Smart home state diagnosis method based on life prediction and reliability analysis

Abstract

The invention relates to the technical field of fault diagnosis and life prediction of intelligent furniture, and discloses an intelligent home state diagnosis method based on life prediction and reliability analysis. The operation life of the whole circuit in the intelligent furniture is predicted and the reliability of the whole circuit in the intelligent furniture is evaluated by analyzing the electric leakage signal conditioning circuit, so that a circuit system in the intelligent furniture is accurately and comprehensively diagnosed, and the operation safety of the intelligent furniture system is improved; and a powerful guarantee is provided for stable operation of the intelligent furniture system. The technical problems of low fault detection precision, limited detection range, poor anti-interference capability, high data dependence and the like in the field of intelligent furniture systems are mainly solved, and the actual application requirements of intelligent furniture products can be better met.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis of smart homes, and in particular to a smart home status diagnosis method based on life prediction and reliability analysis. Background Art

[0002] With the continuous development of big data and artificial intelligence technologies, smart furniture has become a very common product in people's daily lives, and the fault diagnosis and service life prediction technologies of smart furniture have also been more and more widely applied to products. For example, by analyzing the switch records of smart locks, the motor operation data of smart curtains, etc., the fault trend and service life of the device can be predicted, and maintenance can be carried out in advance.

[0003] The existing smart furniture systems have initially formed a system in terms of fault diagnosis and life prediction. With the continuous development of technologies such as big data and artificial intelligence, the accuracy and reliability of these technologies have been continuously improved. At the same time, corresponding problems have also arisen. Smart furniture requires corresponding systems and methods to complete both fault diagnosis and service life prediction. Its algorithms are complex and require a large amount of computing resource support, resulting in high costs for initial development and deployment; the operation data of smart homes may be affected by various factors, such as sensor accuracy, data transmission stability, etc., which may lead to a decrease in the accuracy and reliability of the data; in addition, while smart homes collect user data, it also brings risks of privacy leakage and data security.

[0004] To solve the above technical problems, those skilled in the art analyzed the fault detection method by combining the spatial correlation of perception data and the correlation between multiple attributes; first, determined the abnormal source of the fault node; secondly, aiming at the defects of large communication volume or the need for prior knowledge in the node fault detection algorithm, the solution proposed a solution combining self-detection and similarity detection, and determined the self-state of the fault node through majority voting and similarity judgment with known state nodes; then, an autoregressive model data aggregation algorithm was used to obtain the aggregation result by fitting measurement values, calculating prediction values and the measurement variance of actual values. The feature of this solution is to combine two algorithms of node fault detection with spatio-temporal correlation and data aggregation based on autoregressive model, design and implement a prototype system for node soft fault detection and processing; use temperature sensors to collect data, judge the sensor state, and perform aggregation processing on the measurement values.

[0005] However, the above technology still has the following problems:

[0006] 1) Limited detection range: The above technology may not be able to comprehensively cover all devices and systems in smart furniture, resulting in some faults not being detected in time;

[0007] 2) Insufficient real-time performance: The real-time performance of fault diagnosis is crucial for the maintenance of smart furniture, but the above technologies cannot provide timely fault diagnosis in all cases;

[0008] 3) Poor anti-interference ability: Smart homes are usually deployed in a home environment and may be interfered by various environmental factors, such as electromagnetic interference, physical obstacles, etc. These interferences may affect the accuracy of fault diagnosis and may also cause the original fault diagnosis technology to fail or require frequent updates;

[0009] 4) Dependence on a large amount of data: The above technologies cannot efficiently process data, resulting in low diagnostic efficiency and unable to meet the actual usage requirements of smart furniture. Summary of the Invention

[0010] The purpose of the present invention is to solve the above problems and design a smart home status diagnosis method based on life prediction and reliability analysis.

