Home appliance fault diagnosis methods, devices, electronic equipment and readable storage media

By acquiring and mapping the conducted signal characteristics of home appliances through heterogeneous sensing nodes and calculating the fault confidence, the problem of being unable to identify progressive faults in existing technologies is solved, and quantitative diagnosis of the health status of home appliances and prediction of potential faults are realized.

CN123113798APending Publication Date: 2026-11-17GREE ELECTRIC APPLIANCE INC OF ZHUHAI
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611197904.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-07
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing fault diagnosis methods for smart home appliances can only identify pre-stored known fault types, and cannot identify progressive, non-preset fault types, which may cause the equipment to suddenly break down before the fault is fully manifested.

Method used

By deploying heterogeneous sensing nodes with sensing capabilities outside the target home appliances, the current transmission characteristics of the transmitted signals are obtained and mapped to the feature dimensions corresponding to standard operating conditions. The fault confidence is calculated, the health status of the home appliances is determined, and the gradual process is quantitatively characterized.

Benefits of technology

It enables the prediction of potential deterioration trends in the health status of home appliances, and can reveal abnormal evolution of equipment before the failure mode is fully manifested, thus expanding the applicable boundaries of fault diagnosis and avoiding sudden equipment damage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113798A_ABST
    Figure CN123113798A_ABST
Patent Text Reader

Abstract

This application discloses a method, apparatus, electronic device, and readable storage medium for diagnosing faults in home appliances, belonging to the field of smart home appliance technology. The method includes: acquiring the current conduction characteristics of the conducted signals generated by the target home appliance based on heterogeneous sensing nodes, and mapping the current conduction characteristics to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the conducted signals are used to characterize the operating status of the target home appliance; the heterogeneous sensing nodes consist of home appliances deployed outside the target home appliance and having sensing functions; determining the drift trend of the conduction characteristics based on the current conduction characteristics and the reference conduction characteristics of the conducted signals of the target home appliance under standard operating conditions, and calculating the fault confidence; determining the health status of the target home appliance based on the drift trend and the fault confidence; the method provided by this application can identify non-preset, gradual fault types and avoid sudden damage to home appliances.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application belongs to the field of smart home appliance technology, specifically relating to a method, device, electronic equipment, and readable storage medium for diagnosing faults in home appliances. Background Technology

[0002] With the rapid development of IoT technology and the widespread adoption of smart home appliances, timely diagnosis of malfunctions in smart home appliances is an important means of ensuring user experience.

[0003] Existing fault diagnosis methods for smart home appliances typically employ a pre-set fault database. When the operating data of a smart home appliance matches an entry in the fault database, a fault is identified in the smart home appliance.

[0004] However, the pre-set fault database is fixed, so this method can only identify known fault types that are stored in advance. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, electronic device and readable storage medium for diagnosing faults in home appliances, which can identify non-preset, progressive fault types and avoid sudden damage to home appliances.

[0006] In a first aspect, embodiments of this application provide a method for diagnosing faults in household appliances, the method comprising: Based on heterogeneous sensing nodes, the current conduction characteristics of the conducted signals generated by the target home appliance are obtained, and the current conduction characteristics are mapped to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the conducted signals are used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances deployed outside the target home appliance and having sensing functions; Based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions, the drift trend of the conduction characteristics is determined, and the fault confidence is calculated. Based on the drift trend and the fault confidence level, the health status of the target home appliance is determined.

[0007] Secondly, embodiments of this application provide a fault diagnosis device for household appliances, the device comprising: The feature acquisition module, based on heterogeneous sensing nodes, acquires the current transmission characteristics of the transmitted signal generated by the target home appliance, and maps the current transmission characteristics to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the transmitted signal is used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances deployed outside the target home appliance and having sensing functions; The fault identification module is used to determine the drift trend of the conduction characteristics and calculate the fault confidence level based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions. The status diagnosis module is used to determine the health status of the target home appliance based on the drift trend and the fault confidence level.

[0008] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0009] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0010] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the method as described in the first aspect.

[0011] In this embodiment, a heterogeneous sensing node is formed by a home appliance with sensing function deployed outside the target home appliance. The current conduction characteristics of the conduction signal generated by the target home appliance are obtained. By mapping the current conduction characteristics to the feature dimension corresponding to the standard operating conditions of the target home appliance, the feature data collected under different operating conditions can be unified into a comparable reference space for processing. Based on this, the drift trend of the conduction characteristics is determined and the fault confidence is calculated according to the current conduction characteristics and the reference conduction characteristics of the target home appliance under standard operating conditions. The health status of the target home appliance is determined based on the drift trend and the fault confidence. This achieves a quantitative representation of the gradual process of the home appliance from normal to fault, rather than simply making a binary judgment of "normal / fault". It has the ability to predict the potential deterioration trend of the health status of the home appliance, so that the diagnostic results can not only reflect whether a fault has occurred, but also characterize the evolution stage of the health status of the home appliance. Compared with existing methods that can only identify fixed known fault types in a preset database, this method can reveal the abnormal evolution of the target home appliance when the fault mode is not fully revealed and the fault type is not predefined, thus expanding the applicable boundary of home appliance fault diagnosis. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating the steps of an embodiment of a method for diagnosing faults in household appliances provided by this invention. Figure 2 This is a flowchart illustrating the steps of another embodiment of a home appliance fault diagnosis method provided by the present invention; Figure 3 This is a structural block diagram of a home appliance fault diagnosis device provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] The terms "first," "second," etc., used in the specification and claims of this invention are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, the first object can be one or more. Furthermore, the term "and / or" in the specification and claims is used to describe the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. In embodiments of this invention, the term "multiple" refers to two or more, and other quantifiers are similar.

[0017] Method Implementation Examples Reference Figure 1 The diagram illustrates a flowchart of an embodiment of a home appliance fault diagnosis method according to the present invention, the method comprising: Step 101: Based on heterogeneous sensing nodes, obtain the current conduction characteristics of the conduction signal generated by the target home appliance, and map the current conduction characteristics to the feature dimension corresponding to the standard operating condition of the target home appliance; the conduction signal is used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances deployed outside the target home appliance and having sensing functions.

[0018] Heterogeneous sensing nodes consist of deployed home appliances that are independent of the target home appliances, including but not limited to deployed smart door locks, smart doors and windows, smart switches, robot vacuum cleaners, floor scrubbers, smart speakers, smart displays, routers, human presence radars, etc. These home appliances have built-in sensing functions. For example, floor scrubbers have built-in inertial measurement units (IMUs), and routers have built-in microphone sensing units or channel state information (CSI) sensing units.

[0019] The heterogeneous sensing node and the target home appliance are independent devices. In this application, the heterogeneous sensing node has a built-in local processing module, which enables the acquisition of conducted signals to be completed by the heterogeneous sensing node without any hardware additions or modifications to the target home appliance. In addition, the heterogeneous sensing node can also extract conducted features through the local processing module, map the current conducted features to the feature dimensions corresponding to the standard operating conditions of the target home appliance, determine the drift trend of the conducted features, and calculate the confidence level of local faults.

[0020] Conducted signals are signals emitted by target home appliances that characterize their operating status. Specifically, they can be vibration signals or acoustic signals. Conducted signals are emitted by target home appliances and propagate through the walls, floors, or air paths of the home space, and are collected by heterogeneous sensing nodes.

[0021] For scenarios where target home appliances experience load fluctuations, heterogeneous sensing nodes can collect the conducted signals emitted by the target appliances and extract their current conducted characteristics. Load fluctuations refer to the fluctuations caused by the target home appliance not operating under preset standard steady-state conditions (fluctuations caused by non-standard loads). Load fluctuations can affect the accuracy of fault diagnosis; therefore, it is necessary to map the current conducted characteristics to the feature dimensions corresponding to the standard operating conditions of the target home appliance to eliminate the interference of the differences between non-standard loads and standard operating conditions on the subsequent diagnostic process.

[0022] The mapping method can employ a linear mapping algorithm, using basic operation types such as addition, multiplication, and threshold comparison, to save the computing power of heterogeneous sensing nodes.

[0023] Step 102: Based on the current conduction characteristics and the reference conduction characteristics of the target home appliance under the standard operating conditions, determine the drift trend of the conduction characteristics and calculate the fault confidence.

[0024] The layout of each family's house, the location of home appliances, the wall material, the floor material, and the furniture arrangement are fixed in the long term. Therefore, the transmission signal generated by the target home appliance under standard operating conditions will form a fixed and unique transmission characteristic after being transmitted through the family space.

[0025] When components of home appliances undergo gradual aging, such as the aging of shock absorbers in washing machines or the aging of compressor feet in air conditioners, the characteristics of the vibration / sound signature signals generated during operation will change continuously and slowly. This leads to a systematic, space-wide synchronous drift in the transmission characteristics, rather than a sudden change at a single moment. By analyzing the drift trend of the transmission characteristics, it is possible to determine whether the home appliance has malfunctioned.

[0026] The drift trend analysis can be performed using a judgment algorithm based on spatial conduction consistency verification and temporal drift trajectory fitting. Temporal drift trajectory fitting is performed on the conduction feature sequence for each heterogeneous sensing node, and the change trend of conduction features is judged by the slope of the fitted trajectory. Spatial conduction consistency verification is performed on all heterogeneous sensing nodes, and the spatial conduction consistency rate is calculated to filter out local transient environmental interference.

[0027] Based on the slope of the fitted trajectory and the spatial transmission consistency rate, the local fault confidence of the heterogeneous sensing nodes can be calculated. The fault confidence is used to characterize the degree of certainty that the diagnostic method provided in this application can determine the fault of the target home appliance. By fusing the local fault confidence of each effective heterogeneous sensing node, the global fault confidence can be obtained.

[0028] Step 103: Determine the health status of the target home appliance based on the drift trend and the fault confidence level.

