively high.

[0007] Although existing methods can handle small household appliance failures to a certain extent, they still expose many problems:

[0008] Problem 1: In the traditional after-sales maintenance model, poor information transmission and slow response lead to inefficient fault handling, increasing users' time and economic costs.

[0009] Problem two, the user self-help approach due to the difficulty of information acquisition and judgment, resulting in poor failure solution effect, and may be due to improper operation to make the problem worse.

[0010] Problem three, the limitations of simple detection tools cannot meet the diagnosis needs of complex faults of small household appliances, and it is difficult to provide comprehensive and effective solutions.

[0011] Problem four, the existing method generally lacks intuitive and intelligent interaction experience, and cannot clearly show the user the fault location and maintenance process, resulting in the user often being at a loss when facing a fault.

[0012] Therefore, an innovative small household appliance diagnosis system based on AI customer service and AR imaging is urgently needed, which uses the intelligent interaction and data analysis capabilities of AI customer service and the visualization advantages of AR imaging to break through the limitations of traditional methods and achieve accurate diagnosis, efficient maintenance guidance, and intelligent management and optimization of the whole life cycle of small household appliances. SUMMARY

[0013] Technical problems solved

[0014] In view of the deficiencies of the prior art, the small household appliance diagnosis system based on AI customer service and AR imaging is provided, which solves the problems in the above background art.

[0015] Technical scheme

[0016] To achieve the above purpose, the following technical scheme is used: the small household appliance diagnosis system based on AI customer service and AR imaging comprises a management main system, the management main system comprises an AI intelligent interaction module, an AR visualization presentation module, a multi-source data fusion module, an intelligent diagnosis decision module, a fault prediction and prevention module, a user feedback learning module, a device linkage coordination module and a system core control unit;

[0017] The overall operation steps of the management main system are as follows:

[0018] Sp1, system initialization and connection establishment: start the management main system to activate the AI intelligent interaction module, initialize the AR visualization presentation module to prepare for device imaging display, start the system core control unit to coordinate the operation of each module, and establish a user terminal connection interface for receiving user fault feedback and operation instructions;

[0019] Sp2, multi-source data acquisition and preliminary fusion: the AI intelligent interaction module receives the user's description of the small household appliance fault and the operation record data, the multi-source data fusion module collects the household appliance operation parameters using sensors, including temperature, voltage, and speed, and preliminarily integrates the user description data and the operation parameters, and transmits them to the intelligent diagnosis decision module;

[0020] Sp3, AR imaging real-time display and fault area positioning: the AR visualization presentation module obtains the appearance image of the household appliance through the camera of the user's mobile phone, combines the information provided by the multi-source data fusion module, and generates an AR image of the internal structure of the household appliance in real time. The intelligent diagnostic decision module analyzes the data to determine the fault area, and the AR visualization presentation module accurately marks the fault area on the image;

[0021] Sp4, intelligent diagnosis and scheme generation: the intelligent diagnostic decision module analyzes the fused multi-source data based on a deep learning algorithm, matches a common fault mode library, generates a fault diagnosis result and a corresponding unique repair scheme, and the scheme includes repair steps, required tool and accessory information;

[0022] Sp5, fault prediction and prevention analysis: the fault prediction and prevention module predicts the time and type of future faults of the household appliance according to historical fault data, current operating parameters and environmental factors, generates prevention measures in advance, and transmits them to the system core control unit using big data analysis and machine learning algorithms;

[0023] Sp6, user feedback and learning optimization: user feedback on the diagnosis result and repair scheme is collected through the user feedback learning module, and the AI intelligent interaction module updates the fault mode library and diagnosis algorithm according to the feedback information to improve the diagnosis accuracy and the effectiveness of the scheme. The optimized results are transmitted to the intelligent diagnostic decision module;

[0024] Sp7, device linkage and collaborative processing: when the fault involves multiple small household appliances and is associated with the smart home system, the device linkage coordination module adjusts the operating state of the related devices according to the instructions of the intelligent diagnostic decision module to avoid the expansion of the fault and provide a temporary replacement scheme, such as adjusting the indoor temperature and humidity by linking the air purifier when the air conditioner fails;

[0025] Sp8, comprehensive diagnosis report generation and output: the system core control unit integrates the diagnosis result, repair scheme, fault prediction information and device linkage situation to generate a detailed comprehensive diagnosis report, which is output to the user through the user terminal connection interface, and the report data is stored for subsequent analysis and system optimization;

[0026] Sp9, cyclic monitoring and continuous improvement: the multi-source data fusion module continuously collects household appliance operating parameters, the intelligent diagnostic decision module analyzes the data in real time, the fault prediction and prevention module dynamically adjusts the prediction model, and the system core control unit drives the system to run in a cycle, continuously optimizing the diagnosis process and improving the service quality;

[0027] The management main system is sequentially and gradually run through the running steps to realize accurate diagnosis, efficient repair guidance, and intelligent management and optimization of the whole life cycle of small household appliances.

[0028] Preferably, the AI intelligent interaction module adopts a natural language processing and sentiment analysis algorithm fusion mechanism, wherein the natural language processing algorithm converts user natural language into machine understandable text features through morphological analysis, syntactic analysis and semantic understanding; the sentiment analysis algorithm judges the user text sentiment tendency relying on the combination architecture of the recurrent neural network and the convolutional neural network of deep learning, and then provides more targeted and personalized responses when interacting with the user.

[0029] Preferably, the multi-source data fusion module uses Kalman filtering algorithm and principal component analysis algorithm to implement real-time filtering of these parameters such as speed and removes noise interference to improve data accuracy; the principal component analysis algorithm performs dimensionality reduction processing on the filtered data, extracts main feature components, reduces data redundancy, and improves subsequent data analysis efficiency, so as to more effectively preliminarily fuse them with user description data.

[0030] Preferably, in the intelligent diagnosis and decision module, the fault diagnosis result generation adopts support vector machine and Bayesian classification algorithm for processing, wherein the support vector machine builds an optimal classification hyperplane to classify the fused multi-source data and preliminarily determine the fault category; the Bayesian classification algorithm corrects and optimizes the classification result of the support vector machine according to the prior probability of fault occurrence and the conditional probability of data characteristics, improves the fault diagnosis accuracy, and then generates the corresponding unique repair scheme.

[0031] Preferably, the fault prediction and prevention module uses a long short-term memory network and a grey prediction model fusion mechanism to predict the time and type of future faults of the household appliance, wherein the long short-term memory network processes the time series information in the historical fault data and the current operating parameters to capture the long-term dependence of the data; the grey prediction model generates and processes the original data through accumulation for the case of less data and high uncertainty, establishes a grey differential equation model, and predicts the fault trend of the household appliance, thereby improving the fault prediction reliability and generating preventive measures in advance.

[0032] Preferably, the user feedback learning module adopts a reinforcement learning and online learning algorithm fusion mechanism to update the fault mode library and the diagnosis algorithm, wherein the reinforcement learning algorithm optimizes the behavior strategy of the AI intelligent interaction module and the intelligent diagnosis and decision module by setting a reward mechanism and according to the satisfaction of the user feedback to the diagnosis result and the repair scheme; the online learning algorithm updates the model parameters in real time in the process of continuous user feedback, so that the system can quickly adapt to new fault conditions and user needs, continuously improve the diagnosis accuracy and the effectiveness of the scheme.

[0033] Preferably, the device linkage coordination module uses a genetic algorithm combined with a fuzzy control algorithm strategy when coordinating the adjustment of the operating state of the related devices, searches for the optimal device linkage scheme through the genetic algorithm, and through the simulation of the biological genetic evolution process, selects, crosses and mutates different device linkage combinations to optimize the device linkage effect; the fuzzy control algorithm formulates specific device control rules according to the fuzzy information of the fault type, severity and real-time state of the related devices, realizes accurate and flexible regulation and control of the devices, avoids the expansion of the fault, and provides an effective temporary replacement scheme.

