Household potential safety hazard intelligent detection and early warning system based on AI

Abstract

The invention, which relates to the technical field of household hidden danger detection, discloses an AI-based household hidden danger intelligent detection and early warning system comprising an early warning center, a data acquisition module, a data processing module, an AI analysis module and an execution module. Operation data of each household device in a household environment is acquired through a data acquisition module, and data cleaning, numerical value normalization processing and feature interval screening are performed on the operation data of each household device through a data processing module, so that principal component feature data corresponding to each household device is obtained; the AI analysis module constructs an AI analysis model for carrying out hidden danger analysis on the household equipment, inputs principal component feature data of each household equipment to the AI analysis model to obtain hidden danger items and hidden danger coefficients of the household equipment, delimits hidden danger levels of the household equipment according to the hidden danger coefficients, and finally, determines the hidden danger levels of the household equipment according to the hidden danger coefficients. And the execution module adopts corresponding execution operation according to the hidden danger level of the household equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of home hazard detection, and specifically to an intelligent detection and early warning system for home safety hazards based on AI. Background Art

[0002] With the popularization of smart homes, the types and quantities of home devices have been increasing continuously, and the real-time monitoring and early warning of potential safety hazards in their operation have become an important topic. Most existing home safety systems rely on manual inspections or single-sensor threshold alarms, and have the following deficiencies:

[0003] 1. Low data processing efficiency: Traditional methods are difficult to handle the cleaning, normalization, and feature extraction of massive real-time data, resulting in lagged analysis.

[0004] 2. Insufficient accuracy in hazard identification: Relying on static rules or simple models, it is unable to dynamically capture the abnormal change trends of device voltage, power, and temperature.

[0005] 3. Single response mechanism: Lack of hierarchical early warning and intelligent execution strategies. High-risk hazards cannot be quickly blocked, and low-risk hazards are easily ignored.

[0006] In view of the above problems, there is an urgent need for an intelligent detection and early warning system for home safety hazards based on AI to improve the real-time performance, accuracy, and execution efficiency of home hazard detection. Summary of the Invention

[0007] In order to solve the above problems, the purpose of the present invention is to provide an intelligent detection and early warning system for home safety hazards based on AI.

[0008] The purpose of the present invention can be achieved through the following technical solutions: An intelligent detection and early warning system for home safety hazards based on AI, including an early warning center, which is communicatively connected to a data acquisition module, a data processing module, an AI analysis module, and an execution module;

[0009] The data acquisition module is used to collect the operation data of each home device in the home environment;

[0010] The data processing module is used to perform data cleaning, numerical normalization processing, and feature interval screening on the operation data of each home device, and then obtain the principal component feature data corresponding to each home device;

[0011] The AI analysis module constructs an AI analysis model for hazard analysis of home devices, inputs the principal component feature data of each home device into the AI analysis model, obtains the hazard matters and hazard coefficients corresponding to the home devices, and determines the hazard levels of the home devices according to the hazard coefficients;

[0012] The execution module is used to take corresponding execution operations according to the hazard levels of the home devices.

[0013] Further, the process of collecting the operation data of each household device in the home environment includes:

[0014] Number a number of household devices and denote them as i, where i = 1, 2, 3, ……, n, and n is a natural number greater than 0. Build a data buffer for each household device, and each household device stores its own operation data in real time after collection in the data buffer;

[0015] Create a data packaging node in the data buffer of each household device;

[0016] The data packaging node packs the operation data stored in the data buffer, sets the amount of packaged data, and records in real time the data stock of the operation data collected in the data buffer. When the data stock reaches the amount of packaged data, pack the corresponding operation data to generate a data subset, label the corresponding collection period for the data subset, set the data collection period corresponding to each household device, and upload all the data subsets generated by each household device during the data collection period to the data processing module.