[0029] Based on global fault confidence and spatial transmission consistency, the health status of target home appliances can be determined. For example, the health status can be divided into three progressive stages: micro-aging, moderate aging, and severe failure, and corresponding early warning information and fault diagnosis reports can be generated. These reports are then forwarded uplink through a smart home gateway that communicates with heterogeneous sensor nodes, allowing users or technicians to address the issues accordingly.

[0030] The heterogeneous sensors are also equipped with a distributed collaborative computing domain, which can acquire the conduction characteristics of the conducted signals collected by other heterogeneous sensors. This allows for the acquisition of parameters such as reference conduction characteristics, spatial conduction consistency rate, and global fault confidence, or a comprehensive assessment of the health status of the target home appliance. The distributed collaborative computing domain eliminates the need for centralized cloud computing in this solution, ensuring the timeliness and reliability of the diagnostic process.

[0031] The above diagnostic process does not rely on a preset fault feature library or fault codes and can identify latent faults caused by progressive component aging. In this embodiment, heterogeneous sensing nodes are formed by deploying sensing appliances outside the target appliance to obtain the current conduction characteristics of the conducted signals generated by the target appliance. By mapping the current conduction characteristics to the feature dimensions corresponding to the standard operating conditions of the target appliance, feature data collected under different operating conditions can be unified into a comparable reference space for processing. Based on this, the drift trend of the conduction characteristics is determined and the fault confidence is calculated according to the current conduction characteristics and the reference conduction characteristics of the target home appliance under standard operating conditions. The health status of the target home appliance is determined based on the drift trend and the fault confidence. This achieves a quantitative representation of the gradual process of the home appliance from normal to fault, rather than simply making a binary judgment of "normal / fault". It has the ability to predict the potential deterioration trend of the health status of the home appliance, so that the diagnostic results can not only reflect whether a fault has occurred, but also characterize the evolution stage of the health status of the home appliance. Compared with existing methods that can only identify fixed known fault types in a preset database, this method can reveal the abnormal evolution of the target home appliance when the fault mode is not fully revealed and the fault type is not predefined, thus expanding the applicable boundary of home appliance fault diagnosis.

[0032] Optionally, the method further includes: Step A1: Connect the heterogeneous sensing node through the smart home gateway; Step A2: In response to the target home appliance entering the standard operating condition, the smart home gateway synchronously triggers the heterogeneous sensor node to collect the conducted signals generated by the target home appliance; Step A3: Obtain the conduction characteristics based on the conduction signals collected by the different heterogeneous sensing nodes.

[0033] Through the lightweight adaptation layer of the smart home gateway, heterogeneous sensor nodes that are already deployed in the home but are not the target home appliances and have built-in sensing functions can be connected to form a distributed sensor network, ensuring the multi-source reliability of transmitted signal acquisition and transmitted characteristics.

[0034] By sending broadcasts to heterogeneous sensor nodes through a smart home gateway, the data collection time of all heterogeneous sensor nodes can be synchronized, ensuring the time sequence consistency of the data collected by each heterogeneous sensor node.

[0035] The specific method for home gateways to achieve synchronous triggering can be as follows: the smart home gateway broadcasts Beacon synchronization frames with precise timestamps at fixed intervals. After receiving the Beacon synchronization frames, each heterogeneous sensor node completes the calibration of its local clock based on the deviation between the precise timestamp in the frame and its local clock. When the target home appliance enters the standard operating condition, the smart home gateway sends a collection command with a trigger timestamp. All heterogeneous sensor nodes start synchronous collection at the same trigger time point according to the trigger timestamp.

[0036] Preferably, during local clock calibration, the synchronization accuracy between the smart home gateway's reference time and the local time of the heterogeneous sensor nodes is controlled within 10 μs. Local clock calibration only requires that the absolute deviation between the local time of the heterogeneous sensor nodes and the smart home gateway's reference time be less than or equal to 10 μs, regardless of whether the local clock is ahead or behind. That is, synchronization accuracy refers to the upper limit of the absolute value of the time deviation, and the range of synchronization accuracy is [-10 μs, +10 μs]. As long as the time deviation falls within this range, regardless of whether it is positive or negative, it is considered that the local clock calibration is complete and meets the data acquisition requirements of this solution.

[0037] When the target home appliance enters the preset standard operating condition and runs stably, all synchronized heterogeneous sensor nodes are triggered to synchronously collect the conducted signals generated by the target home appliance. Each heterogeneous sensor node completes the signal feature extraction locally, synchronizes the extracted conducted features to the collaborative computing domain, and merges them to generate a unique reference conducted feature of the target home appliance in the current home space.

[0038] When multiple home appliances are running simultaneously, heterogeneous sensing nodes will collect mixed conducted signals. At this time, the specific conducted characteristics of the target home appliance can be extracted through a three-level interference purification mechanism. The specific steps include: Sub-step A21, Frequency Domain Isolation and Screening: Based on the pre-calibrated dedicated operating base frequency of the target home appliance, extract the signal components of the target home appliance's base frequency ±2Hz narrowband and its 2nd / 3rd harmonic frequency band in the frequency domain, and filter out interference signals from other home appliances at other frequencies; Sub-step A22, Conductive Feature Matching: Calculate the amplitude attenuation ratio distribution of signals collected by multiple heterogeneous sensor nodes, perform similarity matching with the pre-constructed reference conductive features, retain only feature components with a matching degree higher than a preset threshold, and remove signal components that do not conform to the conduction law of the target home appliance; Sub-step A23, Distributed Consistency Verification: Based on the feature matching results of each heterogeneous sensor node, remove the data collected by heterogeneous sensor nodes with abnormal matching degree, and only retain the heterogeneous sensor nodes with valid matching to participate in the subsequent calculation process.

[0039] Similarly, sub-steps A21 to A23 are completed in the collaborative computing domain of heterogeneous sensing nodes, without the need for cloud involvement.

[0040] This solution establishes a unified access and control channel by connecting heterogeneous sensor nodes through a gateway, enabling centralized coordination of distributed home appliances at the sensing function level. In response to the target home appliance entering standard operating conditions, the gateway synchronously triggers each heterogeneous sensor node to collect transmitted signals. This ensures that the signals collected by each sensor node correspond to a completely consistent operating state segment of the target home appliance, eliminating interference from operating condition fluctuations introduced by inconsistent collection times and improving the consistency and accuracy of transmitted characteristics. Based on the transmitted signals collected by different heterogeneous sensor nodes, transmitted characteristics are obtained, integrating the sensing information from multiple nodes in different locations and with different sensing capabilities in space. This provides more representative and robust data support for accurately determining whether transmitted characteristics have drifted.

[0041] Optionally, the transmitted signal includes the vibration signal of the target home appliance, and / or the acoustic signature signal of the home appliance; step A3 may include: Sub-step A31: Obtain the vibration signals collected by different heterogeneous sensing nodes, and / or the spectral feature distribution of the acoustic signature signals after fast Fourier transform; Sub-step A32: Obtain the amplitude attenuation ratio, arrival time difference, and phase offset of the vibration signal and / or acoustic signal relative to the reference sensor node; the reference sensor node is the heterogeneous sensor node corresponding to the preset reference position.

[0042] Conducted signals mainly include vibration signals and acoustic signature signals, and their conduction characteristics include, but are not limited to, those of vibration signals and acoustic signatures. (1) The spectral characteristics distribution of the vibration / acoustic signal after Fast Fourier Transform (FFT) includes the proportion of low frequency components in 10-50Hz, the proportion of mid frequency components in 50-500Hz, and the proportion of high frequency components above 500Hz. (2) The amplitude attenuation ratio of the received signal of the heterogeneous sensor node at different locations, that is, the ratio of the amplitude of the received signal of each heterogeneous sensor node to the amplitude of the received signal of the reference node. (3) The arrival time difference and phase offset of signals received by heterogeneous sensing nodes at different locations; A reference node refers to a heterogeneous sensor node corresponding to a preset reference position. When a heterogeneous sensor node is located at a preset reference position, it has the best communication link quality. Generally, the heterogeneous sensor node that is closest to the target home appliance in a straight line can be taken as the reference node.

[0043] In the spectral characteristic distribution, the proportion of low-frequency components is the core feature of degradation of damping and shock absorption components (shock absorbers, shock absorber pads); the proportion of mid-frequency components is the core feature of degradation due to changes in motor load and blockage / leakage in the flow path; and the proportion of high-frequency components is the core feature of wear and increased clearance in rotating bearings.

[0044] The conduction features extracted by each heterogeneous sensor node correspond to the signal conduction results at a fixed location in the home space, and the features of multiple heterogeneous sensor nodes together constitute a unique reference conduction feature.

[0045] Phase offset refers to the relative phase difference, specifically the difference between the phase of the signal acquired by the current heterogeneous sensing node and the phase of the signal acquired by the reference node at the target appliance's dedicated operating frequency (e.g., a washing machine has a 20Hz spin-drying frequency). The unit is degrees or radians. Essentially, phase offset reflects the difference in path length of vibration or acoustic signals transmitted from the target appliance to different heterogeneous sensing nodes. It is one of the core components of transmission characteristics. Appliances in different locations exhibit different phase offset distributions at the same heterogeneous sensing node. Therefore, phase offset can be used for multi-applied interference separation, abnormal transmission path identification, and amplitude attenuation ratio verification, thus improving the anti-interference capability of this solution.

[0046] The calculation of phase offset can be processed within the heterogeneous sensing node itself, including: Each heterogeneous sensing node performs an FFT operation on the time-domain signal (i.e., vibration signal and acoustic signal) to obtain the frequency-domain amplitude spectrum and phase spectrum. The phase spectrum contains the signal phase value corresponding to each frequency point.

[0047] For the heterogeneous sensing node currently numbered i, extract the phase φi corresponding to the target home appliance's specific base frequency from its obtained phase spectrum. For the reference node, obtain the base frequency phase corresponding to the reference node as φref.