[0011] To achieve the above purpose, the technical solution of the present invention is as follows:

[0012] A smart home status diagnosis method based on life prediction and reliability analysis, characterized in that the method includes the following steps:

[0013] Step 1: Taking the leakage signal conditioning circuit in the smart home as the target, assuming that the degradation process follows a linear drift + random noise model, using the accelerated experiment temperature as the accelerated stress, and using the operating current value of the leakage signal conditioning circuit as the degradation data representation, constructing a circuit degradation data acquisition model for the smart home and obtaining the circuit degradation data;

[0014] Step 2: Based on the circuit degradation data and the Wiener process, establish the probability density function of the performance degradation process of the leakage signal conditioning circuit in the smart home, and then use the maximum likelihood estimation method to solve the unknown parameters in the probability density function;

[0015] Step 3: Use the probability density function of the performance degradation process to predict the operating life of the leakage signal conditioning circuit in the smart home, then predict the time length when the operating current value in the circuit exceeds the failure threshold according to the current operating current value and the degradation rate in the smart home, and construct the remaining life distribution function of the overall circuit in the smart home according to this time length to obtain the life prediction result of the overall circuit in the smart home;

[0016] Step 4: Based on the remaining life distribution function of the overall circuit in the smart home and the Arrhenius equation, deduce the conversion relationship of the circuit degradation data under normal stress conditions and obtain the normal remaining life distribution function f of the overall circuit in the smart home under normal stress conditions L(t; T0), and then obtain the real-time reliability of the overall circuit in the smart home based on the normal predicted lifetime distribution function;

[0017] Step Five, diagnose the operating state of the overall circuit in the smart home based on the real-time normal lifetime prediction result of the overall circuit in the smart home, and then perform self-regulation or remote regulation on the operating state of the smart home according to the diagnosis result and the reliability.

[0018] It should be noted that the diagnosis result is obtained through logical judgment based on the normal lifetime prediction result;

[0019] Among them, the main function of the normal lifetime prediction result is to provide a judgment basis for the regulation of the operating state of the smart home (including self-intelligent regulation or remote manual regulation); when the normal lifetime prediction result is close to the aging limit of the overall circuit in the smart home, the smart home can adjust the relationship between its working time and working current through regulation (including self-intelligent regulation or remote manual regulation), and the smart home can control the magnitude and on / off of its internal working current according to the magnitude of the working time. For example, when the normal lifetime prediction result is close to the aging limit of the overall circuit in the smart home, the main control program of the smart home can regulate the length of the working time and the magnitude of the working current according to the degree to which the normal lifetime prediction result is close to the aging limit. The length value of the working time and the value of the working current of the smart home are both inversely proportional to the degree to which the normal lifetime prediction result is close to the aging limit. That is to say, the closer the normal lifetime prediction result is to the aging limit, the shorter the working time of the smart home, the smaller the working current of the smart home, and the smart home enters a power-saving, low-intensity operation mode or enters a shutdown mode.

[0020] It should be noted that the aging limit is the time length set manually; if the time length of the normal lifetime prediction result is close to but greater than the time length of the aging limit, the smart home can ensure normal operation by adjusting the working duration and working current; if the time length of the normal lifetime prediction result is less than the time length of the aging limit, the smart home will enter the shutdown mode for self-protection.

[0021] The main role of reliability in this application is to provide a basis for judging the switching of the regulation mode of the smart home (mainly the switching between the self-intelligent regulation mode and the remote manual regulation mode). That is, during the operation of the smart home, if the normal life prediction result is close to the aging limit of the overall circuit in the smart home, the main control program of the smart home will automatically switch the regulation mode according to the size of the reliability; when the reliability is higher than the set standard, the smart home adopts the self-intelligent regulation method, which does not require manual intervention and will not issue an alarm; when the reliability is lower than the set standard, the main control program of the smart home will send an alarm signal to the remote control center through the network and automatically switch to the remote manual regulation mode, and the maintenance personnel can adjust its operation status according to the situation of the smart home or visit for maintenance.

[0022] The construction process of the circuit degradation data acquisition model in the first step includes:

[0023] Under the accelerated experimental temperature stress T, the action current value of the leakage signal conditioning circuit is collected as the degradation data, then the circuit degradation data acquisition model is:

[0024] X(t) = x0 + β(T)·t + σ W ·W(t) (1)

[0025] In the formula, X(t) is the action current value at time t, x0 is the initial action current value, β(T) is the degradation rate at temperature T, σ W is the diffusion coefficient, and W(t) is the standard Brownian motion.