[0017] Further, the process of data cleaning and numerical normalization of the operation data of each household device includes:

[0018] Set the computing power ratio period, cleaning period, and normalization period;

[0019] During the computing power ratio period, according to the number of data subsets uploaded by each household device to the data processing module, retrieve computing power resources and ratio the computing power for processing the operation data corresponding to each household device;

[0020] During the cleaning period, according to the computing power resources allocated to each household device, start the pre-built data cleaning program, and sequentially clean the data subsets of each household device, thereby correcting the error data in the data subsets, removing the redundant data in the data subsets, and filling the missing data in the data subsets, and then constructing several standardized data subsets corresponding to each household device at each collection period;

[0021] During the normalization period, perform numerical normalization on several standardized data subsets of each household device to obtain the normalization coefficients of the real-time operation data of each household device at different collection periods; specifically including the normalization coefficients of the device voltage, device power, and device temperature respectively.

[0022] Further, the process of performing feature interval screening to obtain the principal component feature data corresponding to each household device includes:

[0023] Set the first screening interval, the second screening interval, and the third screening interval;

[0024] The first screening range is used to screen the effective voltage to be analyzed;

[0025] The second screening range is used to screen the effective power to be analyzed;

[0026] The third screening range is used to screen the effective temperature to be analyzed;

[0027] Integrate the effective voltage to be analyzed, the effective power to be analyzed, and the effective temperature to be analyzed of each household device, and then obtain the principal component feature data corresponding to each household device.

[0028] Furthermore, the process of constructing an AI analysis model for hazard analysis of household devices includes:

[0029] Obtain the operation data of household devices with a certain amount of data at historical time nodes. The operation data of household devices at historical time nodes is used to record the operation status of household devices, including household devices with normal operation status and their corresponding characteristic component data, and also including household devices with abnormal operation status and their corresponding characteristic component data;

[0030] Construct an initial AI analysis model according to AI technology, input the operation data of household devices at historical time nodes into the initial AI analysis model, analyze the characteristic component data of each household device, and output the operation status of each household device;

[0031] When the operation status output by the initial AI analysis model is consistent with the actual operation status of the household device, record a correct analysis behavior, otherwise, record an incorrect analysis behavior;

[0032] Obtain the model prediction accuracy of the initial AI analysis model, denoted as Sc;

[0033]

[0034] Set an iteration threshold, denoted as Dd;

[0035] When Sc < Dd, perform model iteration on the initial AI analysis model, and obtain the model prediction accuracy of the AI analysis model after each model iteration and judge the numerical size with the iteration threshold;

[0036] When Sc ≥ Dd, stop model iteration and construct the final AI analysis model.

[0037] Furthermore, the process of inputting the principal component feature data of each household device into the AI analysis model to obtain the corresponding hazard matters and hazard coefficients of the household device includes:

[0038] Input the principal component feature data of each home appliance into the AI analysis model in sequence;

[0039] The AI analysis model synchronously constructs a time series prediction layer, an anomaly detection layer, and a knowledge graph mapping layer;

[0040] The time series prediction layer is used to sort the effective voltage to be analyzed, the effective power to be analyzed, and the effective temperature to be analyzed included in the principal component feature data in the order of the operation timestamps of the home appliances, and analyze the dynamic change trends of the device voltage, device power, and device temperature during the operation of the home appliances;

[0041] The anomaly detection layer is used to identify and mark the operation timestamps when the device voltage, device power, and device temperature of the home appliance deviate from the normal change trend, obtain the abnormal device operating points, and acquire the detailed operation data of the home appliance at the abnormal device operating points, analyze and obtain all potential hazard items of the home appliance, and the hazard coefficient corresponding to each potential hazard item;

[0042] The knowledge graph mapping layer is used to construct a hazard causal chain for each home appliance, and the structure of the hazard causal chain is: hazard cause - hazard item - hazard coefficient; create a mapping node for each home appliance in the constructed knowledge graph mapping layer, and map the hazard causal chain of each home appliance into its corresponding mapping node.

[0043] Further, the process of defining the hazard level of the home appliance according to the hazard coefficient includes:

[0044] Denote the hazard coefficient of the home appliance as YH;

[0045] Set a first hazard interval and a second hazard interval, denoted as Δ1 and Δ2 respectively;

[0046] When YH ∈ Δ1, define the hazard level of the corresponding home appliance as the low-risk level;

[0047] When YH ∈ Δ2, define the hazard level of the corresponding home appliance as the high-risk level.