[0048] Then the phase offset Δφi of the heterogeneous sensing node currently numbered i is φi – φref.

[0049] The phase offset Δφi can be normalized to the range [-180°, 180°] for easier calculation. It is understandable that the phase offset of the reference node itself is 0, and the phase offset of the other heterogeneous sensing nodes is the difference between their own phase and the phase of the reference node.

[0050] The calculation logic for amplitude attenuation ratio and arrival time difference is similar to that for phase offset, and will not be repeated here.

[0051] It should be noted that the transmitted signals used to calculate the amplitude attenuation ratio, arrival time difference, and phase offset are exclusive transmitted signals generated solely by the operation of the target home appliance after undergoing the three-level interference purification process of sub-steps A21 to A23. These signals can uniquely correspond to the operating status and spatial transmission characteristics of the target home appliance, rather than the mixed signals originally collected by heterogeneous sensor nodes.

[0052] The fusion of vibration and acoustic signature signals, combining heterogeneous information from two different physical fields—mechanical vibration and acoustics—improves the probability of capturing potential fault characteristics. By acquiring the spectral feature distribution, the time-domain mixed signal is analyzed into a frequency-domain energy distribution, enabling the clear presentation of subtle changes in characteristic frequency components. Simultaneously, the amplitude attenuation ratio, arrival time difference, and phase shift are acquired, providing multi-dimensional spatial discrimination criteria for accurately identifying global feature drift.

[0053] Optionally, step 101 may include: Sub-step 1011: Obtain the current fundamental frequency amplitude of the target home appliance and the reference fundamental frequency amplitude of the target home appliance under the standard operating conditions, and normalize the current fundamental frequency amplitude and the reference fundamental frequency amplitude. Sub-step 1012: The product of the load correction coefficient and the target ratio is used as the load deviation coefficient; the target ratio is the ratio of the normalized reference fundamental frequency amplitude to the normalized current fundamental frequency amplitude; Sub-step 1013: The product of the current conduction feature and the reciprocal of the load deviation coefficient is used as the current conduction feature mapped to the feature dimension corresponding to the standard operating condition.

[0054] Sub-steps 1011 to 1013 describe the mapping process from the conduction features to the feature dimensions corresponding to the quasi-operating condition. First, the reference fundamental frequency amplitude A1 of the target home appliance under standard operating conditions and the current fundamental frequency amplitude A2 of the target home appliance under the current operating condition are collected, and the two are normalized to the interval [0, 1]. Specifically, the normalization process can be achieved by multiplying the derivative of the required fundamental frequency amplitude with the derivative of the pre-stored maximum fundamental frequency amplitude value, which is the reciprocal of the maximum value of the fundamental frequency amplitude, as the normalized fundamental frequency amplitude value.

[0055] Then, based on the normalized current fundamental frequency amplitude A2 and the normalized reference fundamental frequency amplitude A1, the load deviation coefficient α can be calculated using the formula α = K × (A2 / A1), where K is the load correction coefficient, and A2 / A1 is the target ratio, i.e., the ratio of the normalized reference fundamental frequency amplitude to the normalized current fundamental frequency amplitude. The load deviation coefficient α can be truncated with upper and lower limits of 0.8 to 1.2; that is, when α is greater than 1.2, α is forced to be 1.2, or when α is less than 0.8, α is forced to be 0.8. This avoids mapping distortion caused by extreme loads. The load correction coefficient K ranges from 0.8 to 1.2.

[0056] Finally, the eigenvalue F of the transmission characteristics under the current operating condition is multiplied by the reciprocal of the load deviation coefficient to obtain the eigenvalue F mapped to the standard operating condition dimension.

[0057] The mapping result F can be slightly modified using the formula F′ = F + (1 - α) × 0.05, where F′ is the modified mapping result and α is the load deviation coefficient. By modifying the mapping result F, nonlinear deviations can be further eliminated.

[0058] By acquiring and normalizing the current and reference fundamental frequency amplitudes, the dimensions and scales of different signal components are unified, making the changes in operating intensity reflected by the fundamental frequency amplitude more prominent. The load deviation coefficient is obtained by calculating the product of the load correction coefficient and the target ratio. Utilizing the monotonic correlation between the fundamental frequency amplitude and the equipment load intensity, a correction factor reflecting the degree of load deviation between the current operating condition and the standard operating condition is constructed. The product of the current conduction characteristic and the reciprocal of the load deviation coefficient is used as the mapped current conduction characteristic, achieving reverse compensation for the scaling effect caused by load differences in the current conduction characteristic. This allows the mapped current conduction characteristic to be directly compared with the reference conduction characteristic, and the reflected differences mainly originate from changes in the equipment's own health status, effectively suppressing the interference caused by load changes on fault diagnosis.

[0059] Optionally, step 102 may include: Sub-step 1021: For each heterogeneous sensing node, within a sliding time window formed by multiple standard operating condition cycles, fit the drift degree sequence of the conduction feature to obtain the drift slope of the heterogeneous sensing node; the drift degree is used to characterize the degree of change of the current conduction feature relative to the reference conduction feature; the drift slope is used to characterize the rate of change of the conduction feature with the number of standard operating condition cycles; Sub-step 1022: Determine the spatial conduction consistency rate based on the drift slope of all the heterogeneous sensing nodes.

[0060] A sliding time window is set, the length of which is the number of standard operating cycles, ranging from 7 to 30. Each element in the sliding time window represents a transmission characteristic sample under one standard operating cycle. The time-drift trajectory fitting can employ a least squares algorithm, which only involves addition, subtraction, and multiplication operations, reducing the computational overhead of heterogeneous sensing nodes. The specific calculation formula for time-drift trajectory fitting can be: k = [nΣ(xy) - Σx Σy] / [nΣ(x²) - (Σx)²] Where k is the drift slope of the heterogeneous sensing node, x is the sequence number of the effective operating cycle of the target home appliance, y is the drift degree of the heterogeneous sensing node in the current operating cycle, and n is the number of samples within the sliding time window; when the drift slope k≈0, it indicates that the conduction characteristics are stable and the target home appliance has no aging; when the drift slope k continuously increases slightly in the positive direction, it indicates that the target home appliance has gradual aging; when the drift slope k changes instantaneously, it is determined that there is transient interference, not a fault.

[0061] The drift slope k refers to the rate of change of conduction characteristics (such as the proportion of low-frequency components and the fundamental frequency amplitude) with the effective operating period. k>0 is a positive slope (indicating an increase in conduction characteristics), k<0 is a negative slope (indicating a decrease in conduction characteristics), and the smaller the absolute value of k, the more stable the conduction characteristics.

[0062] The drift slope of a single heterogeneous sensor node can only serve as a preliminary indicator of anomalies and cannot independently determine progressive aging or transient interference. This is because the drift slope of a single heterogeneous sensor node can be caused by various abnormal factors: such as hardware drift or calibration failure within the heterogeneous sensor node itself, leading to a shift in its conduction characteristics; or it could be due to localized transient interference near the heterogeneous sensor node (such as localized pressure from heavy objects or people leaning on it), causing localized signal changes. Therefore, further judgment requires relying on the spatial consistency of multiple heterogeneous sensor nodes.

[0063] The core judgment basis for distinguishing between gradual aging and instantaneous interference in this solution is the spatial consistency of the drift slopes of multiple heterogeneous sensing nodes. When effective heterogeneous sensing nodes accounting for a percentage greater than or equal to a first preset percentage threshold (e.g., 80%) exhibit drift slope changes in the same direction and of the same magnitude, it can be determined that the main body of the target household electrical appliance device has undergone gradual aging.

[0064] The gradual aging of a household electrical appliance device will lead to a global intensity change of the signal conducted by the household electrical appliance device, and this change will be synchronously reflected on all distributed heterogeneous sensing nodes through spatial conduction, so the vast majority of heterogeneous sensing nodes will present drift slope changes in the same direction.

[0065] When only a single heterogeneous sensing node, or adjacent heterogeneous sensing nodes accounting for a percentage less than or equal to a second preset percentage threshold (e.g., 20%) exhibit abrupt drift slope anomalies, and the slopes of the vast majority of the remaining heterogeneous sensing nodes remain stable, it is determined as local instantaneous interference, and the corresponding abnormal data is not included in the subsequent calculation process.

[0066] Instantaneous interference (such as personnel touch, local foreign matter) only affects signals in a small range of space and will not be transmitted to all nodes in the whole house, so only a few adjacent heterogeneous sensing nodes will have anomalies.

[0067] When a single heterogeneous sensing node has long-term continuous drift slope anomalies and the rest of the heterogeneous sensing nodes remain stable all the time, it is determined that the heterogeneous sensing node itself has a hardware fault, and a calibration prompt for this heterogeneous sensing node can be triggered.

[0068] It should be noted that the drift slope itself is a continuously valued parameter, which can be divided into three major categories: Stable zero slope: |k|<kth (e.g., the change of conduction characteristic is less than 0.5% every 100 operation cycles), representing that the operation state of the target household electrical appliance device is stable; Positive slope (indicating that the conduction characteristic rises): k>kth, which is the typical trend of aging of most mechanical components (e.g., aging of damping parts and bearing wear will lead to an increase in vibration amplitude); Negative slope (indicating that the conduction characteristic decreases): k<-kth, that is, the conduction characteristic decreases continuously with the operation cycles (e.g., when a new household electrical appliance device is just deployed, the fitting degree of internal parts is not high and the vibration amplitude is relatively high; as the number of operations of the household electrical appliance device increases, the fitting degree of internal parts improves and the vibration amplitude decreases).

[0069] Wherein, kth is a stability threshold, used for characterizing the maximum value of the drift slope of the heterogeneous sensing node when the target household electrical appliance device is in a stable operation state.

[0070] Negative slope also follows the spatial consistency judgment rule, that is, if most heterogeneous sensor nodes have a synchronous negative slope, it is judged that the state of the home appliance has changed, and if a single heterogeneous sensor node has a negative slope, it is judged that there is local interference or node failure.