[0034] Preferably, the system core regulation unit adopts an analytic hierarchy process combined with an association rule algorithm in data mining when generating the comprehensive diagnosis report, determines the weights of the diagnosis result, the maintenance scheme, the fault prediction information and the device linkage situation in the comprehensive diagnosis report through the analytic hierarchy process, so as to highlight the key information; the association rule algorithm mines the potential relationship between these factors, contains the association between different fault types and common maintenance tools, integrates the association information into the comprehensive diagnosis report, so that the content of the report is more rich and comprehensive, the report is output to the user through the user terminal connection interface, and the report data is stored for subsequent analysis and system optimization.

[0035] Preferably, the management main system dynamically adjusts the parameters and model structures of the multi-source data fusion module, the intelligent diagnosis decision module and the fault prediction and prevention module through the adaptive resonance theory, so that they can adapt to the changes of the home appliance operating environment and the fault mode; the particle swarm optimization algorithm optimizes the regulation and control strategy of the system core regulation unit, finds the optimal system operating parameter configuration by simulating the foraging behavior of a bird swarm, and continuously optimizes the diagnosis process and improves the service quality.

[0036] The system has a distributed architecture and can support edge computing, so that the data processing and analysis capability is distributed to the edge device close to the data source, thereby improving the real-time response capability and reducing the data transmission delay, and the edge computing node has intelligent decision-making capability and can perform preliminary fire cause identification and early warning locally, thereby reducing the data transmission amount and delay.

[0037] Advantages

[0038] The application provides a small household appliance diagnosis system based on AI customer service and AR imaging.

[0039] 1、The small household appliance diagnosis system based on AI customer service and AR imaging of the application utilizes a multi-source data fusion module to accurately integrate user description data obtained by an AI intelligent interaction module and household appliance operation parameters collected by a sensor. With the help of Kalman filtering algorithm and principal component analysis algorithm, data noise is effectively removed, and key features are extracted, laying a solid data foundation for an intelligent diagnosis decision module. In the intelligent diagnosis decision module, support vector machines and Bayesian classification algorithms work together, greatly improving the accuracy of fault diagnosis. Taking electric kettle fault diagnosis as an example, traditional methods often require repeated troubleshooting, which is time-consuming and prone to errors. However, the system can quickly and accurately locate the fault type and position, significantly reducing the fault diagnosis time and the idle time of household appliances caused by fault troubleshooting, and effectively improving the efficiency of household appliances.

[0040] 2、In the application, the AI intelligent interaction module uses natural language processing and sentiment analysis algorithm to make the communication between the user and the system as natural and smooth as daily conversation. Users only need to describe the fault in their usual language, and the system can accurately understand the intention and give personalized responses according to the sentiment analysis results. The AR visualization module obtains the appearance image of the household appliance through the camera of the mobile phone, generates the AR imaging of the internal structure immediately, and clearly marks the fault area. This allows non-professional users to intuitively see the fault location, completely changing the situation in traditional after-sales maintenance where users have difficulty understanding complex fault descriptions. For example, in air purifier fault diagnosis, users can clearly see the filter screen blockage position with the help of AR imaging, without waiting for the after-sales personnel to explain, they can prepare to replace the filter screen themselves, greatly enhancing the user's autonomy and convenience in handling faults, and effectively improving the user's satisfaction and trust in the system.

[0041] 3、The fault prediction and prevention module of the application uses a long short-term memory network and a gray prediction model fusion mechanism to comprehensively consider historical fault data, current operation parameters and environmental factors of household appliances, and accurately predict the time and type of fault occurrence in advance. Compared with traditional fault prediction methods, the system can detect potential fault risks earlier. For example, in small bread machine fault prediction, it can predict motor failure in advance, allowing users to have enough time to arrange maintenance or replace parts, effectively avoiding sudden failures that disrupt normal use. At the same time, the system also generates detailed prevention measures suggestions to help users prepare for maintenance work in advance, effectively extending the service life of household appliances and reducing maintenance costs, fundamentally improving the reliability and stability of small household appliances.

[0042] 4. In the present invention, when the fault is associated with multiple small household appliances and involves the smart home system, the device linkage coordination module plays a key role. Using genetic algorithms and fuzzy control algorithms, the system can quickly screen out the optimal device linkage plan and coordinate related equipment to adjust the operating status in an orderly manner. For example, when the air conditioner fails, the system will automatically link the air purifier to cleverly adjust the indoor temperature and humidity, maintain a comfortable and pleasant indoor environment, and prevent the overall home environment from being damaged by the failure of a single device. The system's core control unit uses adaptive resonance theory and particle swarm optimization algorithm to continuously optimize the parameters and operating strategies of each module. With the passage of time, the system's diagnostic accuracy and operating efficiency have steadily improved, creating a continuously optimized and upgraded smart home appliance management service for users, and creating a more intelligent, comfortable and efficient home environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is the system mind map of the present invention;

[0044] Figure 2 It is the overall framework diagram of the present invention;

[0045] Figure 3 This is a table of diagnostic accuracy experimental data of the present invention;

[0046] Figure 4 This is a table of fault prediction reliability experimental data of the present invention. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention. Specific embodiment one:

[0049] like Figures 1-2 As shown in the figure, the small household appliance diagnosis system based on AI customer service and AR imaging includes a main management system, which includes an AI intelligent interaction module, an AR visualization presentation module, a multi-source data fusion module, an intelligent diagnosis and decision module, a fault prediction and prevention module, a user feedback learning module, an equipment linkage coordination module, and a system core control unit;

[0050] The overall operation steps of the management main system are as follows:

[0051] Sp1, system initialization and connection establishment: turn on the management master system to activate the AI intelligent interaction module, initialize the AR visualization presentation module to prepare for device imaging display, start the system core regulation unit to coordinate the operation of each module, and establish a user end connection interface for receiving user fault feedback and operation instructions;

[0052] Sp2, multi-source data acquisition and preliminary fusion: the AI intelligent interaction module receives user descriptions and operation record data of small household appliance faults, the multi-source data fusion module collects household appliance operating parameters using sensors, including temperature, voltage, and speed, and preliminarily integrates user description data and operating parameters, and transmits them to the intelligent diagnosis decision module;

[0053] Sp3, AR imaging real-time display and fault area positioning: the AR visualization presentation module obtains household appliance appearance images through the user's mobile phone camera, combines the information provided by the multi-source data fusion module, and generates real-time AR imaging of the internal structure of the household appliance. The intelligent diagnosis decision module analyzes the data to determine the fault area, and the AR visualization presentation module accurately labels the fault area on the imaging;

[0054] Sp4, intelligent diagnosis and scheme generation: the intelligent diagnosis decision module analyzes the fused multi-source data based on deep learning algorithms, matches the common fault mode library, generates fault diagnosis results and corresponding unique repair schemes, and the scheme includes repair steps, required tools and accessory information;

[0055] Sp5, fault prediction and prevention analysis: the fault prediction and prevention module uses big data analysis and machine learning algorithms to predict the time and type of future faults based on historical fault data, current operating parameters and environmental factors, and generates preventive measures in advance, which are transmitted to the system core regulation unit;

[0056] Sp6, user feedback and learning optimization: user feedback on diagnosis results and repair schemes is collected through the user feedback learning module, the AI intelligent interaction module updates the fault mode library and diagnosis algorithm based on feedback information, improves diagnosis accuracy and scheme effectiveness, and optimizes the results transmitted to the intelligent diagnosis decision module;

[0057] Sp7, device linkage and collaborative processing: when the fault involves multiple small household appliances and is associated with the smart home system, the device linkage coordination module adjusts the operating state of related devices according to the instructions of the intelligent diagnosis decision module to avoid the expansion of the fault and provide temporary replacement solutions, such as linking air purifiers to adjust indoor temperature and humidity when the air conditioner fails;

[0058] Sp8, comprehensive diagnostic report generation and output: the system core control unit integrates diagnostic results, repair schemes, fault prediction information and equipment linkage conditions to generate detailed comprehensive diagnostic reports, which are output to users through user terminal connection interfaces, and the report data is stored for subsequent analysis and system optimization;

[0059] Sp9, cycle monitoring and continuous improvement: the multi-source data fusion module continuously collects household appliance operation parameters, the intelligent diagnostic decision module analyzes data in real time, the fault prediction and prevention module dynamically adjusts the prediction model, and the system core control unit drives the system to run in a cycle, continuously optimizing the diagnostic process and improving service quality;

[0060] Among them, the management main system is gradually run through the running steps, realizing the precise diagnosis, efficient repair guidance and intelligent management and optimization of the whole life cycle of small household appliances.