[0026] The construction process of the probability density function of the performance degradation process in the second step is:

[0027] Based on the circuit degradation data, and the performance degradation process is a time-varying drift Wiener process, then the probability density function of the performance degradation process is:

[0028]

[0029] The solution process of the unknown parameters in the probability density function of the performance degradation process in the second step is to estimate the unknown parameters using the maximum likelihood estimation method. Based on the circuit degradation data, that is, the degradation data X(t i ), a log-likelihood function is constructed:

[0030]

[0031] In the formula, L is the predicted life of the leakage signal conditioning circuit; N is the total number of action current degradation data points collected; maximizing lnL can solve the unknown parameters β(T) and σ W ;

[0032] The process of predicting the overall circuit operation life in the smart home by using the probability density function of the performance degradation process in Step 3 is as follows:

[0033] First, define the failure threshold as D. Then the value of L is the time length when the degradation process first reaches D, that is:

[0034] L = inf{t ≥ 0|X(t) ≥ D} (4)

[0035] Second, based on the analytical solution of the Wiener process, and assuming that the life distribution of the overall circuit in the smart home under the accelerated stress condition follows the inverse Gaussian distribution, the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition is:

[0036]

[0037] Finally, substitute the operating current value X(t c ) at the current time t c into the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition to obtain the remaining life distribution f L (t - t c ). Combine the estimated β(T) and σ W to calculate the remaining life distribution f L (t - t c ), and take the mean or quantile of the remaining life distribution f L (t - t c ). Use the mean and quantile as the life prediction results of the overall circuit in the smart home under the accelerated stress condition.

[0038] The conversion relationship of the circuit degradation data under the normal stress condition is derived based on the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition and the Arrhenius equation in Step 4 as follows:

[0039]

[0040] In the formula, E a is the activation energy, k is the Boltzmann constant, T is the accelerated test temperature, and T0 is the normal use temperature.

[0041] The normal predicted life distribution function f L (t; T0) of the overall circuit inside the smart home reconstructed in Step 4 is:

[0042]

[0043] The reliability R(t1) in Step 4 is the probability that the life exceeds the time length t1, and its calculation formula is:

[0044]

[0045] Compared with the prior art, the present application has the following advantages:

[0046] 1. The present application adopts high-precision random degradation modeling, which is different from the traditional method that assumes the degradation process is purely deterministic (such as linear) or simply random (white noise). By using the Wiener process to simultaneously characterize the trend drift and Brownian motion randomness, it is closer to the actual circuit aging behavior (such as the cumulative effects of oxidation and material fatigue). The measurement of a certain type of leakage protection circuit shows that the coverage rate of the model for the action current fluctuation range reaches 98%, while the traditional linear model only covers 65%. In the 1000-hour accelerated test, the correlation coefficient R2 between the model's predicted life and the true first failure time is 0.93, significantly higher than the R2 of 0.68 of the exponential model.

[0047] 2. The present application adopts a physical-statistical fusion cross-temperature life prediction method, while the prior art usually directly extrapolates the results of the accelerated test (such as the Arrhenius model is only used for calculating the acceleration factor) without coupling with the degradation kinetic equation. The present application embeds the temperature stress into the drift rate β(T) of the Wiener process through the Arrhenius equation, realizing the deep combination of physical mechanisms and statistical models. In the degradation rate conversion from 85°C to 25°C, the average error of the traditional method is 28%, while the error of this model is <5%. It can quantify the effects of other stresses such as radiation and humidity (by extending the Eyring equation), and thus better meet the requirements of multi-environment applications in smart homes.

[0048] 3. The present application can realize the dynamic switching of the control method by combining reliability analysis on the basis of the normal life prediction results of smart homes, and switch the control method depending on the change of reliability (that is, the switching between the self-intelligent control method and the remote manual control method); while the traditional maintenance strategy is based on a fixed cycle or a single threshold, its maintenance timing is inaccurate and the maintenance cost is high; in addition, the present application can also realize hierarchical early warning by using reliability (such as R(t) < 95% prompts inspection, R(t) < 80% forces maintenance), thereby improving the user experience. Description of the Drawings

[0049] Figure 1 is a flowchart of a smart home status diagnosis method based on life prediction and reliability analysis according to the present invention;

[0050] Figure 2 is a schematic diagram of a leakage signal conditioning circuit according to the present invention;

[0051] Figure 3 is a parameter definition table of a smart home status diagnosis method based on life prediction and reliability analysis according to the present invention;

[0052] Figure 4 It is a comparative analysis table of the present invention and the prior art in terms of method content;

[0053] Figure 5 It is a comparative analysis table of the present invention and the prior art in terms of economic benefits;

[0054] Figure 6 It is a diagnostic result curve graph of a smart home status diagnosis method based on life prediction and reliability analysis according to the present invention. Detailed implementation manners