[0071] By fitting the drift degree sequence within a sliding time window to obtain the drift slope, the long-term rate of change of conduction characteristics with the number of uses was extracted from historical data of multiple standard operating cycles. The introduction of the sliding time window ensures that drift trend analysis is always based on recent continuous observation data, which can both accumulate sufficient samples for statistical fitting to eliminate the interference of single random fluctuations and automatically adapt to the stage characteristics of equipment state evolution. Furthermore, the spatial conduction consistency rate is determined based on the drift slope of all heterogeneous sensor nodes, realizing cross-validation of multiple nodes in the spatial dimension, enhancing the reliability and anti-interference ability of drift trend determination.

[0072] Optionally, sub-step 1022 includes: Sub-step 10221: The ratio of the total number of nodes drifting in the same direction to the total number of heterogeneous sensing nodes is used as the spatial conduction consistency rate; Sub-step 10222: The same-direction drifting node refers to the heterogeneous sensing nodes with a drift slope greater than the stability threshold, which is the larger of the two: heterogeneous sensing nodes with a drift slope greater than the stability threshold and heterogeneous sensing nodes with a drift slope less than the stability threshold.

[0073] The spatial conduction consistency rate C is calculated as follows: C = Total number of nodes drifting in the same direction / Total number of heterogeneous sensor nodes. When the spatial conduction consistency rate is greater than or equal to the fourth preset consistency rate threshold (e.g., C ≥ 70%), it is determined to be a full-space synchronous drift, which is a high-probability real fault. When the spatial conduction consistency rate is less than the fifth preset consistency rate threshold (e.g., C < 40%), it is determined to be a local single-point mutation, which is an environmental interference and is directly filtered out. It should be noted that the total number of heterogeneous sensor nodes refers to the total number of valid heterogeneous sensor nodes obtained after sub-steps A21 to A23. Valid heterogeneous sensor nodes have excluded heterogeneous sensor nodes that are offline in communication, have failed in signal, are out of sync in timing, or have abnormal hardware.

[0074] The statistical basis for the total number of nodes drifting in the same direction is also based on effective heterogeneous sensing nodes. The core criterion for determining nodes drifting in the same direction is that the drift slope direction is consistent with the globally dominant drift direction, and the absolute value of the slope exceeds the stability threshold. The specific process is as follows: First, a stability threshold kth is preset to distinguish between normal random fluctuations and effective drift trends. Normal random fluctuations are defined as |k| ≤ kth (transduction characteristics show no obvious trend, belonging to normal operation noise). Heterogeneous sensor nodes with stable fluctuations are not included in the statistics of heterogeneous sensor nodes with effective drift trends, avoiding interference from minor random fluctuations in the determination. Effective drift trends include: 1. Positive drift: drift slope k > kth (transduction characteristics continuously increase); 2. Negative drift: drift slope k < -kth (transduction characteristics continuously decrease).

[0075] Subsequently, the globally dominant drift direction is determined. The number of heterogeneous sensing nodes drifting positively and the number drifting negatively are counted among all valid heterogeneous sensing nodes. The direction corresponding to the one with the higher percentage is the globally dominant drift direction.

[0076] Finally, the total number of nodes drifting in the same direction is counted. Among all effective heterogeneous sensing nodes, the heterogeneous sensing nodes whose drift slope direction is consistent with the globally dominant drift direction and whose absolute drift slope value exceeds the stability threshold kth are considered to be drifting nodes, and their total number is the total number of drifting nodes in the same direction.

[0077] The ratio of the total number of nodes drifting in the same direction to the total number of heterogeneous sensor nodes is used as the spatial conduction consistency rate, providing a quantitative measure of spatial consistency. Heterogeneous sensor nodes with small drift amplitudes and still in a state of random fluctuation are filtered out using a stable drift threshold, retaining only those exhibiting a clear drift trend for inclusion in the directional consistency statistics. Subsequently, heterogeneous sensor nodes exhibiting significant drift are divided into two groups: positive drift and negative drift. The number of nodes in the majority group is taken as the total number of effective nodes drifting in the same direction, and the majority direction is identified as the globally dominant drift direction. This approach ensures that the spatial conduction consistency rate robustly reflects the degree of consistency in the trend of conduction characteristics changes caused by changes in the device's state, pointing to the same direction as the majority of sensor nodes in space, effectively reducing the interference of local abnormal nodes on global judgment.

[0078] Optionally, the drift slope and the spatial conduction consistency rate are each assigned a weight; step 102 may include: Sub-step 1023: The weighted result of the drift slope and the spatial conduction consistency rate is used as the local fault confidence of the heterogeneous sensing node; Sub-step 1024: Remove heterogeneous sensing nodes whose local fault confidence is less than a preset local fault confidence threshold, and perform a weighted average of the local fault confidence of the remaining heterogeneous sensing nodes to obtain the global fault confidence.

[0079] The formula for calculating the confidence level P of a local fault is: P = w1*|k| + w2*C, where w1 is the weight of the drift slope and w2 is the weight of the spatial transmission consistency rate, with w1 + w2 = 1 and P∈[0,1].

[0080] The specific method for obtaining the global confidence score is as follows: discard heterogeneous sensing nodes with low confidence scores P < 0.2, and take the weighted average of the local fault confidence scores P of the remaining heterogeneous sensing nodes to obtain the global fault confidence score Pall.

[0081] Drift slope and spatial conduction consistency rate are each assigned a weight. The weighted result of these two factors is used as the local fault confidence of heterogeneous sensing nodes. This ensures that the fault confidence of each heterogeneous sensing node depends both on the rate of degradation of its own perceived conduction characteristics over time and on the degree to which its drift direction aligns with the drift directions of the majority of heterogeneous sensing nodes globally. When relying solely on drift slope to calculate confidence, a node experiencing a drastic unidirectional drift due to local environmental changes may lead to an artificially high local confidence. However, by introducing spatial conduction consistency rate as an adjustment factor, the confidence is only sufficiently enhanced when the node's drift direction aligns with the drift direction of the majority of nodes in space. If the directions are inconsistent, the confidence is suppressed. This achieves confidence assessment under multi-node spatial collaborative verification, making the local fault confidence more reflective of the device itself rather than changes in the local environment. Building upon this, after removing heterogeneous sensing nodes whose local fault confidence is less than a preset local fault confidence threshold, the local fault confidence of the remaining sensing nodes is weighted and averaged to obtain the global fault confidence, achieving fusion inference from local to global perspectives. The final global fault confidence integrates cross-validation information from multiple heterogeneous sensing nodes that have passed spatial transmission consistency verification. Compared to schemes relying on a single or a few sensors, its results exhibit higher reliability and robustness.

[0082] Optionally, step 104 may include: Sub-step 1041: If the spatial conduction consistency rate is greater than or equal to the first preset consistency rate threshold and the global fault confidence is within the first confidence interval, determine that the target home appliance is in a slightly aging state. Sub-step 1042: If the spatial conduction consistency rate is greater than or equal to the second preset consistency rate threshold, and the global fault confidence level is within the second confidence level interval, determine that the target home appliance is in a moderately aging state; the second preset consistency rate threshold is greater than the first preset consistency rate threshold; the minimum value of the second confidence level interval is greater than the maximum value of the first confidence level interval; Sub-step 1043: If the spatial conduction consistency rate is greater than or equal to the third preset consistency rate threshold, and the global fault confidence is within the third confidence interval, determine that the target home appliance is in a severely aged state; the third preset consistency rate threshold is greater than the second preset consistency rate threshold; the minimum value of the third confidence interval is greater than the maximum value of the second confidence interval.

[0083] The health status determination rules provided in this solution are as follows: Mild aging state: The drift slope increases slowly, the spatial conduction consistency rate is greater than or equal to the first preset consistency rate threshold (e.g., spatial conduction consistency rate C≥60%), and the global fault confidence is in the first confidence interval (e.g., Pall∈[0.3, 0.5]). Then it is determined that the component performance is slightly degraded, there is no fault, and it enters the observation period. Moderate aging state: The drift slope continues to rise, the spatial conduction consistency rate is greater than or equal to the second preset consistency rate threshold (e.g., spatial conduction consistency rate C≥75%), and the global fault confidence is in the second confidence interval (e.g., global fault confidence Pall∈[0.5, 0.8]). Then it is determined that the component is aging and the performance has obviously deteriorated. Maintenance suggestions are pushed. Severe aging state: The drift slope rises rapidly, the spatial conduction consistency rate is greater than or equal to the third preset consistency rate threshold (e.g., spatial conduction consistency rate C≥85%), and the global fault confidence is in the third confidence range (e.g., global fault confidence Pall>0.8). Then, it is determined that the component is close to failure, an emergency alarm is pushed, and the user is prompted to stop using the device immediately.

[0084] In all three judgment processes mentioned above, the drift slope is required to be greater than the stability threshold.

[0085] This scheme combines spatial conduction consistency rate and global failure confidence level as the grading criteria for health status, constructing a progressive grading judgment framework. In the mild aging state, most nodes in the space show early, weak feature drift with a consistent direction. In the moderate aging state, both the spatial conduction consistency rate threshold and the confidence level interval increase accordingly, reflecting the mutual confirmation of deepening degradation and enhanced spatial conduction synergy. In the severe aging state, both the threshold and interval settings are at the highest level, meaning that equipment structural degradation has been highly consistent throughout the entire space and has accumulated to a significant level. This grading framework makes the boundaries between mild, moderate, and severe aging states clear and logically consistent, enabling precise division of the gradual process from initial degradation to near failure, avoiding misjudgments based on a single indicator.