[0048] Further, the process of taking corresponding execution operations according to the hazard level of the home appliance includes:

[0049] When the hazard level of the home appliance is the low-risk level, the AI analysis model retrieves the hazard causal chain of the home appliance in the knowledge graph mapping layer, and creates a maintenance suggestion corresponding to the current home appliance according to the hazard causal chain;

[0050] When the hazard level of the home appliance is the high-risk level, push an APP notification to the corresponding owner user of the home appliance, cut off the power supply of the corresponding home appliance, and arrange relevant maintenance personnel to perform equipment maintenance on the powered-off home appliance.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] 1. By dynamically packing and collecting data in the data buffer area, combining with the computing power ratio mechanism to optimize the allocation of computing power resources, the efficiency of data cleaning, numerical normalization processing, and feature screening is improved. Feature interval screening is adopted to eliminate redundant data and retain the principal component feature data, reducing the computational complexity of the AI analysis model and effectively improving the data processing efficiency.

[0053] 2. An AI analysis model with a multi-level relationship of a time series prediction layer, an anomaly detection layer, and a knowledge graph mapping layer is constructed to dynamically analyze the operation trend of the device, identify abnormal working condition points, and generate a hidden danger causal chain. Based on the hidden danger coefficient, the low-risk and high-risk levels of household equipment are divided to realize the quantitative assessment of hidden danger risks and effectively improve the accuracy of hidden danger identification.

[0054] 3. When in the low-risk level, targeted maintenance suggestions are automatically generated. When in the high-risk level, an APP notification is immediately pushed, and after remotely cutting off the power, relevant maintenance personnel are dispatched to form a "monitoring - analysis - response" closed loop, effectively reducing the accident rate, realizing the rapid blocking of high-risk hidden dangers in household equipment, and the daily monitoring of low-risk hidden dangers. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is the schematic diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] As Figure 1 shown, the intelligent detection and early warning system for home safety hazards based on AI includes an early warning center, and the early warning center is communicatively connected to a data collection module, a data processing module, an AI analysis module, and an execution module;

[0057] The data collection module is used to collect the operation data of each household device in the home environment;

[0058] The data processing module is used to perform data cleaning, numerical normalization processing, and feature interval screening on the operation data of each household device, and then obtain the principal component feature data corresponding to each household device;

[0059] The AI analysis module constructs an AI analysis model for analyzing hidden dangers of household devices, inputs the principal component feature data of each household device into the AI analysis model, obtains the hidden danger matters and hidden danger coefficients corresponding to the household devices, and determines the hidden danger levels of the household devices according to the hidden danger coefficients;

[0060] The execution module is used to take corresponding execution operations according to the hidden danger levels of the household devices.

[0061] It should be further noted that in the specific implementation process, the process of collecting the operation data of each household device in the home environment includes:

[0062] Number a number of household devices, and record the numbers as i, then i = 1, 2, 3,..., n, where n is a natural number greater than 0. For each household device, construct a data buffer area, and each household device stores its own operation data in real time after collection in the data buffer area;

[0063] Create a data packaging node in the data buffer area of each household device;

[0064] The data packaging node packages the operation data stored in the data buffer area, sets the amount of packaged data, and records in real time the data stock of the operation data collected in the data buffer area. When the data stock reaches the amount of packaged data, package the corresponding operation data to generate a data subset, and label the corresponding collection period for the data subset;

[0065] Set the data collection period for each household device, and upload all the data subsets generated by each household device during its data collection period to the data processing module, and the data processing module performs subsequent processing operations;

[0066] It should be noted that if, when the data collection period of the household device ends, the data stock of the remaining operation data in the data buffer area still does not reach the amount of packaged data, calculate the difference between the amount of packaged data and the data stock, and fill the null character data corresponding to the difference value into the remaining operation data to reach the amount of packaged data, and construct a data subset.