[0061] The AI intelligent interaction module adopts a natural language processing and sentiment analysis algorithm fusion mechanism. The natural language processing algorithm converts user natural language into machine-understandable text features through morphological analysis, syntactic analysis and semantic understanding methods. The sentiment analysis algorithm relies on the combination architecture of recurrent neural networks and convolutional neural networks based on deep learning to determine the sentiment tendency of user text, and then provides more targeted and personalized responses when interacting with users.

[0062] The multi-source data fusion module uses Kalman filtering algorithm combined with principal component analysis algorithm to implement real-time filtering of parameters such as speed and rotational speed, removes noise interference, and improves data accuracy. The principal component analysis algorithm reduces the dimensionality of the filtered data, extracts the main characteristic components, reduces data redundancy, and improves the efficiency of subsequent data analysis, so as to more effectively preliminarily fuse it with user description data.

[0063] In the intelligent diagnostic decision module, the fault diagnosis result generation adopts support vector machine and Bayesian classification algorithm for processing. The support vector machine builds an optimal classification hyperplane to classify the fused multi-source data and preliminarily determine the fault category. The Bayesian classification algorithm corrects and optimizes the classification results of the support vector machine based on the prior probability of fault occurrence and the conditional probability of data characteristics, improves the fault diagnosis accuracy, and then generates the corresponding unique repair scheme.

[0064] The fault prediction and prevention module utilizes a long short-term memory network and a grey prediction model fusion mechanism to predict the time and type of future failures of the household appliance. The long short-term memory network processes time series information in historical failure data and current operating parameters to capture long-term dependencies in the data. The grey prediction model generates processed data by accumulating the original data to establish a grey differential equation model for predicting the failure trend of the household appliance. The combination of the two improves the reliability of fault prediction and generates preventive measures in advance.

[0065] The user feedback learning module adopts a reinforcement learning and online learning algorithm fusion mechanism to update the fault mode library and diagnostic algorithm. The reinforcement learning algorithm optimizes the behavior strategy of the AI intelligent interaction module and the intelligent diagnostic decision module by setting a reward mechanism and considering the satisfaction of the user feedback on the diagnostic results and repair schemes. The online learning algorithm updates the model parameters in real time during continuous user feedback to enable the system to quickly adapt to new fault conditions and user needs, continuously improving diagnostic accuracy and scheme effectiveness.

[0066] The device linkage coordination module uses a genetic algorithm and fuzzy control algorithm combination strategy when coordinating related devices to adjust operating states. The genetic algorithm searches for the optimal device linkage scheme, and the fuzzy control algorithm formulates specific device control rules based on fuzzy information such as fault type, severity, and real-time state of related devices to achieve precise and flexible control of devices, avoid fault expansion, and provide effective temporary replacement schemes.

[0067] The system core regulation unit adopts an analytic hierarchy process and association rule algorithm in data mining for fusion when generating a comprehensive diagnostic report. The analytic hierarchy process determines the weight of factors such as diagnostic results, repair schemes, fault prediction information, and device linkage conditions in the comprehensive diagnostic report to highlight key information. The association rule algorithm mines the potential relationships between these factors, including the association between different fault types and common repair tools, and integrates this association information into the comprehensive diagnostic report to make the content more comprehensive and complete. The report data is output to the user through a user terminal connection interface and stored for subsequent analysis and system optimization.

[0068] The management main system dynamically adjusts the parameters and model structure of the multi-source data fusion module, intelligent diagnostic decision module, and fault prediction and prevention module through adaptive resonance theory to adapt to changes in the operating environment and fault modes of the household appliance. The particle swarm optimization algorithm optimizes the regulation strategy of the system core regulation unit by simulating the foraging behavior of bird flocks to find the optimal system operating parameter configuration, continuously optimizing the diagnostic process and improving service quality.

[0069] The system has a distributed architecture and can support edge computing, which can decentralize data processing and analysis capabilities to edge devices close to data sources, thereby improving real-time response capabilities and reducing data transmission delays. In addition, the edge computing node has intelligent decision-making capabilities, which can perform preliminary fire cause identification and early warning locally, reducing data transmission volume and delay.

[0070] The data transmission paths between the modules in the management main system are as follows: the AI intelligent interaction module transmits the user-related text feature data converted through natural language processing and sentiment analysis to the multi-source data fusion module through the internal data transmission channel of the system. The multi-source data fusion module accurately transmits the preliminary fusion data integrated with user description data and sensor-collected home appliance operation parameters to the intelligent diagnostic decision module. The intelligent diagnostic decision module transmits the fault diagnosis results and corresponding maintenance scheme data to the system core control unit, and simultaneously feeds back the diagnosis result data to the fault prediction and prevention module for optimizing the prediction model. The fault prediction and prevention module generates prevention measure suggestion data and transmits it to the system core control unit. The user feedback learning module collects feedback data from users on the diagnosis results and maintenance schemes, and transmits the data required for optimization to the AI intelligent interaction module and the intelligent diagnostic decision module for updating the fault mode library and diagnostic algorithm. The device linkage coordination module receives instruction data from the intelligent diagnostic decision module, adjusts the operating state of related devices according to the instructions, and transmits device linkage execution result data to the system core control unit. The system core control unit transmits comprehensive diagnosis report data to the user through the user terminal connection interface, and stores the report data for subsequent analysis and system optimization. According to the data transmitted by each module, the system core control unit generates control signals and transmits them to each corresponding module to drive the orderly operation of each module of the management main system, ensuring the smooth progress of the system's cyclic monitoring and continuous improvement process.

[0071] After the system is started, the management main system is started, the AI intelligent interaction module is activated, and it is ready to receive user fault feedback and operation instructions. At the same time, the AR visualization presentation module is initialized for subsequent device imaging display. The system core control unit is started, which is responsible for coordinating the operation of each module and establishing a user terminal connection interface.

[0072] In the multi-source data acquisition and preliminary fusion stage, the AI intelligent interaction module uses natural language processing and sentiment analysis algorithm fusion mechanism to convert the user's description and operation record data of small household appliance faults into machine understandable text features, which are transmitted to the multi-source data fusion module through the system internal data transmission channel. The multi-source data fusion module uses sensors to collect the temperature, voltage, speed and other operating parameters of the household appliance, uses Kalman filter algorithm combined with principal component analysis algorithm to real-time filter and reduce dimension of the operating parameters, reduces data redundancy and improves accuracy, and then integrates the data with the user description data and transmits them to the intelligent diagnosis decision module.