[0055] The present invention will be specifically described below with reference to the accompanying drawings, as Figure 1-6 shown;

[0056] A smart home status diagnosis method based on life prediction and reliability analysis, the method comprising the following steps:

[0057] Step 1, taking the leakage signal conditioning circuit in the smart home as the target, assuming that the degradation process follows a linear drift + random noise model, taking the accelerated test temperature as the acceleration stress, and the accelerated test temperature is a high temperature set according to experimental requirements, generally 125°C, that is, the accelerated test temperature is the acceleration stress under experimental conditions or theoretical conditions, and it is not affected by other factors. Then, taking the operating current value of the leakage signal conditioning circuit as the representation of degradation data, constructing a circuit degradation data acquisition model for the smart home and obtaining circuit degradation data;

[0058] Among them, the acquisition process of the circuit degradation data is as follows:

[0059] First, set multiple groups of accelerated temperatures T1, T2,..., T n (such as 85°C, 105°C, 125°C... etc.);

[0060] Then, test the circuit degradation data acquisition model under each temperature condition, and record the sequence of the operating current changing with time each time after the test is completed, that is, X(t i ), X(t i ) is the operating current value recorded after the i-th test is completed;

[0061] It should be noted that the change of the operating current value of the leakage signal conditioning circuit can better reflect the degradation of the entire circuit in the smart home. Therefore, the present application selects the operating current value of the leakage signal conditioning circuit as the standard and basis for measuring the degradation degree of the internal circuit of the smart home;

[0062] Step 2, based on the circuit degradation data and the Wiener process, establish the probability density function of the performance degradation process of the leakage signal conditioning circuit in the smart home, and then use the maximum likelihood estimation method to solve the unknown parameters in the probability density function;

[0063] Step 3: Use the probability density function of the performance degradation process to predict the operating life of the leakage signal conditioning circuit in the smart home. Then, based on the current operating current value and the degradation rate in the smart home, predict the time length when the operating current value in the circuit exceeds the failure threshold, and construct the remaining life distribution function of the overall circuit in the smart home according to this time length to obtain the life prediction result of the overall circuit in the smart home;

[0064] Step 4: Based on the remaining life distribution function of the overall circuit in the smart home and the Arrhenius equation, deduce the conversion relationship of the circuit degradation data under normal stress conditions and obtain the normal remaining life distribution function f L (t; T0) of the overall circuit in the smart home under normal stress conditions, that is, reconstruct the normal prediction life distribution function of the overall circuit inside the smart home based on the normal stress (normal operating temperature T0) and the normal circuit degradation data under normal stress conditions (degradation rate β(T0) at temperature T0), and then obtain the real-time reliability of the overall circuit in the smart home based on the normal prediction life distribution function;

[0065] Step 5: Diagnose the operating state of the overall circuit in the smart home based on the real-time reliability of the overall circuit in the smart home and the normal life prediction result, and then remotely regulate the operating state of the smart home according to the diagnosis result.

[0066] The creative point of the present invention lies in fitting the Wiener process parameters β(T) and σ through the accelerated degradation test data W , establishing a stochastic process description of the degradation trajectory, and then using the Arrhenius equation to convert the degradation rate β(T) at high temperature to β(T0) at normal temperature T0 to achieve equivalent life prediction from the accelerated test to the actual working condition; finally, combining the inverse Gaussian life distribution with the reliability function R(t), real-time evaluate the remaining life of the circuit and perform remote regulation (such as diagnosing potential faults when R(t) < 95%).

[0067] The construction process of the circuit degradation data acquisition model in Step 1 includes:

[0068] Under the accelerated experimental temperature stress T, collect the operating current value of the leakage signal conditioning circuit as the degradation data. Then, the circuit degradation data acquisition model is:

[0069] X(t) = x0 + β(T)·t + σ W ·W(t) (1)

[0070] Wherein, X(t) is the operating current value at time t, which is used as the degradation feature quantity, with the unit of A; x0 is the initial operating current value, the current value in the normal state, with the unit of A; β(T) is the degradation rate at temperature T, which is related to the temperature, with the unit of A / hour (ampere per hour); σ W is the diffusion coefficient, mainly used to characterize the random fluctuation intensity, with the unit of A / √hour (ampere per square root hour); W(t) is the standard Brownian motion, whose mean value is 0 and variance is t.