[0067] It should be further noted that in the specific implementation process, the process of data cleaning and numerical normalization processing of the operation data of each household device includes:

[0068] Set the computing power ratio period, cleaning period, and normalization period;

[0069] During the computing power ratio period, according to the number of data subsets uploaded by each household device to the data processing module, retrieve computing power resources and allocate the computing power size for processing the operation data of each household device;

[0070] Record the number of data subsets of the operation data corresponding to the household device numbered i as S[i], set the unit computing power, and record the value of the unit computing power as cal. The unit computing power is used to process a data subset, and record the computing power size of the total operation computing power for processing the household device numbered i as Cal[i]. Then the calculation formula of Cal[i] is as follows:

[0071] Cal[i] = S[i] × cal;

[0072] During the cleaning period, according to the computing power resources allocated to each household device, start the pre-built data cleaning program, and sequentially perform data cleaning on several data subsets of each household device, so as to correct the incorrect data in the data subsets, eliminate the redundant data in the data subsets, and fill in the missing data in the data subsets;

[0073] Furthermore, construct several corresponding standardized data subsets for each household device at each acquisition period;

[0074] During the normalization period, perform numerical normalization on several standardized data subsets of each household device, so as to obtain the normalization coefficients corresponding to the real-time operation data of each household device at different acquisition periods;

[0075] The specific process of data normalization is as follows:

[0076] For a certain household device, record the instantaneous operation data at a certain acquisition moment corresponding to the standardized data subset at each acquisition period as D[t], where t is a certain acquisition moment in the acquisition period. Among them, the instantaneous operation data corresponding to the household device at each acquisition moment includes the device instantaneous voltage, the device instantaneous power, and the device instantaneous temperature;

[0077] The instantaneous operation data D[t] is expressed as follows:

[0078] D[t] = [V[t] - s, P[t] - s, C[t] - s];

[0079] Among them, V[t] - s, P[t] - s, and C[t] - s are the device instantaneous voltage, the device instantaneous power, and the device instantaneous temperature corresponding to the household device at the t acquisition moment, respectively;

[0080] Taking a certain household device as an example, sequentially perform numerical normalization on the instantaneous operation data at each acquisition moment in each acquisition period. Specifically: obtain the voltage extreme values, power extreme values, and temperature extreme values of the household device at each acquisition period, and combine the device instantaneous voltage, the device instantaneous power, and the device instantaneous temperature at each acquisition moment in each acquisition period to calculate the normalization coefficients of the operation data corresponding to the current household device at each acquisition period;

[0081] Among them, the voltage extreme values include the maximum voltage and the minimum voltage, which are denoted as V-max and V-min respectively;

[0082] The power extreme values include the maximum power and the minimum power, which are denoted as P-max and P-min respectively;

[0083] The temperature extreme values include the highest temperature and the lowest temperature, which are denoted as C-max and C-min respectively;

[0084] For the home appliance device numbered i, the normalization coefficients of the device voltage, device power, and device temperature corresponding to the instantaneous operation data at each acquisition moment during a certain acquisition period are denoted as follows:

[0085] The normalization coefficient of the device voltage is denoted as α[i];

[0086] The normalization coefficient of the device power is denoted as β[i];

[0087] The normalization coefficient of the device temperature is denoted as γ[i];

[0088] α[i] = (V[t]-s - V-min) / (V-max - V-min);

[0089] β[i] = (P[t]-s - P-min) / (P-max - P-min);

[0090] γ[i] = (C[t]-s - C-min) / (C-max - C-min).

[0091] It should be further noted that in the specific implementation process, the process of performing feature interval screening to obtain the principal component feature data corresponding to each home appliance device includes:

[0092] Set the first screening interval, the second screening interval, and the third screening interval;

[0093] The first screening interval is used to screen the effective voltage to be analyzed;

[0094] The second screening interval is used to screen the effective power to be analyzed;

[0095] The third screening interval is used to screen the effective temperature to be analyzed;

[0096] The first screening interval, the second screening interval, and the third screening interval are respectively denoted as Ω1, Ω2, and Ω3, where Ω1 = [0.2, 0.8], Ω2 = [0.3, 0.75], and Ω3 = [0.4, 0.9];

[0097] All device voltages with α[i] ∈ Ω1 are screened as the effective voltage to be analyzed for the home appliance device, and all device voltages outside this range are excluded;

[0098] All device powers with β[i] ∈ Ω2 are screened as the effective power to be analyzed for the home appliance device, and all device powers outside this range are excluded;

[0099] All device temperatures with γ[i] ∈ Ω3 are screened as the effective temperature to be analyzed for the home appliance device, and Eliminate all the device temperatures;

[0100] Integrate the effective voltage to be analyzed, the effective power to be analyzed, and the effective temperature to be analyzed obtained by each home device, and then obtain the principal component feature data corresponding to each home device.