[0073] The AR visualization presentation module obtains the appearance image of the household appliance through the user's mobile phone camera, combines the information provided by the multi-source data fusion module, and generates the AR imaging of the internal structure of the household appliance in real time. The intelligent diagnosis decision module processes the fused multi-source data using support vector machine and Bayesian classification algorithm, matches the common fault mode library, determines the fault area, and the AR visualization presentation module accurately labels the area on the imaging, and generates the fault diagnosis result and the unique repair scheme. The result and scheme data are transmitted to the system core control unit, and the diagnosis result data is also fed back to the fault prediction and prevention module.

[0074] The fault prediction and prevention module uses long short-term memory network and gray prediction model fusion mechanism to predict the future fault time and type of the household appliance according to historical fault data, current operating parameters and environmental factors, and generates preventive measures suggestions in advance and transmits them to the system core control unit. The user feedback learning module collects user feedback data on the diagnosis result and repair scheme, uses reinforcement learning and online learning algorithm fusion mechanism to transmit the data required for optimization to the AI intelligent interaction module and the intelligent diagnosis decision module, and updates the fault mode library and diagnosis algorithm.

[0075] When the fault involves multiple small household appliances and is associated with the smart home system, the device linkage coordination module receives the instruction data from the intelligent diagnosis decision module, uses genetic algorithm and fuzzy control algorithm combined strategy to coordinate the related devices to adjust the operating state, avoid the expansion of the fault and provide temporary replacement scheme, and transmits the device linkage execution result data to the system core control unit.

[0076] The system core control unit integrates the diagnosis result, repair scheme, fault prediction information and device linkage situation, uses analytic hierarchy process and association rule algorithm in data mining to generate a comprehensive diagnosis report, and outputs the report to the user through the user terminal connection interface, and stores the report data for subsequent analysis and system optimization. In addition, the system core control unit generates control signals according to the data transmitted by each module and transmits them to the corresponding modules to drive the orderly operation of the modules of the management main system.

[0077] The whole system has a distributed architecture, supports edge computing, and downgrades data processing and analysis capabilities to edge devices close to data sources. The edge computing node performs preliminary small household appliance fault reason identification and early warning locally, reduces data transmission volume and delay, and improves real-time response capability, thereby realizing accurate diagnosis, efficient maintenance guidance, and intelligent management and optimization of the whole life cycle of small household appliances. Through cyclic monitoring and continuous improvement, the performance of the system is continuously improved. Embodiment two:

[0079] As shown in Figures 1-2 the following is a use case of the scheme in embodiment one:

[0080] Case one: fault diagnosis and maintenance of electric kettle

[0081] User feedback: the user found that the electric kettle at home became slower in boiling water, and there was an abnormal noise during the last few times of boiling water. Through the mobile phone APP, the user entered the system and described the fault to the AI intelligent interaction module. The AI intelligent interaction module uses natural language processing and sentiment analysis algorithms to understand the user's description and identify the user's anxious emotion, and quickly gives a soothing response and guides the user to supplement information such as the daily use frequency of the electric kettle and whether there has been a dry burning situation recently.

[0082] Data acquisition and fusion: the multi-source data fusion module receives the text feature data converted from the user's description transmitted by the AI intelligent interaction module, and at the same time collects the operating parameters such as the temperature, voltage and vibration frequency (reflecting the rotation speed) of the electric kettle through the built-in sensor. After removing parameter noise using Kalman filtering algorithm and dimensionality reduction using principal component analysis algorithm, the two types of data are integrated and transmitted to the intelligent diagnosis decision module.

[0083] AR imaging and diagnosis: the AR visualization presentation module generates AR imaging of the internal heating pipe, temperature controller and other components of the electric kettle on the user's mobile phone screen according to the information provided by the multi-source data fusion module. The intelligent diagnosis decision module uses support vector machine and Bayesian classification algorithm to analyze the fusion data, match the fault mode library, and determine that the heating pipe has slight fouling leading to reduced heating efficiency, and the temperature controller has poor contact causing abnormal noise. The AR visualization presentation module accurately labels the positions of the heating pipe and the temperature controller on the imaging.

[0084] Maintenance scheme generation: the intelligent diagnosis decision module generates a maintenance scheme, including turning off the power, using white vinegar and water mixed at a ratio of 1:10 to clean the heating pipe by soaking for 30 minutes, opening the base to check whether the temperature controller connection line is loose, and if so, replugging and fixing it. At the same time, it lists the required tools such as wrench and screwdriver, and the temperature controller accessories that may need to be replaced, and transmits the scheme to the system core control unit.

[0085] Fault prediction and feedback: The fault prediction and prevention module predicts that if the heating tube is not cleaned and the temperature controller is not repaired in time, the heating tube may be damaged and the kettle may not work normally in the next month based on historical fault data and current operating parameters of the electric kettle. The system core control unit generates preventive measures and transmits them to the system core control unit. The user feedback learning module collects user feedback on the diagnosis results and repair schemes. If the user successfully completes the repair and feedback problem is solved, the reinforcement learning algorithm rewards the system for correct diagnosis, and the online learning algorithm updates the fault mode library and diagnosis algorithm to improve the accuracy of subsequent diagnosis.

[0086] Comprehensive report and optimization: The system core control unit integrates the diagnosis results, repair schemes, and fault prediction information, determines the weight of each part using the analytic hierarchy process, and uses the association rule algorithm to mine the potential relationship between faults and daily use habits to generate a comprehensive diagnosis report. The user outputs the report through the user interface. At the same time, according to the data of each module, control signals are generated to drive the system to continuously optimize, such as adjusting the sensor acquisition frequency to more accurately monitor the operating status of the electric kettle.

[0087] Case two: intelligent air purifier fault handling and equipment linkage

[0088] Fault feedback: The user feels that the noise of the air purifier has become larger when running, and the indoor air quality has not improved significantly. Through voice and AI intelligent interaction module communication. The AI intelligent interaction module analyzes the user's voice and combines emotional analysis to determine the user's concern about indoor air quality, and asks for information such as air purifier indicator light status and filter replacement time.

[0089] Data integration: The AI intelligent interaction module transmits the processed user information to the multi-source data fusion module, which simultaneously collects air purifier wind speed, motor speed, filter resistance (reflected in voltage changes), etc. After Kalman filtering and principal component analysis processing, it is transmitted to the intelligent diagnosis decision module.

[0090] Diagnosis and imaging: The AR visualization rendering module generates AR imaging of the air purifier's internal air duct, motor, filter, and other components. The intelligent diagnosis decision module analyzes the data and determines that there are foreign objects in the air duct causing noise to increase and the filter is clogged, reducing purification effectiveness. The AR visualization rendering module marks the location of the air duct and filter in the imaging.

[0091] Repair and linkage: The intelligent diagnosis decision module generates a repair scheme to clean the air duct and replace the filter, and transmits it to the system core control unit. Because the fault affects indoor air quality, and the user's home has a smart humidifier and a smart home system associated with it, the equipment linkage coordination module receives the instructions from the intelligent diagnosis decision module, uses genetic algorithms and fuzzy control algorithms to reduce the humidity setting of the smart humidifier, and avoid high humidity environment to exacerbate air quality problems. At the same time, the linkage execution result is fed back to the system core control unit.

[0092] Prediction and Learning: The fault prediction and prevention module predicts that the motor may be damaged due to excessive load in the next week if the air duct is not cleaned and the filter is not replaced in time, and generates prevention measures suggestions to transmit to the system core control unit. The user feedback learning module collects user feedback after maintenance, and if the user indicates that the problem is solved, the system updates the fault mode library and diagnostic algorithm.

[0093] Report Output and System Optimization: The system core control unit generates a comprehensive diagnostic report containing fault causes, maintenance solutions, device linkage situations, and fault prediction information, and outputs it to the user. At the same time, based on the data of each module, the system parameters are optimized through the adaptive resonance theory and particle swarm optimization algorithm to improve the efficiency of diagnosis and device linkage.