[0071] The construction process of the probability density function of the performance degradation process in step two is as follows:

[0072] Based on the circuit degradation data and the performance degradation process being a time-varying drift Wiener process, the probability density function of the performance degradation process is:

[0073]

[0074] In the solution process of the unknown parameters in the probability density function of the performance degradation process in step two, the maximum likelihood estimation method is used to estimate the unknown parameters. Based on the circuit degradation data, that is, the degradation data X(t i ), a log-likelihood function is constructed:

[0075]

[0076] Wherein, L is the predicted life of the leakage signal conditioning circuit; N is the total number of collected operating current degradation data points. Specifically: under each temperature stress condition, the operating current values X(t1), X(t2), …, X(t N ) are recorded in time series, and a total of N observation data points are obtained. Each data point corresponds to a time t i (i = 1, 2, …, N), that is, N times of degradation data collection are carried out; maximizing lnL can solve the unknown parameters β(T) and σ W ;

[0077] It should be noted that the value of N can reflect the data volume size and directly affect the accuracy of parameter estimation; the larger N is, the closer the maximum likelihood estimation result is to the true value (law of large numbers); in formula (3), the first term is proportional to the data volume N and penalizes the model complexity (similar to the information criterion); the second summation is the weighted sum of the squared residuals of all N data points, measuring the model fitting degree. During the process of collecting degradation data, it is usually necessary to ensure that N ≥ 30 to meet the statistical significance requirement. If the sampling interval is Δt and the total test duration is t total , then N = t total / Δt (for example, sampling once every 10 minutes and conducting the test for 100 hours, then N = 600).

[0078] The process of predicting the overall circuit operation life in the smart home by using the probability density function of the performance degradation process in step three is as follows:

[0079] First, define the failure threshold as D (for example, when the operating current exceeds 1.5 A, it is considered a failure), then the value of L is the time length when the degradation process first reaches D, that is:

[0080] L = inf{t ≥ 0|X(t) ≥ D} (4)

[0081] Second, based on the analytical solution of the Wiener process and assuming that the life distribution of the overall circuit in the smart home under the accelerated stress condition follows the inverse Gaussian distribution, the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition is:

[0082]

[0083] Finally, substitute the operating current value X(t c ) at the current time t c into the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition to obtain the remaining life distribution f L (t - t c ). Combine the estimated β(T) and σ W to calculate the remaining life distribution f L (t - t c ), and take the mean or quantile of the remaining life distribution f L (t - t c ). Use the mean and quantile as the life prediction results of the overall circuit in the smart home under the accelerated stress condition.

[0084] The conversion relationship of the circuit degradation data under the normal stress condition is derived based on the remaining life distribution function of the overall circuit in the smart home under the accelerated stress condition and the Arrhenius equation in step four:

[0085] Mapping the degradation rate β(T) under the accelerated stress to the normal temperature T0 (such as 25 °C) gives:

[0086]

[0087] In the formula, E a is the activation energy (material property, unit: eV), k is the Boltzmann constant (8.617×10 -5 eV / K), T is the accelerated test temperature (unit: K), and T0 is the normal operating temperature (unit: K).

[0088] In Step 4, the normal predicted life distribution function f L (t; T0) is as follows:

[0089]

[0090] In Step 4, the reliability R(t1) is the probability that the life exceeds the duration t1, and its calculation formula is:

[0091]

[0092] Example 1

[0093] Scenario: Life prediction of the leakage protection circuit of the smart home

[0094] Accelerated test conditions: T = 125°C, and β(125°C) = 0.02 A / hour, σW = 0.005 A / √hour are measured;

[0095] Normal use conditions: T0 = 25°C, activation energy Ea = 0.8 eV;

[0096] Then the life conversion calculation:

[0097]

[0098] If the current operating current X(t c ) = 1.2 A (failure threshold D = 1.5 A), then the predicted median remaining life:

[0099]

[0100] Therefore, the staff will obtain the diagnostic result of "maintenance is required within 8 days", and then use t 50 as t1, and substitute the result of t 50 (200 hours) into the calculation formula of the reliability R(t1) to determine the probability that the overall circuit life in the smart home exceeds t 50 . Furthermore, determine that the regulation method of the smart home is remote manual regulation and send an alarm and regulation request to the command center. The staff in the control center can remotely control the magnitude of the load current and adjust the working state of the smart home according to the predicted median remaining life t 50 and the value of the reliability R(t1), so as to extend the service life of the smart home and control the operation risk of the smart home; the mathematical model system established by this method realizes the accurate diagnosis of smart home circuit faults through the combination of physical degradation mechanism and statistical inference, and at the same time intelligently switches the regulation method of the smart home through reliability. The technical solution of this application is significantly better than the traditional regular maintenance or threshold alarm method.