[0101] It should be further noted that in the specific implementation process, the process of constructing an AI analysis model for hazard analysis of home devices includes:

[0102] Obtain the operation data of home devices corresponding to a number of data volumes at historical time nodes. The operation data of home devices corresponding to a number of data volumes at historical time nodes is used to record the operation status of home devices, including home devices with normal operation status and their corresponding characteristic component data, and also including home devices with abnormal operation status and their corresponding characteristic component data;

[0103] Construct an initial AI analysis model according to AI technology;

[0104] The AI analysis model is used to conduct hazard analysis on home devices;

[0105] Input the operation data of home devices at historical time nodes into the initial AI analysis model. The initial AI analysis model analyzes the characteristic component data of each home device, and then outputs the operation status of each home device. When the operation status output by the initial AI analysis model is consistent with the actual operation status of the home device, record a correct analysis behavior; otherwise, record an incorrect analysis behavior;

[0106] Obtain the model prediction accuracy of the initial AI analysis model, and denote the model prediction accuracy as Sc;

[0107]

[0108] Set an iteration threshold, and denote the iteration threshold as Dd;

[0109] When Sc < Dd, set a training set, a test set, and a validation set, perform model iteration on the initial AI analysis model, and obtain the model prediction accuracy of the AI analysis model after each model iteration. Compare the model prediction accuracy with the iteration threshold;

[0110] When Sc ≥ Dd, stop model iteration and construct the final AI analysis model.

[0111] It should be further noted that in the specific implementation process, the process of inputting the principal component feature data of each home device into the AI analysis model to obtain the hazard matters and hazard coefficients corresponding to the home device includes:

[0112] Input the principal component feature data of each household device into the AI analysis model in sequence;

[0113] The AI analysis model synchronously constructs a time series prediction layer, an anomaly detection layer, and a knowledge graph mapping layer;

[0114] The time series prediction layer is used to sort the effective voltage to be analyzed, the effective power to be analyzed, and the effective temperature to be analyzed included in the principal component feature data in the order of the operating timestamps of the household device, and analyze the dynamic change trends of the device voltage, device power, and device temperature during the operation of the household device;

[0115] The anomaly detection layer is used to identify and mark the operating timestamps when the device voltage, device power, and device temperature of the household device deviate from the normal change trends, thereby obtaining the abnormal device operating points, and obtaining the detailed operation data of the household device corresponding to the abnormal device operating points;

[0116] Based on the detailed operation data of the household device at the abnormal device operating points, analyze and obtain all potential hazard items of the household device, and the corresponding hazard coefficients for each potential hazard item;

[0117] The knowledge graph mapping layer is used to construct a potential hazard causal chain for each household device, and the structure of the potential hazard causal chain is: potential hazard cause - potential hazard item - hazard coefficient; create a mapping node for each household device in the constructed knowledge graph mapping layer, and map the potential hazard causal chain of each household device into its corresponding mapping node.

[0118] Among them, the potential hazard cause is the specific accident cause corresponding to the abnormality of the household device, the potential hazard item is the specific accident type corresponding to the abnormality of the household device, and the hazard coefficient represents the probability corresponding to various specific accident types of the household device.