[0094] Case Three: Fault Diagnosis and Full Life Cycle Management of Small Bread Machine

[0095] User Consultation: The user found that the bread crust was burnt and the inside was not cooked when making bread, and consulted the AI intelligent interaction module. The AI intelligent interaction module guides the user to describe the bread machine model, the bread making program used, the amount of ingredients, etc.

[0096] Data Collection and Processing: The multi-source data fusion module receives the user information processed by the AI intelligent interaction module, and at the same time collects the operating parameters such as the temperature of the heating element of the bread machine, the speed of the stirring motor, and the working time, and transmits them to the intelligent diagnostic decision module after processing by the algorithm.

[0097] Fault Location and Solution: The AR visualization presentation module displays the AR imaging of the heating element, stirring paddle and other components inside the bread machine. The intelligent diagnostic decision module analyzes and determines that the heating element temperature control system is faulty, causing the heating temperature to be too high and the stirring motor speed to be unstable, affecting the uniformity of the dough. The solution of maintaining the temperature control system and checking the motor and transmission components is generated and transmitted to the system core control unit.

[0098] Prevention and Feedback: The fault prediction and prevention module predicts that serious faults such as motor burnout and heating element damage may occur in the future based on the historical data of the bread machine and the current fault, and generates prevention measures suggestions. The user feedback learning module collects user feedback after maintenance, and if there is still a problem after maintenance, the system further optimizes the diagnostic algorithm.

[0099] Full Life Cycle Management: The system core control unit integrates the diagnostic results, maintenance solutions, and fault prediction information to generate a comprehensive diagnostic report. At the same time, by analyzing the historical use data of the bread machine, it provides regular maintenance suggestions such as cleaning the heating element every three months and checking the lubrication of the transmission components, etc., to realize intelligent management of the full life cycle of the bread machine. The system continuously monitors the operating parameters of the bread machine and optimizes the management strategy through the adaptive resonance theory and particle swarm optimization algorithm to improve the performance of the system. Specific embodiment three:

[0101] like Figures 1-2 As shown, the key algorithms mentioned in Example 1 are analyzed in detail below, including their core mathematical formulas and explanations:

[0102] Sentiment Analysis - Sentiment Classification Based on Recurrent Neural Network (RNN):

[0103] For an input sequence x=[x1,x2,...,x T ], the hidden layer state h of RNN t The update formula is:

[0104] h t =σ(W xh x t +W hh h t-1 +b h )

[0105] Among them, the output y t =W hy h t +b y

[0106] where x t is the input vector of the input sequence at time t, corresponding to the text feature vector after preprocessing such as lexical analysis; h t is the state vector of the hidden layer at time t, which combines the information of the current input and the hidden layer state at the previous moment; h t-1 is the state vector of the hidden layer at the previous moment (t-1); W xh is the weight matrix input to the hidden layer, which is used to transform the input vector x t Mapped to the hidden layer; W hh is the weight matrix from hidden layer to hidden layer, which is used to transmit the information of the hidden layer state at the previous moment; W hy is the weight matrix from the hidden layer to the output layer, which is used to map the hidden layer state to the output result; b h is the bias vector of the hidden layer; b y is the bias vector of the output layer; σ is the activation function, usually the sigmoid function, which is used to introduce nonlinear factors so that the model can learn complex patterns. The formula is:

[0107]

[0108] This formula analyzes the user's text sequence word by word and continuously updates the hidden layer state to capture the emotional information in the text. It ultimately outputs a result representing the text's emotional tendency (such as positive, negative, or neutral), helping the AI ​​intelligent interaction module respond to users more humanely.

[0109] Kalman filter algorithm:

[0110] Prediction step:

[0111] State prediction:

[0112] Covariance prediction: P k k-1 = AP k-1|k-1 + Q T

[0113] Update step:

[0114] Kalman gain: K k = P k|k-1 H T (HP k|k-1 H T + R) -1

[0115] State update:

[0116] Covariance update: P k|k = (I - K k H)P k|k-1

[0117] Character interpretation:

[0118] where k is the current time; is the state prediction value at the current time (k) based on the state estimate value at the previous time (k-1); is the state estimate value at the previous time (k-1); A is the state transition matrix, describing the transition relationship of the system state from one time to the next; B is the control input matrix, used to map the control input u k to the state space (in sensor data processing, B and u k can be ignored if there is no control input involved); u k is the control input at time k (for sensor acquisition parameter filtering, it can be considered as zero or irrelevant); P k|k-1 is the covariance matrix of the predicted state , measuring the uncertainty of the prediction; P k-1|k-1 is the covariance matrix of the state estimate value at the previous time; Q is the process noise covariance matrix, reflecting the influence of internal noise on state transition; K k is the Kalman gain, used to balance the contributions of the predicted value and the measured value to the final estimate value; H is the observation matrix, mapping the system state to the observation space, i.e., the relationship matrix between the sensor measurement value and the system state; z k ​Ykis the sensor measurement value at time k; R is the measurement noise covariance matrix, reflecting the size of the sensor measurement noise; I is the unit matrix.

[0119] This algorithm formula uses the state equation and observation equation of the system, combines the state estimation at the last time and the current measurement value, and performs optimal estimation on the current system state, effectively removes the noise interference in the household appliance operation parameters (such as temperature, voltage, and speed) collected by the sensor, and improves the accuracy of the data.

[0120] Principal component analysis (PCA) algorithm:

[0121] The data matrix X is centered to obtain

[0122] The covariance matrix C is calculated

[0123] The eigenvalue decomposition of the covariance matrix C is performed, C = UΛU T , where U is the eigenvector matrix, and Λ is the eigenvalue diagonal matrix; the eigenvectors corresponding to the first k largest eigenvalues are selected to form the transformation matrix W = [u1, u2,..., uk]; the dimension-reduced data matrix Y is k

[0124] X: original household appliance operation parameter data matrix, each row represents a sample (such as multiple parameter values measured at one time), and each column represents a feature (such as temperature, voltage, and different parameters); n: sample number; p: feature number; The centered data matrix is centered to make the mean of the data zero, facilitating subsequent calculation; C: covariance matrix, used to measure the correlation between different features; U: eigenvector matrix, whose column vectors are the eigenvectors of the covariance matrix C; Λ: eigenvalue diagonal matrix, the elements on the diagonal are the eigenvalues of the covariance matrix C; ui: the ith eigenvector; W: transformation matrix, composed of the eigenvectors corresponding to the first k largest eigenvalues; Y: dimension-reduced data matrix, with a dimension of n × k, retaining the main information of the original data and reducing data redundancy. i

[0125] This algorithm formula reduces the dimension of the multi-dimensional household appliance operation parameter data collected by the sensor, converts high-dimensional data into low-dimensional data, while retaining the main features and information of the data as much as possible, improves the efficiency of subsequent data analysis, and facilitates the fusion and analysis of user description data.

[0126] Support vector machine (SVM) - classification in linearly separable case:

[0127] ​​For linearly separable datasets, the goal is to find a hyperplane w T x + b = 0 such that the two classes of samples can be separated correctly and the margin is maximized. The margin is determined by solving the following optimization problem:

[0128]

[0129] s.t. y i (w T x i +b) ≥ 1, i = 1,..., n

[0130] where w: the normal vector of the hyperplane, determines the direction of the hyperplane; b: the intercept of the hyperplane, determines the position of the hyperplane; x i : the feature vector of the i-th sample (the vector representation of the fused multi-source data after feature extraction); y i : the class label of the i-th sample, taking values +1 or -1, representing different fault categories (or normal / fault status); n: the number of samples; ||w||: the norm of the vector w, usually using the two-norm, representing the length of the vector.