[0086] In addition to natural aging of materials, the following progressive mechanical failures can also be identified by this solution, including: Connection and fastening loosening faults: The preload of bolts and clips gradually decreases, from slight loosening at the beginning to complete loosening later, corresponding to a gradual increase in the intermittent impact (corresponding to the sensing characteristic of signal peak). Fluid passage blockage / leakage faults: oil stains clogging the range hood, slow refrigerant leakage in the air conditioner, and minor blockage in the refrigerator's refrigeration pipes. The characteristics of these blockages change continuously with the degree of buildup / leakage. Rotary support wear failure: continuous wear of bearing raceways and slow drying of grease belong to the broad category of wear-related degradation, which is consistent with the aging logic; Slippage faults in power transmission: Wear and stretching of the drive belt, and decrease in tension result in a gradual decrease in transmission efficiency and fluctuations in transmission characteristics; The above-mentioned faults are all latent faults that are difficult to identify in the early stages by related technologies. This solution does not require separate pre-storage of fault characteristics. It can complete the identification and classification of these latent faults by relying on the logic of spatial synchronous drift.

[0087] Optionally, the method further includes: Step B1: If the spatial conduction consistency rate is greater than the fourth preset consistency rate threshold, determine the target dimension of the conduction feature that dominates the drift; the fourth preset consistency rate threshold is greater than the first preset consistency rate threshold and less than the second preset consistency rate threshold. Step B2: Based on the correlation between the general mechanical structure and the drift law of the conduction feature, determine the target general mechanical structure that has failed, corresponding to the target dimension of the conduction feature; the target device includes at least one of the general mechanical structures.

[0088] The system pre-stores the correspondence between the general mechanical structure of home appliances and the drift law of conduction characteristics, rather than pre-setting feature templates for specific faults. When the spatial conduction consistency rate is detected to be greater than the fourth preset consistency rate threshold (e.g., spatial conduction consistency rate ≥ 70%), that is, when full-space synchronous drift is detected, the target dimension of the conduction characteristic that dominates the drift is identified, and the corresponding general mechanical structure is matched according to the drift law of the target dimension of the conduction characteristic to locate the root cause component of the fault. The correspondence between the general mechanical structure and the drift law of conduction characteristics is obtained based on the vibration principle of the general mechanical structure and is independent of the specific brand and model of the home appliance.

[0089] The specific steps for obtaining the target dimension of the transmitted feature can be as follows: Sub-step B11, Single-dimensional synchronous drift rate calculation: For each feature dimension, first select the nodes drifting in the same direction under that dimension, and calculate the average relative drift rate of all nodes drifting in the same direction. The average relative drift rate is calculated as follows: (mean of each transmission feature within the current sliding time window - mean of the reference transmission feature) / mean of the reference transmission feature * 100%. At the same time, verify that the proportion of nodes drifting in the same direction in this dimension is ≥ 80%, confirming that this dimension is an effective dimension for synchronous drift across the entire space. Dimensions that do not meet the criteria are judged as random fluctuations and are not included in the subsequent sorting.

[0090] Sub-step B12, Drift Contribution Rate Calculation: Calculate the drift contribution rate for each feature dimension, quantifying its contribution to the overall drift: Contribution Rate = Absolute value of the average relative drift rate of that feature dimension / Sum of the absolute values ​​of the drift rates of all feature dimensions * 100% Sub-step B13, Dominant Dimension Determination and Homology Verification: Dominant dimension determination includes single dominant dimension determination and joint dominant dimension determination. Single dominant dimension is determined if the contribution rate of a certain feature dimension is ≥60%, indicating that the dimension is dominant and drifting, corresponding to a single type of component degradation. Joint dominant dimension is determined if the sum of the contribution rates of the top two dimensions is ≥80%, and neither of them is less than 60%, indicating that the two dimensions are jointly dominant, corresponding to a composite type of component degradation. Homology verification is determined if the drift rate of all feature dimensions is <3%, further confirming that the drift originates from the target home appliance itself, rather than external interference.

[0091] The pre-stored correspondences form a category-level general mechanical structure law knowledge base, constructed according to the types of common core components of home appliances, without relying on the exclusive characteristics of specific brands or models. Each correspondence includes elements such as the specific category of the component, degradation mechanism, conduction characteristic drift law, and applicable home appliance category. General mechanical structures can include damping and shock absorption components, rotating support components, connection and fastening components, power output components, and fluid passage components, etc., and the following correspondences exist for these components: I. Damping and vibration reduction components; Specific component categories: washing machine suspension springs / shock absorbers, air conditioner compressor rubber shock-absorbing feet, refrigerator compressor support feet, range hood whole machine shock-absorbing pads; Degradation mechanism: Polymer rubber materials are subjected to alternating stress for a long time, resulting in fatigue aging, elasticity decay, and a decrease in damping coefficient, which leads to a continuous reduction in the overall vibration damping capacity of the machine. Conductive drift characteristics: All heterogeneous sensing nodes drift synchronously in the positive direction. The amplitude proportion of the low-frequency band (10~60Hz) where the fundamental frequency is located continues to increase monotonically, and the amplitudes of the second and third integer harmonics increase synchronously and proportionally. There are no discrete high-frequency impact peaks, the overall spectrum shape remains stable, and only the overall intensity increases uniformly. The drift rate is uniform, which conforms to the quasi-linear slope characteristics of progressive aging. Compatible home appliance categories: washing machines, air conditioners, refrigerators, and range hoods.

[0092] II. Rotary support components (bearing wear and aging); Specific component categories: washing machine motor bearings, air conditioner indoor and outdoor fan bearings, refrigerator compressor crankshaft bearings, and range hood impeller bearings; Degradation mechanism: bearing raceway wear, grease drying and failure, increased clearance between rolling elements and raceway, resulting in periodic micro-impacts during operation; Characteristic drift pattern: All nodes drift synchronously in the positive direction, the fundamental frequency amplitude of the equipment is basically stable, but the sideband spectrum of bearing fault characteristic frequency appears in the mid-to-high frequency band (300~2000Hz), and the proportion of impact characteristics continues to increase; in the middle and late stages of wear, the time domain signal shows periodic impact envelopes, and the frequency of high frequency peaks gradually increases; the drift rate gradually accelerates with the deepening of wear, showing a nonlinear acceleration trend.

[0093] Compatible appliance categories: Appliances with rotary motors.

[0094] III. Connecting and fastening components (loosening fault) Specific component categories: washing machine inner drum fixing bolts, air conditioner pipe fixing clips, refrigerator compressor connecting plate bolts, and appliance casing assembly fasteners; Degradation mechanism: Long-term alternating vibration causes the preload of fasteners to decrease and loosen, resulting in gaps between mating parts and intermittent collisions during operation; Characteristic drift pattern: Non-monotonic intermittent drift, with random transient peaks in the time domain signal, and the spectrum exhibits wideband spurious characteristics; weak spatial synchronization: Only the amplitude of neighboring nodes near the loosened position increases significantly, while the change in distant nodes is not obvious, and the proportion of heterogeneous sensing nodes drifting in the same direction is usually less than 60%; The impact characteristics do not have a fixed occurrence period and are strongly correlated with the operating load and speed.

[0095] Compatible home appliance categories: All types of home appliances.

[0096] IV. Power output components (motor performance degradation, corresponding to negative slope drift) Specific component categories: permanent magnet synchronous motors for washing machines, brushless DC motors for air conditioners, and inverter compressor motors for refrigerators; Degradation mechanism: Aging of motor winding insulation and demagnetization of permanent magnets lead to a decrease in rated output torque and speed, and a continuous reduction in the excitation force of the whole machine; Characteristic drift pattern: All nodes synchronously drift negatively, the fundamental frequency amplitude continues to decrease slowly, the proportion of each frequency band component remains basically unchanged, and the overall characteristics decrease proportionally; the spectrum shape has no obvious distortion, only the overall intensity decays uniformly; the drift rate is slow and uniform, showing quasi-linear characteristics. Compatible home appliance categories: Inverter appliances.

[0097] V. Fluid passage components (flow path blockage or leakage) Specific component categories: oil accumulation at the air outlet of range hood duct, refrigerant circulation piping of air conditioner, refrigeration system piping of refrigerator, and drain pump flow path of washing machine.

[0098] Degradation mechanism: Deposits and blockages in the flow path increase resistance, or refrigerant leaks reduce load, ultimately altering the operating load of the power components. Characteristic drift patterns: Blockage scenario: Increased wind resistance / flow resistance, increased motor load, positive drift of acoustic / vibration amplitude in the mid-frequency range (100~500Hz), and slight increase in fundamental frequency; Leakage scenario: Decreased compressor load, negative drift of fundamental frequency vibration amplitude, accompanied by slight fluctuations in operating frequency; Good synchronization across all nodes, and the drift rate gradually changes with the degree of blockage / leakage.

[0099] Compatible home appliance categories: range hoods, air conditioners, refrigerators, washing machines.

[0100] When full-space synchronous drift is detected, the target dimension of the dominant drift transmission characteristic is determined, pinpointing the most prominent physical phenomenon during the degradation of home appliances. Subsequently, based on the correlation between the general mechanical structure and the drift pattern of the transmission characteristic, the target general mechanical structure corresponding to the target dimension that has failed is identified, achieving reverse inference from the transmission characteristic to the specific structural component. This approach enables diagnostics not only to determine the overall health of the equipment but also to locate the specific structural component type that has failed based on the degradation pattern of the general mechanical structure, without relying on a priori fault database for the specific equipment model. This enhances the precision of the diagnostic method and its value for repair guidance.

[0101] Optionally, the method further includes: Step C1: Update the reference conduction characteristics within a sliding time window formed by multiple standard operating condition cycles; The sliding time window is used to accommodate a fixed number of conduction characteristics under the standard operating conditions; step C1 may include: Sub-step C11: The ratio of the difference between the first conduction feature most recently added to the sliding time window and the second conduction feature most recently removed from the sliding time window to the fixed number is used as the update amount; Sub-step C12: Sum the reference conduction feature and the update amount to obtain the updated reference conduction feature.