[0119] It should be further noted that in the specific implementation process, the process of defining the potential hazard level of the household device includes:

[0120] Record the hazard coefficient of the household device as YH, and the value of YH is a real number in the interval (0.4, 1);

[0121] Set the first potential hazard interval and the second potential hazard interval;

[0122] Record the first potential hazard interval and the second potential hazard interval as Δ1 and Δ2 respectively;

[0123] Among them, Δ1 = (0.4, 0.7], Δ2 = (0.7, 1);

[0124] When YH ∈ Δ1, define the potential hazard level of the corresponding household device as the low-risk level;

[0125] When YH ∈ Δ2, the hidden danger level corresponding to the home appliance is defined as the high-risk level.

[0126] It should be further noted that in the specific implementation process, the process of taking corresponding execution operations according to the hidden danger level of home appliances includes:

[0127] When the hidden danger level of the home appliance is the low-risk level, the hidden danger causal chain of the home appliance in the knowledge graph mapping layer is retrieved through the AI analysis model, and the maintenance suggestions corresponding to the current home appliance are created according to the hidden danger causal chain;

[0128] Among them, the maintenance suggestions include checking the circuit load, checking the mechanical performance, checking the device function completion degree, and checking the device appearance, etc.;

[0129] When the hidden danger level of the home appliance is the high-risk level, an APP notification is pushed to the owner user corresponding to the home appliance, the corresponding home appliance is powered off, and relevant maintenance personnel are arranged to perform equipment maintenance on the powered-off home appliance.

[0130] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An AI-based intelligent detection and early warning system for home safety hazards, including an early warning center, characterized in that: The early warning center is communicatively connected to a data acquisition module, a data processing module, an AI analysis module and an execution module; The data collection module is used to collect the operation data of each household appliance in the household environment; The data processing module is used to perform data cleaning, numerical normalization and feature interval screening on the operation data of each household appliance, thereby obtaining the principal component feature data corresponding to each household appliance; The AI ​​analysis module constructs an AI analysis model for analyzing hidden dangers of household appliances, inputs the principal component feature data of each household appliance into the AI ​​analysis model, obtains hidden danger items and hidden danger coefficients corresponding to the household appliances, and defines the hidden danger level of the household appliances according to the hidden danger coefficients; The execution module is used to take corresponding execution operations according to the hidden danger level of the household equipment.

2. The AI-based home safety hazard intelligent detection and early warning system according to claim 1 is characterized in that: The process of collecting the operating data of each home appliance in the home environment includes: A number of home appliances are numbered and recorded as i, i=1, 2, 3, ..., n, where n is a natural number greater than 0, and a data cache is constructed for each home appliance, and each home appliance collects its own operating data in real time and stores it in the data cache; Create a data packaging node in the data cache area of ​​each home device; The data packaging node packages the operating data stored in the data cache area, sets the packaged data volume, and records the data inventory of the collected operating data in the data cache area in real time. When the data inventory reaches the packaged data volume, the corresponding operating data is packaged to generate a data subset, and the corresponding collection time period is marked for the data subset. The data collection cycle corresponding to each home appliance is set, and all data subsets generated by each package within the data collection cycle are uploaded to the data processing module.

3. The AI-based home safety hazard intelligent detection and early warning system according to claim 2 is characterized in that: The process of data cleaning and numerical normalization of the operating data of each home device includes: Set the computing power allocation period, cleaning period, and normalization period; During the computing power allocation period, according to the number of data subsets uploaded to the data processing module by each home appliance, computing power resources are called up to allocate the computing power of each home appliance for data processing corresponding to the operating data; During the cleaning period, according to the computing resources allocated to each home appliance, a pre-built data cleaning program is started to clean several data subsets of each home appliance in turn, thereby correcting the erroneous data in the data subset, eliminating the redundant data in the data subset, and filling the missing data in the data subset, thereby constructing several standard data subsets corresponding to each home appliance in each collection period; During the normalization period, numerical normalization processing is performed on several standard data subsets of each household device to obtain the normalization coefficient of the real-time operating data of each household device in different collection periods; specifically, the normalization coefficients of the device voltage, device power and device temperature are included.