[0131] The formula algorithm in the intelligent diagnosis decision module, by finding the optimal hyperplane, classifies the fused multi-source data, and preliminarily determines the fault category of small household appliances, providing a basis for subsequent more accurate diagnosis and maintenance scheme generation. Specific embodiment four:

[0133] As Figures 1-2 shown below is the specific application logic step description of each module and algorithm in the small household appliance diagnosis system based on AI customer service and AR imaging:

[0134] 1. AI intelligent interaction module

[0135] Interactive access: system startup, module activation, waiting for user input fault description and operation record, receiving information through chat window, voice recognition, etc.

[0136] Natural language processing: first lexical analysis, split the user input sentence into meaningful words according to the dictionary, such as "electric kettle does not heat" into "electric kettle" and "does not heat"; then syntactic analysis, clarify the grammatical relationship between words, judge the subject-predicate-object structure; finally semantic understanding, convert natural language into machine-understandable text features, such as converting "electric kettle does not heat" into feature data representing the fault phenomenon.

[0137] Sentiment analysis: use deep learning model, input processed text features, model analyzes sentiment tendency in text, judges user emotions such as anxious, calm, etc.

[0138] Response generation and feedback: According to natural language and sentiment analysis results, give targeted response. If the user is anxious, first calm down and then guide to supplement information. At the same time, the processed user text feature data is transmitted to the multi-source data fusion module.

[0139] 2. Multi-source data fusion module

[0140] Data collection: Receive user text feature data from AI intelligent interaction module, and collect household appliance operation parameters such as temperature, voltage, heating element working time of electric kettle, etc. with sensors.

[0141] Data processing: First, denoise the data collected by the sensor to remove abnormal data caused by environmental interference; then, through principal component analysis, filter out key data features and remove redundant information to simplify the data structure.

[0142] Data fusion and transmission: Integrate the processed user data and operation parameters to form fusion data and transmit it to the intelligent diagnosis and decision module.

[0143] 3. Intelligent diagnosis and decision module

[0144] Data reception and preparation: Receive fusion data from multi-source data fusion module, arrange data format, and prepare for diagnosis analysis.

[0145] Fault area positioning: Based on deep learning algorithm, combined with the imaging information of household appliance internal provided by AR visualization presentation module, analyze fusion data. First, use support vector machine to classify data and preliminarily judge fault category; then use Bayesian classification algorithm to correct and optimize classification results according to fault prior probability and data feature conditional probability, and determine fault area.

[0146] Fault area labeling: Feedback the determined fault area information to the AR visualization presentation module for labeling in AR imaging.

[0147] Maintenance scheme generation: According to the fault diagnosis result, match the corresponding maintenance scheme from the fault mode library, including maintenance steps, required tools and accessory information. Transmit the diagnosis result and maintenance scheme data to the system core control unit, and feedback the diagnosis result to the fault prediction and prevention module.

[0148] 4. AR visualization presentation module

[0149] Image acquisition: Use user's mobile phone camera to take pictures of household appliance appearance.

[0150] AR imaging generation: Combine the information provided by the multi-source data fusion module to generate three-dimensional imaging of the internal structure of the household appliance using AR technology.

[0151] Fault area marking: Receives fault area information from the intelligent diagnosis and decision-making module and marks the fault location with eye-catching logos on the AR image for users to view intuitively.

[0152] 5. Fault prediction and prevention module

[0153] Data collection and preparation: Collect historical appliance failure data, current operating parameters, and environmental data, such as ambient temperature and humidity. Also receive diagnostic results from the intelligent diagnosis and decision-making module to optimize the prediction model.

[0154] Fault prediction: Utilize long-short-term memory networks to analyze time series information in historical and current data and capture long-term data change trends. For data with small amounts and high uncertainty, use a gray prediction model to predict the time and type of future home appliance failures.

[0155] Preventive measures generation: Based on the fault prediction results, preventive measures suggestions are generated, such as regular maintenance items, early replacement of wearing parts, etc., and transmitted to the system core control unit.

[0156] 6. User Feedback Learning Module

[0157] Feedback collection: Collect user feedback on diagnostic results and maintenance plans through system interface, evaluation questionnaires, etc.

[0158] Algorithm fusion optimization: Using reinforcement learning, we adjust the behavioral strategies of the AI ​​intelligent interaction module and the intelligent diagnosis and decision-making module based on user satisfaction feedback; using online learning, we update model parameters in real time to adapt to new fault conditions and user needs.

[0159] Data transmission and update: The data required for optimization is transmitted to the AI ​​intelligent interaction module and the intelligent diagnosis and decision-making module, and the fault mode library and diagnosis algorithm are updated.

[0160] 7. Equipment linkage coordination module

[0161] Command reception: Receive fault and equipment linkage commands from the intelligent diagnosis and decision module.

[0162] Linkage solution search: Use genetic algorithms to simulate the biological genetic evolution process, try different equipment linkage combinations, and find the optimal solution.

[0163] Equipment control rule formulation: Based on the fault type, severity and real-time status of related equipment, fuzzy control algorithms are used to formulate equipment control rules and determine equipment adjustment parameters.

[0164] Equipment linkage execution and feedback: According to the control rules, coordinate related equipment to adjust the operating status, avoid the expansion of faults or provide temporary alternatives, and feedback the execution results to the system core control unit.

[0165] 8. System core regulation unit

[0166] Module coordination: When the system starts, activate and coordinate the orderly operation of each module. During operation, generate control signals based on the data of each module to ensure continuous monitoring and improvement of the system.

[0167] Comprehensive diagnostic report generation: Use the analytic hierarchy process to determine the importance of diagnostic results, repair plans, fault predictions, and device interactions in the report. Use association rule algorithms to mine potential relationships between these information. Integrate the information to generate a comprehensive diagnostic report and output it to the user.

[0168] Data storage and system optimization: Store report data for subsequent analysis. Use adaptive resonance theory to dynamically adjust the parameters and model structure of multi-source data fusion, intelligent diagnostic decision-making, and fault prediction and prevention modules. Use particle swarm optimization algorithms to optimize the control strategy and improve system performance. Specific embodiment five:

[0170] As shown in Figures 1-2 the following is a detailed hardware composition and hardware description of each module in embodiment one:

[0171] AI intelligent interaction module

[0172] Hardware composition:

[0173] Smart terminal device: Choose a smartphone, tablet computer, or dedicated intelligent diagnostic device. These devices are equipped with high-resolution display screens, input boxes, voice input buttons, and conversation record display areas on the screen to facilitate user input of fault descriptions and receipt of system responses. The device has a built-in microphone for voice recognition, converting user voice commands into text information; it also has a speaker for playing system-generated voice replies.

[0174] Communication module: Built-in Wi-Fi module for wireless communication with other modules in the management main system device or server, ensuring that user input data is transmitted to the multi-source data fusion module in a timely manner, while receiving feedback information from the system.

[0175] Hardware description: The smart terminal device is the only entry point for users to interact with the system, and its portability and multifunctionality are key to smooth communication with users. Through the display screen and input devices, users can easily communicate with the system; the communication module ensures stable data transmission, allowing the system to respond to user needs in real time.

[0176] 2. Multi-source data fusion module

[0177] Hardware composition:

[0178] Sensor group: Contains temperature sensor, voltage sensor, and speed sensor. The temperature sensor is a thermistor sensor, installed near the heating components of small household appliances, such as the heating element of an electric kettle or the motor part, to monitor the temperature changes in real time when the appliance is working. The voltage sensor is a resistance voltage divider sensor, used to measure the working voltage of the appliance to ensure it operates within the normal voltage range. The speed sensor is a Hall effect sensor, used to monitor the speed of rotating parts such as fan motors or stirrer motors.