[0101] The above technical solution only reflects the preferred technical solution of the technical solution of the present invention. Some changes that those skilled in the art of the present technology may make to some parts thereof all reflect the principle of the present invention and fall within the protection scope of the present invention.

Claims

1. A smart home status diagnosis method based on life prediction and reliability analysis, characterized in that: The method comprises the following steps: Step 1: Accelerate the degradation process by targeting the leakage signal conditioning circuit in the smart home, then use the action current value of the leakage signal conditioning circuit as the degradation data representation, build a circuit degradation data collection model for the smart home and obtain circuit degradation data; Step 2: Based on the circuit degradation data and the Wiener process, a probability density function of the performance degradation process of the leakage signal conditioning circuit in the smart home is established, and then the maximum likelihood estimation method is used to solve the unknown parameters in the probability density function; Step three, using the probability density function of the performance degradation process to predict the operating life of the leakage signal conditioning circuit in the smart home, and then predicting the time length that the action current value in the circuit exceeds the failure threshold according to the current action current value and degradation rate in the smart home, and constructing the remaining life distribution function of the overall circuit in the smart home according to the time length to obtain the life prediction result of the overall circuit in the smart home; Step 4: Based on the remaining life distribution function of the overall circuit in the smart home and the Arrhenius equation, the conversion relationship of the circuit degradation data under normal stress conditions is derived and the normal remaining life distribution function of the overall circuit in the smart home under normal stress conditions is obtained, and then the reliability of the overall circuit in the smart home is obtained based on the normal predicted life distribution function; Step five: diagnose the operating status of the entire circuit in the smart home based on the normal life prediction results, and then remotely control the operating status of the smart home based on the diagnosis results and reliability.

2. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 1, characterized in that: The circuit degradation data acquisition model in step 1 is: X(t)=x0+β(T)·t+σ W ·W(t) (1) Where X(t) is the operating current value at time t, x0 is the initial operating current value, β(T) is the degradation rate at temperature T, σ W is the diffusion coefficient, and W(t) is the standard Brownian motion.

3. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 2, characterized in that: The probability density function of the performance degradation process in step 2 is:

4. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 3, characterized in that: The process of solving the unknown parameters by using the maximum likelihood estimation method in step 2 is to construct a log-likelihood function: Where L is the predicted life of the leakage signal conditioning circuit, and N is the total number of collected action current degradation data points. Maximizing lnL can solve the unknown parameters β(T) and σ W .

5. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 4, characterized in that: The life prediction process of the overall circuit in the smart home in step 3 is as follows: First, define the failure threshold as D, then the value of L is the length of time it takes for the degradation process to reach D for the first time, that is: L=inf{t≥0|X(t)≥D} (4) Secondly, based on the analytical solution of the Wiener process, and assuming that the lifetime distribution of the overall circuit in the smart home under accelerated stress conditions obeys the inverse Gaussian distribution, the remaining lifetime distribution function of the overall circuit in the smart home under accelerated stress conditions is: Finally, the current time t c The operating current value X(t c ) is brought into the remaining life distribution function of the overall circuit in the smart home under accelerated stress conditions to obtain the remaining life distribution f L (tt c ), for the remaining life distribution f L (tt c ) is calculated and the mean or quantile is taken, and the mean and quantile are used as the life prediction results of the overall circuit in the smart home under accelerated stress conditions.

6. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 5, characterized in that: The conversion relationship of the circuit degradation data in step 4 under normal stress conditions is: In the formula, E a is the activation energy, k is the Boltzmann constant, T is the accelerated test temperature, and T0 is the normal use temperature.

7. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 6, characterized in that: The normal predicted life distribution function f of the overall circuit inside the smart home in step 4 is L (t; T0) is:

8. The method for diagnosing smart home status based on life prediction and reliability analysis according to claim 7, characterized in that: The reliability R(t1) in step 4 is the probability that the lifespan exceeds the duration t1, and its calculation formula is:

Images