4. The AI-based home safety hazard intelligent detection and early warning system according to claim 3 is characterized in that: The process of screening the feature intervals and obtaining the principal component feature data corresponding to each home appliance includes: Set a first screening interval, a second screening interval, and a third screening interval; The first screening interval is used to screen effective voltages to be analyzed; The second screening interval is used to screen the effective power to be analyzed; The third screening interval is used to screen effective temperatures to be analyzed; The effective voltage to be analyzed, the effective power to be analyzed, and the effective temperature to be analyzed of each household device are integrated to obtain the principal component characteristic data corresponding to each household device.

5. The AI-based home safety hazard intelligent detection and early warning system according to claim 4 is characterized in that: The process of building an AI analysis model for potential hazards analysis of home devices includes: Acquire a certain amount of home appliance operation data of the home appliance at a historical time node, where the home appliance operation data at the historical time node is used to record the operation status of the home appliance, including home appliances in a normal operation status and their corresponding characteristic component data, and also including home appliances in an abnormal operation status and their corresponding characteristic component data; Construct an initial AI analysis model based on AI technology, input the home appliance operation data at historical time nodes into the initial AI analysis model, analyze the characteristic component data of each home appliance, and output the operation status of each home appliance; When the operating status output by the initial AI analysis model is consistent with the actual operating status of the home device, a correct analysis behavior is recorded; otherwise, an incorrect analysis behavior is recorded; Obtain the model prediction accuracy of the initial AI analysis model, denoted as Sc; Set the iteration threshold and record it as Dd; When Sc<Dd, the initial AI analysis model is iterated, and the model prediction accuracy of the AI ​​analysis model after each model iteration is obtained and the value of the iteration threshold is judged; When Sc≥Dd, stop model iteration and build the final AI analysis model.

6. The AI-based intelligent detection and early warning system for home safety hazards according to claim 5 is characterized in that: The process of inputting the principal component characteristic data of each household device into the AI ​​analysis model to obtain the hidden danger items and hidden danger coefficients corresponding to the household devices includes: Input the principal component feature data of each household device into the AI ​​analysis model in turn; The AI ​​analysis model simultaneously builds the time series prediction layer, anomaly detection layer, and knowledge graph mapping layer; The time series prediction layer is used to sort the effective voltage to be analyzed, the effective power to be analyzed and the effective temperature to be analyzed included in the main component feature data according to the operation timestamp sequence of the home appliances, and analyze the dynamic change trend of the device voltage, device power and device temperature during the operation of the home appliances; The anomaly detection layer is used to identify and mark the operation timestamps of the device voltage, device power and device temperature of the home equipment that deviate from the normal change trend, obtain the abnormal device operating point, and obtain the detailed operation data of the home equipment at the abnormal device operating point, analyze and obtain all the hidden dangers of the home equipment, and the hidden danger coefficient corresponding to each hidden danger; The knowledge graph mapping layer is used to construct the hidden danger causal chain of each household device. The structure of the hidden danger causal chain is: hidden danger cause - hidden danger item - hidden danger coefficient; in the constructed knowledge graph mapping layer, a mapping node is created for each household device, and the hidden danger causal chain of each household device is mapped to its corresponding mapping node.

7. The AI-based home safety hazard intelligent detection and early warning system according to claim 6 is characterized in that: The process of determining the hazard level of household equipment based on the hazard coefficient includes: The hidden danger coefficient of household equipment is recorded as YH; Set a first hidden danger interval and a second hidden danger interval, which are denoted as Δ1 and Δ2 respectively; When YH∈Δ1, the hidden danger level of the corresponding household equipment is defined as low-risk level; When YH∈Δ2, the hidden danger level of the corresponding household equipment is defined as a high-risk level.

8. The AI-based intelligent detection and early warning system for home safety hazards according to claim 7 is characterized in that: The process of taking appropriate action based on the potential risk level of home equipment includes: When the hidden danger level of the home appliance is low-risk, the hidden danger causal chain of the home appliance in the knowledge graph mapping layer is retrieved through the AI ​​analysis model, and the corresponding maintenance suggestions for the current home appliance are created based on the hidden danger causal chain; When the hidden danger level of a home appliance is high-risk, an APP notification is pushed to the corresponding user of the home appliance, the corresponding home appliance is powered off, and relevant maintenance personnel are arranged to inspect and repair the home appliance that has been powered off.

Images