[0179] Data acquisition card: Responsible for collecting analog signals output by various sensors and converting them into digital signals for subsequent processing. The data acquisition card has 8 analog input channels, can connect multiple sensors simultaneously, and has a sampling precision of 16 bits and a sampling rate of 10 kHz, ensuring accurate and fast acquisition of sensor data.

[0180] Edge computing device: Raspberry Pi 4B is selected as the edge computing device to perform preliminary processing and preprocessing of the collected data near the data source. It runs data processing algorithms, performs noise reduction and filtering operations, removes noise interference in sensor data, and selects key data features through principal component analysis to reduce data redundancy. The processed data is then transmitted to the intelligent diagnosis and decision-making module through the network.

[0181] Hardware description: The sensor group is responsible for acquiring various parameters of the appliance's operation, which is the core component for the system to understand the working state of the appliance. The data acquisition card realizes the conversion from analog signals to digital signals, ensuring that the data can be processed by the computer. The edge computing device performs preliminary processing of the data locally, reducing the computational burden of the subsequent modules and improving data processing efficiency and system response speed.

[0182] 3. Intelligent diagnosis and decision-making module

[0183] Hardware composition:

[0184] High-performance computer: Equipped with Intel Xeon Platinum 8380 processor, NVIDIA A100 GPU, and 128GB memory, with powerful computing capacity to run deep learning algorithms and various diagnostic and decision-making models. The computer is equipped with a 10Gbps high-speed network interface to ensure fast reception of fusion data transmitted by the multi-source data fusion module, and transmission of diagnostic results and maintenance solutions to the system core control unit and the fault prediction and prevention module.

[0185] Storage device: Samsung 980 PRO solid state drive is used to store common fault mode library, trained deep learning model and other data related to diagnostic decision. The fault mode library stores a large number of common fault cases of different types of small household appliances and their corresponding feature data and solutions for matching use during diagnosis; the trained model is based on a large number of historical data training, used for analysis and diagnosis of fusion data.

[0186] Hardware description: high-performance computer is the core of intelligent diagnostic decision, its powerful computing capacity ensures the efficient operation of complex deep learning algorithm and diagnostic model; storage device provides data support for diagnostic process, so that the system can quickly retrieve and match fault cases, accurately generate diagnostic results and repair schemes.

[0187] 4. AR visualization presentation module

[0188] Hardware composition:

[0189] Augmented reality display device: choose a smart phone with AR function. The phone is equipped with a high-resolution camera for taking pictures of household appliances, and built-in gyroscopes, accelerometers and other sensors for real-time tracking of the position and attitude changes of the device, ensuring that AR imaging can accurately match and overlay display with household appliances in the real scene. The phone screen is used to display AR imaging.

[0190] Graphics processing unit (GPU): Qualcomm Adreno 660 GPU built-in the phone, used to accelerate the rendering and generation process of AR imaging. The powerful graphics processing capability of GPU processes a large amount of three-dimensional model data and image information in real time, making AR imaging more smooth and realistic, clearly showing the internal structure and fault area of household appliances to users.

[0191] Hardware description: augmented reality display device is the tool for users to intuitively obtain the internal AR imaging of household appliances, its display effect and interactive performance directly affect the user experience; GPU is the key hardware to realize high-quality AR imaging, ensuring the smoothness and accuracy of imaging, helping users more accurately understand the fault location and internal structure of household appliances.

[0192] 5. Fault prediction and prevention module

[0193] Hardware composition:

[0194] Data storage device: uses Western Digital Purple disk to store historical fault data, long-term running parameters and environmental data of household appliances. These data are the basis for fault prediction, through long-term accumulation and analysis, the rules and trends of household appliance failure are mined.

[0195] Computing device: Shares computing resources of high-performance computers with the intelligent diagnostic decision module. Runs fault prediction algorithms such as long short-term memory networks and grey prediction models to analyze and process stored data. The computing power of high-performance computers can handle complex time series analysis and model operations to ensure accurate prediction of future fault time and type of home appliances.

[0196] Hardware description: The data storage device stores a large amount of historical data, providing a rich information source for fault prediction; the computing device uses advanced algorithms to analyze these data, predict appliance failures, and take preventive measures to ensure stable operation of appliances.

[0197] 6. User feedback learning module

[0198] Hardware composition:

[0199] User terminal device: The same as the smartphone used by the AI intelligent interaction module. Through feedback entry such as evaluation questionnaire and feedback button set on the system interface, users input satisfaction evaluation and specific feedback opinions on the diagnosis results and maintenance solutions. The Wi-Fi communication module of the mobile phone transmits user feedback data to the management main system for processing.

[0200] Server: Choose Dell PowerEdge R740xd server, used to receive and store user feedback data, and run reinforcement learning and online learning algorithms. The server is equipped with Intel Xeon Silver 4210R processor and 64GB memory, with certain computing power, and can optimize and adjust the related models and strategies of the AI intelligent interaction module and the intelligent diagnostic decision module according to user feedback data.

[0201] Hardware description: User terminal device provides feedback channel for users, facilitating user participation in system optimization process; server is responsible for collecting, storing and processing feedback data, using algorithms to continuously improve the system, improving the diagnosis accuracy and service quality of the system.

[0202] 7. Equipment linkage coordination module

[0203] Hardware composition:

[0204] Smart gateway: Choose Xiaomi smart multi-mode gateway as the hub connecting small appliances and smart home systems. The gateway has Wi-Fi, Bluetooth, ZigBee communication interfaces, and can communicate with different types of small appliances and smart home devices. It receives device linkage instructions from the intelligent diagnostic decision module, converts the instructions into corresponding communication protocols, and sends them to related devices to realize linkage control between devices.

[0205] Actuator: Install a relay as an actuator on small appliances and smart home devices that require linkage. The actuator controls the running state of the device according to the instructions sent by the smart gateway, such as turning on or off the appliance, adjusting the working parameters of the device, etc. For example, when the air conditioner fails, the smart gateway controls the air purifier to adjust the indoor temperature and humidity through the relay.

[0206] Hardware Description: The smart gateway realizes communication and instruction transmission between different devices, which is the key to device linkage; the actuator is the hardware that specifically executes the device linkage operation, ensuring that the relevant devices adjust the running state according to the instructions, avoiding the expansion of faults or providing temporary alternative solutions.

[0207] 8. System Core Control Unit

[0208] Hardware Composition:

[0209] Central Server: Huawei FusionServer Pro2488HV5 server is selected as the core control device of the entire management main system. The server is equipped with two Intel Xeon Platinum8260 processors, 512GB memory and 25Gbps high-speed network interface. It is responsible for coordinating the operation between modules, receiving data from each module, and generating control signals according to the system operation logic and sending them to the corresponding module to ensure the stable operation and cyclic monitoring of the system.

[0210] Data Storage Array: Huawei OceanStor5310V5 storage array is used to build a large-capacity and high-reliability data storage array using redundant array of independent disks (RAID5) technology. It is used to store various data generated during system operation, including comprehensive diagnostic reports, intermediate data of each module, and historical data required for system optimization, etc. The storage array has data backup and recovery functions to ensure data security and integrity.

[0211] Hardware Description: The central server is the "brain" of the system, responsible for coordinating the collaborative work of each module to ensure the orderly operation of the system; the data storage array provides reliable data storage protection for the system, enabling the system to analyze and utilize historical data, continuously optimize the diagnostic process and improve service quality.

[0212] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other inventors can develop based on the same general inventive concepts embodied by the described embodiments. That is, although the present application is described in terms of particular embodiments and implementations, it is to be understood that the terminology used is for the purpose of descriptive clarity and that it is intended to be limited only by the words of the patent claims. A reference to an aspect of the present application employing a particular aspect, feature or structure of the described embodiments is not to be interpreted as an indication that all or even any aspects of the present application have such feature or structure in some way.