[0102] Perform real-time update of the reference conduction feature for each feature dimension of the conduction feature (such as the low-frequency component proportion of each individual heterogeneous sensing node, the amplitude attenuation ratio between heterogeneous sensing nodes, the phase offset, etc.), and all conduction features are updated independently.

[0103] The update process includes: Sliding time window setting: set a fixed window length N=10 (only the feature samples of the latest 10 valid normal working conditions are retained).

[0104] Update trigger condition: Only when the current operation cycle meets the standard working condition, the conduction signal is valid, there is no redundant equipment interference, and it is determined to be in a normal state (global fault confidence < 0.2), the conduction feature of the current operation cycle will be included in the update window; samples with faults, interference, and load abnormalities will not participate in the update, so as to avoid the reference conduction feature from being polluted.

[0105] The update logic of the reference conduction feature is: When the sliding time window is not full (cumulative valid samples < N): directly add new samples to the window, and use the arithmetic mean of all cumulative samples as the new reference conduction feature; When the window is full (cumulative valid samples ≥ N): eliminate the oldest 1 old sample in the window, add the latest 1 sample, and complete the calculation of the update amount through the added and deleted samples. The specific calculation formula is: Updated reference conduction feature = reference conduction feature before update + (conduction feature of the new sample - conduction feature of the eliminated old sample) / window length.

[0106] The reference conduction feature only involves basic operations, with low computational complexity, which is fully adapted to the computing power of heterogeneous sensing nodes.

[0107] For example, taking a washing machine as the target household appliance, taking the conduction feature of the 10-50Hz low-frequency component proportion of the balcony smart door lock (reference node) as an example, the window length N=10, the initial reference conduction feature is obtained by the first 3 normal dehydration calibrations, and the reference conduction feature is updated subsequently with the normal operation of the washing machine.

[0108] The update process is as follows: In the initial accumulation phase (when the sliding time window is not full), the low-frequency percentages of the first three effective normal dehydration cycles are 34.8%, 35.2%, and 35.0%, respectively; therefore, the initial reference conduction characteristic = (34.8% + 35.2% + 35.0%) ÷ 3 = 35.0%; the effective percentages of the subsequent 4th to 10th normal dehydration cycles are 34.9%, 35.1%, 35.0%, 34.7%, 35.3%, 34.8%, and 35.2%, respectively. By the 10th run, the sliding time window is full, and the 10 samples are: [34.8, 35.2, 35.0, 34.9, 35.1, 35.0, 34.7, 35.3, 34.8, 35.2] (unit: %). At this point, the reference conduction characteristic = (34.8 + 35.2 + 35.0 + 34.9 + 35.1 + 35.0 + 34.7 + 35.3 + 34.8 + 35.2) ÷ 10 = 35.00%. During the sliding update phase (window full, incremental replacement), the 11th normal dehydration showed an effective low-frequency proportion of 35.1%. Incremental update was then performed: the oldest old sample (34.8%) was removed, and the latest 11th sample (35.1%) was added. Therefore, the reference conduction characteristic = 35.00% + (35.1% - 34.8%) ÷ 10 = 35.00% + 0.03% = 35.03%.

[0109] The updated samples within the sliding time window are: [35.2, 35.0, 34.9, 35.1, 35.0, 34.7, 35.3, 34.8, 35.2, 35.1] (unit: %).

[0110] The 12th normal dehydration cycle showed an effective low-frequency rate of 34.9%. (Continued updates will continue.) If we remove the oldest sample in the current window (35.2%) and add the latest 12th sample (34.9%), then the reference transmission characteristic = 35.03% + (34.9% - 35.2%) ÷ 10 = 35.03% - 0.03% = 35.00%.

[0111] The update of the reference conduction feature can include abnormal sample filtering: if someone leans against the washing machine during the 13th spin cycle, causing the low-frequency proportion to suddenly increase to 38.2%, the system determines this as abnormal interference (exceeding the 3σ normal fluctuation range), then this sample will not be included in the update window, and the reference conduction feature will remain unchanged at 35.00%, avoiding the pollution of the reference conduction feature by a single instantaneous interference.

[0112] By updating the reference conduction characteristics within a sliding time window, the reference baseline can adaptively evolve along with the normal, slow aging process of the equipment, avoiding misjudgments of normal aging accumulation due to long-term static reference conduction characteristics. The update method employs incremental updates, dividing the difference between the newly added first conduction characteristic and the most recently removed second conduction characteristic by a fixed amount as the update increment, which is then added to the reference conduction characteristic to obtain the updated reference conduction characteristic, resulting in low computational resource consumption. The updated reference conduction characteristic represents the average level of conduction characteristics under all standard operating conditions within the window, smoothly reflecting the long-term slow-changing trend caused by normal aging of the equipment. While maintaining sensitivity to abnormal drift detection, it effectively reduces the probability of falsely reporting natural aging as a fault.

[0113] Reference Figure 2 The diagram illustrates a flowchart of another embodiment of the home appliance fault diagnosis method of the present invention, the method comprising: Step S1: Node adaptation and spatiotemporal synchronization; This step is equivalent to steps A1 and A2 above.

[0114] Step S2: Construction of spatial intrinsic conduction fingerprint baseline; Spatial intrinsic conduction fingerprint refers to conduction characteristics, while spatial intrinsic conduction fingerprint baseline refers to reference conduction characteristics.

[0115] Step S3: Baseline unsupervised adaptive update; This step is equivalent to step C1 above.

[0116] Step S4: Normalization of operating conditions; This step is equivalent to step 101 above.

[0117] Step S5: Fault / Anomaly Identification and Confidence Measurement; This step is equivalent to step 102 above.

[0118] Step S6: Fault root cause location and graded early warning.

[0119] This step is equivalent to steps 103, B1, and B2 mentioned above.

[0120] The following section uses the diagnosis of aging shock absorbers in a washing machine as an example to introduce the above steps S1 to S6.

[0121] A family home has three bedrooms and one living room. A brand-name front-loading washing machine (non-smart, no internet connectivity, no built-in vibration sensor, used for 3 years) is installed on the balcony. The heterogeneous sensor nodes deployed in the home include: the IMU built into the balcony smart door lock, the IMU built into the living room robot vacuum cleaner, the balcony smart door and window sensors (door magnets), the microphone array of the living room smart speaker, and the vibration sensor built into the master bedroom smart switch. The sliding time window is set to 15 spin-drying cycles. During mapping, the linear mapping algorithm has preset parameters: K=1.0, and the maximum fundamental frequency amplitude A=500mV. When calculating the fault confidence, preset parameters are: w1=0.6, w2=0.4. Then... Step S1, Node Adaptation and Synchronization: The home smart home gateway automatically connects to the above 5 heterogeneous sensor nodes through the Matter protocol, completes node registration and clock calibration, and achieves a time synchronization accuracy of 8μs for all nodes; the balcony smart door lock is determined as the reference node (closest to the washing machine, with the best link quality).

[0122] Step S2, Standard Operating Condition Definition and Baseline Construction: Sub-step S21: Define the standard operating conditions of the washing machine: no load, maximum spin speed of 1200 rpm, single spin cycle duration of 3 minutes, and the base frequency of the washing machine motor is 20 Hz under this condition; Sub-step S22: After the washing machine is installed for the first time, the system automatically triggers three standard operating conditions for spin-drying. During each operation, the gateway sends a synchronous acquisition command, and five nodes synchronously acquire the vibration and acoustic signals generated by the washing machine during spin-drying. Sub-step S23: Each heterogeneous sensing node performs signal sampling (IMU sampling frequency 1kHz, microphone sampling frequency 48kHz) and FFT spectrum transformation locally to extract the proportion of low-frequency components in the 10-50Hz range, the amplitude attenuation ratio of each node, phase offset and other conduction characteristics. Sub-step S24: The collaborative computing domain integrates the conduction characteristic data from the three runs to construct the reference conduction of the washing machine: Under normal standard operating conditions, the proportion of the 10-50Hz low-frequency component is stable at 35%±2%, the amplitude attenuation ratio of the balcony smart door lock and the living room robot vacuum is stable at 1:0.4±0.03, and the fluctuation range of the phase offset of each node is ≤5%; the fundamental frequency amplitude of the reference node under standard operating conditions is A1=400mV.

[0123] Step S3, Baseline Adaptive Update: Each time the user uses the washing machine's spin cycle, the system automatically collects operational data. If the characteristic fluctuation of this operation is within the normal fluctuation range of the reference transmission characteristic, the reference transmission characteristic is updated through a moving average algorithm to adapt to minor changes in the washing machine's installation location, changes in the placement of furniture on the balcony, and other environmental influences. At the same time, it filters out instantaneous interference such as people walking around and closing the door.

[0124] Step S4, Operating Condition Normalization: When a user uses the washing machine at full load (5kg of clothes) for spin-drying, the reference node collects the current fundamental frequency amplitude A2=480mV, and the current low-frequency component ratio F=0.6 (60%); normalization is performed using a lightweight linear mapping algorithm: Sub-step S41: Normalize the fundamental frequency amplitude: A1=400 / 500=0.8, A2=480 / 500=0.96; Sub-step S42, load deviation coefficient calculation: α = 1.0 × (0.96 / 0.8) = 1.2, reaching the upper limit, take α = 1.2; Sub-step S43, Linear mapping: F = 0.6 × (1 / 1.2) = 0.5; Sub-step S44, result correction: F′=0.5+(1-1.2)×0.05=0.49.

[0125] Step S5: Fault / Anomaly Identification and Confidence Measurement (Shock Absorber Aging Identification): Sub-step S51, Single-node temporal drift trajectory fitting: After 3 years of use, the 4 shock absorbers show progressive aging, and the damping performance decreases. During spin-drying, the eccentric vibration of the inner drum is amplified. Within the sliding time window (15 spin-drying samples), the system performs first-order linear trajectory fitting on the drift degree sequence of 5 heterogeneous sensing nodes to obtain the drift slope of each node: smart door lock for balcony k1=0.032, robot vacuum cleaner for living room k2=0.029, smart door and window sensor for balcony k3=0.031, smart speaker for living room k4=0.027, smart switch for master bedroom k5=0.028. The drift slope of all nodes increases slightly in the positive direction, showing a progressive upward trend.

[0126] Sub-step S52, Spatial conduction consistency verification: Statistically analyze the drift slope direction of all heterogeneous sensing nodes. All of them are rising in the positive direction. The number of nodes drifting in the same direction is 5. The total number of effective nodes is 5. The spatial conduction consistency rate C = 5 / 5 = 100% ≥ 70%. It is determined to be a synchronous drift in the whole space. Environmental interference is excluded, and it is confirmed as a high probability real fault.

[0127] Sub-step S52, Distributed Confidence Fusion: Each heterogeneous sensor node locally calculates the local fault confidence: P1=0.6×0.032+0.4×1.0=0.419, P2=0.6×0.029+0.4×1.0=0.417, P3=0.6×0.031+0.4×1.0=0.419, P4=0.6×0.027+0.4×1.0=0.416, P5=0.6×0.028+0.4×1.0=0.417; The local fault confidence of all heterogeneous sensor nodes is greater than 0.2, so they are not discarded. The weighted average of the above local fault confidence is taken to obtain the global fault confidence P=0.418, and the average drift slope k=0.029.

[0128] Step S6: Fault Root Cause Location and Graded Early Warning: Sub-step S61, Root Cause Location: The system identifies the dominant drift feature of this anomaly as a surge in low-frequency components and synchronous amplification of amplitude across all nodes. Matching the general mechanical structure of the washing machine: the shock absorber is the core damping component of the washing machine. After aging, the damping performance decreases, which will directly lead to a significant increase in the low-frequency vibration component of the inner drum rotation, and the attenuation amplitude of the vibration signal transmitted through the ground is significantly reduced. This matches the drift pattern of the transmission characteristics in this case, and the root cause of the fault is the aging of the washing machine's shock absorber.

[0129] Sub-step S62, graded early warning: According to the judgment rules, the global fault confidence P=0.418∈[0.3, 0.5], the spatial transmission consistency rate C=100%≥60%, and the average drift slope slowly increases, which is judged to be the micro-aging stage (mild aging state). A reminder is pushed to the user's mobile application: "Your washing machine shock absorber has shown slight aging, and its performance has begun to decline. It has entered the observation period. It is recommended to pay attention to the operating status regularly."

[0130] Sub-step S63, Follow-up Tracking: As the shock absorber ages further, the drift slope within the sliding time window continues to rise. After 3 months, the average drift slope k=0.058, the spatial transmission consistency rate C=90%, and the global fault confidence level P=0.63, indicating a moderate aging stage (moderate aging state). A maintenance reminder is pushed: "Your washing machine's shock absorber has experienced moderate aging, and its damping performance has decreased. It will vibrate violently during spin-drying. It is recommended to replace the shock absorber as soon as possible to avoid damage to the washing machine from the inner drum collision." At the same time, the corresponding washing machine model's shock absorber replacement plan and information on nearby after-sales service outlets are pushed. If the user does not handle it in time, the shock absorber ages further. After 5 months, the average drift slope k=0.095, the spatial transmission consistency rate C=95%, and the global fault confidence level P=0.86, indicating a severe failure stage (severe aging state). An emergency alarm is pushed, prompting the user to immediately stop using the washing machine to avoid equipment damage and safety accidents.

[0131] In summary, in this embodiment of the application, by deploying home appliances with sensing functions outside the target home appliance to form a heterogeneous sensing node, the current transmission characteristics of the transmission signal generated by the target home appliance are obtained. By mapping the current transmission characteristics to the feature dimensions corresponding to the standard operating conditions of the target home appliance, the feature data collected under different operating conditions can be unified into a comparable reference space for processing. Based on this, the drift trend of the conduction characteristics is determined and the fault confidence is calculated according to the current conduction characteristics and the reference conduction characteristics of the target home appliance under standard operating conditions. The health status of the target home appliance is determined based on the drift trend and the fault confidence. This achieves a quantitative representation of the gradual process of the home appliance from normal to fault, rather than simply making a binary judgment of "normal / fault". It has the ability to predict the potential deterioration trend of the health status of the home appliance, so that the diagnostic results can not only reflect whether a fault has occurred, but also characterize the evolution stage of the health status of the home appliance. Compared with existing methods that can only identify fixed known fault types in a preset database, this method can reveal the abnormal evolution of the target home appliance when the fault mode is not fully revealed and the fault type is not predefined, thus expanding the applicable boundary of home appliance fault diagnosis.

[0132] Device Examples Reference Figure 3 The diagram shows a logic block diagram of a home appliance fault diagnosis device 20 provided in an embodiment of the present invention. The device 20 may include: Feature acquisition module 201 is used to acquire the current conduction features of the conduction signal generated by the target home appliance based on heterogeneous sensing nodes, and map the current conduction features to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the conduction signal is used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances deployed outside the target home appliance and having sensing functions; The fault identification module 202 is used to determine the drift trend of the conduction characteristics and calculate the fault confidence level based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions. The status diagnosis module 203 is used to determine the health status of the target home appliance based on the drift trend and the fault confidence.

[0133] Optionally, the device 20 may further include: A smart home gateway access module is used to access the heterogeneous sensor node through a smart home gateway. The synchronous triggering module is used to respond to the target home appliance entering the standard operating condition by synchronously triggering the heterogeneous sensing node to collect the conducted signals generated by the target home appliance through the smart home gateway. The initial reference conduction feature acquisition module is used to acquire the conduction features based on the conduction signals collected by the different heterogeneous sensing nodes.

[0134] Optionally, the conducted signal includes the vibration signal of the target home appliance, and / or the acoustic signature signal of the home appliance; the initial reference conducted feature acquisition module includes: The spectral feature analysis submodule is used to acquire the vibration signals collected by different heterogeneous sensing nodes, and / or the acoustic signature signals, after fast Fourier transform to obtain the spectral feature distribution. The signal parameter submodule is used to acquire the amplitude attenuation ratio, arrival time difference, and phase offset of the vibration signal and / or the acoustic signal relative to the reference sensor node; the reference sensor node is the heterogeneous sensor node corresponding to the preset reference position.

[0135] Optionally, the feature acquisition module 201 includes: The fundamental frequency amplitude submodule is used to obtain the current fundamental frequency amplitude of the target home appliance and the reference fundamental frequency amplitude of the target home appliance under the standard operating conditions, and to normalize the current fundamental frequency amplitude and the reference fundamental frequency amplitude. The load deviation coefficient acquisition submodule is used to take the product of the load correction coefficient and the target ratio as the load deviation coefficient; the target ratio is the ratio of the normalized reference fundamental frequency amplitude to the normalized current fundamental frequency amplitude; The mapping execution submodule is used to take the product of the current conduction feature and the reciprocal of the load deviation coefficient as the current conduction feature mapped to the feature dimension corresponding to the standard operating condition.

[0136] Optionally, the fault identification module 202 may include: The drift slope acquisition submodule is used to fit the drift degree sequence of the conduction feature within a sliding time window formed by multiple standard operating condition cycles for each heterogeneous sensing node, and obtain the drift slope of the heterogeneous sensing node; the drift degree is used to characterize the degree of change of the current conduction feature relative to the reference conduction feature; the drift slope is used to characterize the rate of change of the conduction feature with the number of standard operating condition cycles; The spatial conduction consistency rate acquisition submodule is used to determine the spatial conduction consistency rate based on the drift slope of all the heterogeneous sensing nodes.

[0137] Optionally, the spatial propagation consistency rate acquisition submodule may include: The spatial conduction consistency rate acquisition unit is used to take the ratio of the total number of nodes drifting in the same direction to the total number of heterogeneous sensing nodes as the spatial conduction consistency rate. The term "same-direction drifting node" refers to the heterogeneous sensing nodes with a drift slope greater than the stability threshold, where the number of these nodes is greater than the stability threshold, and the number of these nodes is less than the stability threshold.

[0138] Optionally, the drift slope and the spatial conduction consistency rate are each configured with weights; the fault identification module may include: The local fault confidence acquisition submodule is used to take the weighted result of the drift slope and the spatial conduction consistency rate as the local fault confidence of the heterogeneous sensing node. The global fault confidence acquisition submodule is used to remove heterogeneous sensing nodes whose local fault confidence is less than a preset local fault confidence threshold, and to perform a weighted average of the local fault confidence of the remaining heterogeneous sensing nodes to obtain the global fault confidence.

[0139] Optionally, the status diagnosis module 203 may include: The first diagnostic submodule is used to determine that the target home appliance is in a slightly aging state when the spatial conduction consistency rate is greater than or equal to a first preset consistency rate threshold and the global fault confidence is within a first confidence interval. The second diagnostic submodule is used to determine that the target home appliance is in a moderately aging state when the spatial conduction consistency rate is greater than or equal to a second preset consistency rate threshold and the global fault confidence level is within a second confidence level interval; the second preset consistency rate threshold is greater than the first preset consistency rate threshold; and the minimum value of the second confidence level interval is greater than the maximum value of the first confidence level interval. The third diagnostic submodule is used to determine that the target home appliance is in a severely aged state when the spatial conduction consistency rate is greater than or equal to a third preset consistency rate threshold and the global fault confidence is within a third confidence interval; the third preset consistency rate threshold is greater than the second preset consistency rate threshold; and the minimum value of the third confidence interval is greater than the maximum value of the second confidence interval.