[0213] While the embodiments of the application have been shown and described herein, it will be understood by those skilled in the art that many changes, modifications, substitutions and alterations to these embodiments can be made without departing from the spirit and scope of the present application, which is defined by the appended claims and their equivalents.

Claims

1. A small appliance diagnostic system based on AI customer service and AR imaging, featuring: It includes a main management system, which includes an AI intelligent interaction module, an AR visualization presentation module, a multi-source data fusion module, an intelligent diagnosis and decision-making module, a fault prediction and prevention module, a user feedback learning module, an equipment linkage coordination module and a system core control unit; The overall operation steps of the management main system are as follows: Sp1. System initialization and connection establishment: Start the management main system to activate the AI ​​intelligent interaction module, initialize the AR visualization module to prepare for device imaging display, start the system core control unit to coordinate the operation of each module, and establish a user-side connection interface to receive user fault feedback and operation instructions; Sp2. Multi-source data collection and preliminary fusion: The AI ​​intelligent interaction module receives the user's description of the small household appliance fault and the operation record data. The multi-source data fusion module uses sensors to collect the household appliance operating parameters, including temperature, voltage, and speed. It preliminarily integrates the user description data with the operating parameters and transmits them to the intelligent diagnosis and decision module. Sp3, AR imaging real-time display and fault area location: The AR visualization module uses the user's mobile phone camera to obtain the appearance image of the home appliance. Combined with the information provided by the multi-source data fusion module, it generates an AR image of the internal structure of the home appliance in real time. The intelligent diagnosis and decision module analyzes the data to determine the fault area, and the AR visualization module accurately marks the fault area on the image. Sp4, Intelligent Diagnosis and Solution Generation: The intelligent diagnosis and decision-making module uses a deep learning algorithm to analyze and integrate multi-source data, match common fault mode libraries, and generate fault diagnosis results and corresponding unique maintenance solutions, which include maintenance steps, required tools, and accessories. Sp5. Fault prediction and prevention analysis: The fault prediction and prevention module uses big data analysis and machine learning algorithms to predict the time and type of future home appliance failures based on historical fault data, current operating parameters and environmental factors, and generates preventive measures in advance, which are transmitted to the system's core control unit; Sp6, User Feedback and Learning Optimization: User feedback on diagnostic results and maintenance solutions is collected through the user feedback learning module. The AI ​​intelligent interaction module updates the fault mode library and diagnostic algorithm based on the feedback information, improving diagnostic accuracy and solution effectiveness, and the optimization results are transmitted to the intelligent diagnosis decision module. Sp7, Device Linkage and Collaborative Processing: When a fault involves multiple small appliances and is associated with the smart home system, the device linkage coordination module, based on the instructions of the intelligent diagnosis and decision module, coordinates the relevant devices to adjust their operating status, prevent the fault from escalating, and provide temporary alternatives. For example, if an air conditioner fails, it can be linked to an air purifier to adjust the indoor temperature and humidity. Sp8, Comprehensive diagnostic report generation and output: The system core control unit integrates diagnostic results, maintenance plans, fault prediction information and equipment linkage status to generate a detailed comprehensive diagnostic report, which is output to the user through the user terminal connection interface and the report data is stored for subsequent analysis and system optimization; Sp9, cyclic monitoring and continuous improvement: The multi-source data fusion module continuously collects household appliance operating parameters, the intelligent diagnosis and decision-making module analyzes data in real time, the fault prediction and prevention module dynamically adjusts the prediction model, and the system core control unit drives the system to operate cyclically, continuously optimizing the diagnosis process and improving service quality; Among them, the main management system runs step by step through the operation steps to achieve accurate diagnosis of small household appliance faults, efficient maintenance guidance, and intelligent management and optimization of the entire life cycle.

2. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1 is characterized by: The AI ​​intelligent interaction module adopts a fusion mechanism of natural language processing and sentiment analysis algorithms. The natural language processing algorithm converts the user's natural language into machine-understandable text features through lexical analysis, syntactic analysis and semantic understanding; the sentiment analysis algorithm relies on the combination of deep learning recurrent neural network and convolutional neural network architecture to judge the emotional tendency of user text.

3. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1 is characterized by: The multi-source data fusion module uses the Kalman filter algorithm and the principal component analysis algorithm to combine parameters such as square and speed to implement real-time filtering, remove noise interference, and improve data accuracy; the principal component analysis algorithm performs dimensionality reduction processing on the filtered data, extracts the main characteristic components, reduces data redundancy, and improves the efficiency of subsequent data analysis.

4. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: In the intelligent diagnosis and decision-making module, fault diagnosis results are generated using a support vector machine and a Bayesian classification algorithm. The support vector machine constructs an optimal classification hyperplane, classifies the fused multi-source data, and preliminarily determines the fault category. The Bayesian classification algorithm corrects and optimizes the classification results of the support vector machine based on the prior probability of fault occurrence and the conditional probability of data features, thereby improving the accuracy of fault diagnosis and generating a corresponding unique maintenance plan.

5. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: The fault prediction and prevention module uses the fusion mechanism of long-short-term memory network and grey prediction model to predict the time and type of future household appliance failures. The long-short-term memory network processes the time series information of historical fault data and current operating parameters to capture the long-term dependencies of the data. The grey prediction model accumulates and generates raw data to establish a grey differential equation model, predict the failure trend of household appliances, and generate preventive measures in advance.

6. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: The user feedback learning module uses a fusion mechanism of reinforcement learning and online learning algorithms to update the fault mode library and diagnosis algorithm. The reinforcement learning algorithm optimizes the behavior strategies of the AI ​​intelligent interaction module and the intelligent diagnosis and decision module based on user feedback on the satisfaction of the diagnosis results and maintenance plans by setting up a reward mechanism. The online learning algorithm updates the model parameters in real time during the continuous feedback process of users.

7. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: When coordinating the adjustment of the operating status of related equipment, the equipment linkage coordination module uses a combination of genetic algorithms and fuzzy control algorithms to search for the optimal equipment linkage solution through the genetic algorithm. By simulating the biological genetic evolution process, different equipment linkage combinations are selected, crossed, and mutated to optimize the equipment linkage effect. The fuzzy control algorithm formulates specific equipment control rules based on fuzzy information such as the fault type, severity, and the real-time status of related equipment to achieve precise and flexible regulation of the equipment, avoid the expansion of the fault, and provide effective temporary alternative solutions.

8. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: When generating a comprehensive diagnostic report, the system core control unit uses a hierarchical analysis method (AHP) and an association rule algorithm in data mining to integrate them. The AHP determines the weights of factors such as diagnostic results, maintenance plans, fault prediction information, and equipment linkage status in the comprehensive diagnostic report to highlight key information. The association rule algorithm mines the potential relationships between these factors, including the associations between different fault types and common maintenance tools, and incorporates this associated information into the comprehensive diagnostic report, which is output to the user through the user-side connection interface, and the report data is stored for subsequent analysis and system optimization.

9. The small household appliance diagnostic system based on AI customer service and AR imaging according to claim 1, characterized in that: The main management system dynamically adjusts the parameters and model structures of the multi-source data fusion module, the intelligent diagnosis and decision module, and the fault prediction and prevention module through adaptive resonance theory, so that it can adapt to changes in the operating environment and fault modes of home appliances; the particle swarm optimization algorithm optimizes the control strategy of the system's core control unit, and by simulating the foraging behavior of bird flocks, it finds the optimal system operating parameter configuration to continuously optimize the diagnosis process and improve service quality.

Images