[0140] Optionally, device 20 further includes: The target dimension acquisition module for conduction features is used to determine the target dimension of the conduction feature that dominates the drift when the spatial conduction consistency rate is greater than a fourth preset consistency rate threshold; the fourth preset consistency rate threshold is greater than the first preset consistency rate threshold and less than the second preset consistency rate threshold. The faulty mechanical structure determination module is used to determine the target general mechanical structure that has failed, corresponding to the target dimension of the conduction feature, based on the correlation between the general mechanical structure and the drift law of the conduction feature; the target device includes at least one of the general mechanical structures.

[0141] Optionally, the device 20 may further include: The reference conduction feature update module is used to update the reference conduction feature within a sliding time window formed by multiple standard operating condition cycles. The sliding time window is used to accommodate a fixed number of conduction characteristics under the standard operating conditions; the reference conduction characteristic update module may include: The update quantity acquisition submodule is used to take the ratio of the difference between the first conduction feature most recently added to the sliding time window and the second conduction feature most recently removed from the sliding time window to the fixed quantity as the update quantity; The reference conduction feature update execution submodule is used to sum the reference conduction feature with the update amount to obtain the updated reference conduction feature.

[0142] In summary, the device control apparatus provided in this application provides a heterogeneous sensing node composed of home appliances with sensing functions deployed outside the target home appliance, which obtains the current transmission characteristics of the transmission signal generated by the target home appliance. By mapping the current transmission characteristics to the feature dimensions corresponding to the standard operating conditions of the target home appliance, the feature data collected under different operating conditions can be unified into a comparable reference space for processing. Based on this, the drift trend of the conduction characteristics is determined and the fault confidence is calculated according to the current conduction characteristics and the reference conduction characteristics of the target home appliance under standard operating conditions. The health status of the target home appliance is determined based on the drift trend and the fault confidence. This achieves a quantitative representation of the gradual process of the home appliance from normal to fault, rather than simply making a binary judgment of "normal / fault". It has the ability to predict the potential deterioration trend of the health status of the home appliance, so that the diagnostic results can not only reflect whether a fault has occurred, but also characterize the evolution stage of the health status of the home appliance. Compared with existing methods that can only identify fixed known fault types in a preset database, this method can reveal the abnormal evolution of the target home appliance when the fault mode is not fully revealed and the fault type is not predefined, thus expanding the applicable boundary of home appliance fault diagnosis.

[0143] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0144] Reference Figure 4 This is a structural block diagram of an electronic device for device control provided in an embodiment of this application. Figure 4 As shown, the electronic device includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the method as described in the foregoing embodiments.

[0145] The processor can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable devices, transistor logic devices, hardware components, or any combination thereof. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, or a combination of a DSP and a microprocessor.

[0146] The memory may be ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or it may be EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory), magnetic tape, floppy disk, and optical data storage devices, etc.

[0147] This application also provides a non-transitory computer-readable storage medium that, when the instructions in the storage medium are executed by a processor of an electronic device (server or terminal), enables the processor to perform the methods of the foregoing embodiments.

[0148] This application also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0149] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0150] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0151] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0152] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0153] These computer program instructions may also be stored in a computer-readable storage medium capable of directing a computer or other programmable data processing terminal device to operate in a predictive manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0154] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0155] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0156] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0157] The present application provides a detailed description of a method, apparatus, electronic device, and readable storage medium for diagnosing household appliances. Specific examples have been used to illustrate the principles and implementation methods of the present application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present application. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of the present application. Therefore, the content of this specification should not be construed as a limitation of the present application.

Claims

1. A method for diagnosing faults in household appliances, characterized in that, The method includes: Based on heterogeneous sensing nodes, the current conduction characteristics of the conducted signals generated by the target home appliance are obtained, and the current conduction characteristics are mapped to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the conducted signals are used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances with sensing functions deployed outside the target home appliance. Based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions, the drift trend of the conduction characteristics is determined, and the fault confidence is calculated. Based on the drift trend and the fault confidence level, the health status of the target home appliance is determined.

2. The method according to claim 1, characterized in that, The method further includes: The heterogeneous sensing nodes are accessed via a smart home gateway; In response to the target home appliance entering the standard operating condition, the smart home gateway synchronously triggers the heterogeneous sensor node to collect the conducted signals generated by the target home appliance. The conduction characteristics are obtained based on the conduction signals collected by the different heterogeneous sensing nodes.

3. The method according to claim 2, characterized in that, The conducted signal includes the vibration signal of the target home appliance and / or the acoustic signature signal of the home appliance; the step of obtaining the conducted features based on the conducted signals collected by the different heterogeneous sensing nodes includes: The vibration signals collected by different heterogeneous sensing nodes and / or the acoustic signature signals are obtained by fast Fourier transform to obtain the spectral feature distribution. The vibration signal and / or acoustic signature signal are acquired relative to the reference sensor node, and the amplitude attenuation ratio, arrival time difference, and phase offset are collected. The reference sensor node is the heterogeneous sensor node corresponding to the preset reference position.

4. The method according to claim 1, characterized in that, The step of mapping the current transmission feature to the feature dimension corresponding to the standard operating condition of the target home appliance includes: The current fundamental frequency amplitude of the target home appliance and the reference fundamental frequency amplitude of the target home appliance under the standard operating conditions are obtained respectively, and the current fundamental frequency amplitude and the reference fundamental frequency amplitude are normalized. The product of the load correction factor and the target ratio is used as the load deviation factor; the target ratio is the ratio of the normalized reference fundamental frequency amplitude to the normalized current fundamental frequency amplitude; The product of the current conduction feature and the reciprocal of the load deviation coefficient is used as the current conduction feature mapped to the feature dimension corresponding to the standard operating condition.

5. The method according to claim 1, characterized in that, The step of determining the drift trend of the conduction characteristics based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions includes: For each heterogeneous sensing node, within a sliding time window formed by multiple standard operating condition cycles, the drift degree sequence of the conductive feature is fitted to obtain the drift slope of the heterogeneous sensing node; the drift degree is used to characterize the degree of change of the current conductive feature relative to the reference conductive feature; the drift slope is used to characterize the rate of change of the conductive feature with the number of standard operating condition cycles; The spatial conduction consistency rate is determined based on the drift slope of all the heterogeneous sensing nodes.

6. The method according to claim 5, characterized in that, The determination of spatial conduction consistency based on the drift slopes of all the heterogeneous sensing nodes includes: The ratio of the total number of nodes drifting in the same direction to the total number of heterogeneous sensing nodes is used as the spatial conduction consistency rate. The term "same-direction drifting node" refers to the heterogeneous sensing nodes with a drift slope greater than the stability threshold, where the number of these nodes is greater than the stability threshold, and the number of these nodes is less than the stability threshold.

7. The method according to claim 5, characterized in that, The drift slope and the spatial conduction consistency rate are each assigned a weight; The calculation of the fault confidence includes: The weighted result of the drift slope and the spatial conduction consistency rate is used as the local fault confidence of the heterogeneous sensing node; Remove heterogeneous sensing nodes whose local fault confidence is less than a preset local fault confidence threshold, and then perform a weighted average of the local fault confidence of the remaining heterogeneous sensing nodes to obtain the global fault confidence.

8. The method according to claim 7, characterized in that, Determining the health status of the target home appliance based on the drift trend and the fault confidence includes: If the spatial conduction consistency rate is greater than or equal to the first preset consistency rate threshold, and the global fault confidence is within the first confidence interval, the target home appliance is determined to be in a slightly aging state. If the spatial conduction consistency rate is greater than or equal to the second preset consistency rate threshold, and the global fault confidence level is within the second confidence level interval, the target home appliance is determined to be in a moderately aging state; the second preset consistency rate threshold is greater than the first preset consistency rate threshold; the minimum value of the second confidence level interval is greater than the maximum value of the first confidence level interval; If the spatial conduction consistency rate is greater than or equal to a third preset consistency rate threshold, and the global fault confidence level is within a third confidence level interval, the target home appliance is determined to be in a severely aged state; the third preset consistency rate threshold is greater than the second preset consistency rate threshold; and the minimum value of the third confidence level interval is greater than the maximum value of the second confidence level interval.

9. The method according to claim 8, characterized in that, The method further includes: If the spatial conduction consistency rate is greater than the fourth preset consistency rate threshold, the target dimension of the conduction feature that dominates the drift is determined; the fourth preset consistency rate threshold is greater than the first preset consistency rate threshold and less than the second preset consistency rate threshold. Based on the correlation between the general mechanical structure and the drift law of the conduction feature, the target general mechanical structure corresponding to the target dimension of the conduction feature is determined, and the target device includes at least one of the general mechanical structures.

10. The method according to claim 1, characterized in that, The method further includes: The reference conduction characteristics are updated within a sliding time window formed by multiple standard operating condition cycles; The sliding time window is used to accommodate a fixed number of conduction characteristics under the standard operating conditions; within the sliding time window formed by multiple standard operating condition cycles, the reference conduction characteristics are updated, including: The ratio of the difference between the first conduction feature most recently added to the sliding time window and the second conduction feature most recently removed from the sliding time window to the fixed number is used as the update amount; The sum of the reference transmission feature and the update amount is used as the updated reference transmission feature.

11. A fault diagnosis device for household appliances, characterized in that, The device includes: The feature acquisition module, based on heterogeneous sensing nodes, acquires the current conduction features of the conducted signals generated by the target home appliance, and maps the current conduction features to the feature dimensions corresponding to the standard operating conditions of the target home appliance; the conducted signals are used to characterize the operating state of the target home appliance; the heterogeneous sensing nodes are composed of home appliances deployed outside the target home appliance and having sensing functions; The fault identification module is used to determine the drift trend of the conduction characteristics and calculate the fault confidence level based on the current conduction characteristics and the reference conduction characteristics of the conduction signal of the target home appliance under the standard operating conditions. The status diagnosis module is used to determine the health status of the target home appliance based on the drift trend and the fault confidence level.

12. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the method as described in any one of claims 1 to 10.

13. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the method as described in any one of claims 1 to 10.