Equipment reliability prediction method, device, electronic device and storage medium based on q-weibull distribution

Abstract

The present invention relates to equipment reliability technology, and provides an equipment reliability prediction method, device, electronic device and storage medium based on q-weibull distribution, wherein the method comprises: obtaining the estimated operating time of a target device; inputting the estimated operating time into a reliability prediction model pre-constructed based on q-weibull distribution, and obtaining the probability of failure when the operating time of the target device reaches the estimated operating time; the model parameters of the reliability prediction model are determined after iterative optimization of the parameter distribution of the model parameters based on a preset mean and a preset standard deviation. The present invention can quickly determine the model parameters with high precision, and improves the accuracy of the prediction results of the reliability prediction model.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment reliability, and in particular to a method, a device, an electronic device and a storage medium for predicting equipment reliability based on q-weibull distribution. Background Art

[0002] The life distribution model is a commonly used reliability analysis tool. By fitting the historical failure data of the equipment, it can predict the failure probability of the equipment in the future. Its advantage is that it can describe the early failure period, accidental failure period and wear and tear failure period of the equipment, and is suitable for reliability analysis of various types of equipment.

[0003] In reliability analysis, exponential distribution is widely used in electronic devices. GB / T15214-2008 Ultrasonic Diagnostic Equipment Reliability Test Requirements and Methods are formulated on the basis that the failure law obeys the exponential distribution. However, the premise of constructing a life distribution model based on the exponential distribution is that the failure rate is constant, ignoring important realistic factors such as operator proficiency and operating environment. In order to accurately analyze the reliability of equipment, reliability engineers usually use a failure rate function with bathtub characteristics to construct a model to describe the process of equipment reliability level degradation. Among them, Weibull distribution has been widely used because it can describe the three stages of the bathtub curve.

[0004] However, the Weibull distribution cannot use a set of parameters to simultaneously describe the various stages of the failure rate function with bathtub characteristics, which limits the application of the Weibull distribution. To this end, the Weibull distribution is extended by introducing new parameters to solve this problem, and the Q-Weibull distribution is one of the extension methods. However, in the prior art, the Q-Weibull distribution parameter estimation has a large amount of computation and poor fitting accuracy, which affects the accuracy of the reliability prediction results. Summary of the invention

[0005] The present invention aims to provide a method, device, electronic device and storage medium for predicting equipment reliability based on q-weibull distribution.

[0006] The embodiments of the present invention can be implemented as follows:

[0007] In a first aspect, the present invention provides a method for predicting equipment reliability based on q-weibull distribution, the method comprising:

[0008] Get the estimated runtime of the target device;

[0009] The estimated operating time is input into a reliability prediction model pre-constructed based on q-weibull distribution to obtain the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined after iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation.

[0010] In an optional implementation manner, the model parameters are determined as follows:

[0011] Acquire failure data of each of the sample products of a first preset number when the failure occurs for the first time, wherein all the sample products are of the same model as the target device;

[0012] Using the preset mean and the preset standard deviation as the current mean and the current standard deviation respectively;

[0013] Determine a parameter distribution of the model parameter according to the current mean and the current standard deviation and generate a second preset number of current parameters that conform to the parameter distribution, wherein the second preset number is greater than the first preset number;

[0014] Calculate the total error of the predicted value obtained based on each of the current parameters relative to all fault data;

[0015] Determining a plurality of target parameters from the second preset number of current parameters according to the total errors of all the current parameters;

[0016] Calculate the parameter mean and parameter standard deviation of all the target parameters;

[0017] Resetting the parameter mean as the current mean and resetting the parameter standard deviation as the current standard deviation, returning to the step of determining the parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution, until a preset condition is met;

[0018] The model parameters are determined according to the multiple target parameters obtained last time.

[0019] In an optional implementation, the step of calculating the total error of the predicted value obtained based on each of the current parameters relative to all fault data includes:

[0020] For any target current parameter, calculating an intermediate error between each of the fault data and the target current parameter;

[0021] Sum all the intermediate errors to obtain a total error of the predicted value obtained based on the target current parameter prediction relative to all the fault data;

[0022] Each of the current parameters is used as the target current parameter in turn to obtain a total error of a predicted value obtained based on each of the current parameters relative to all fault data.

[0023] In an optional implementation, the fault data includes the running time and the fault probability, the current parameter includes a scale parameter and a first shape parameter and a second shape parameter for determining the shape of the fault rate function, and for any target fault data among all the fault data, the method of calculating the intermediate error between the target fault data and the target current parameter is:

[0024] Using the formula Calculate the intermediate error between the target fault data and the target current parameter, where: is the intermediate error between the target fault data and the target current parameter, is the failure probability of the target failure data, is the running time of the target fault data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is the scale parameter of the current parameter of the target.

[0025] In an optional implementation manner, the step of determining a plurality of target parameters from the second preset number of current parameters according to the total error of all the current parameters includes:

[0026] All total errors are sorted in ascending order, and the multiple total errors before the preset position in the sequence are taken as the target total errors;

[0027] The current parameter corresponding to each total error in the target total error is taken as a target parameter to obtain multiple target parameters.

[0028] In an optional implementation, the step of determining the model parameters according to the multiple target parameters obtained last time includes:

[0029] The target parameter with the smallest total error among the multiple target parameters obtained for the last time is determined as the model parameter.

[0030] In an optional implementation manner, the preset condition is that the number of iterations reaches a preset number or the current mean and the current standard deviation both reach respective error convergence thresholds.

[0031] In a second aspect, the present invention provides a device for predicting equipment reliability based on q-weibull distribution, the device comprising:

[0032] An acquisition module is used to obtain the estimated running time of the target device;

[0033] A prediction module is used to input the estimated operating time into a reliability prediction model pre-constructed based on q-weibull distribution to obtain the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined after iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation.

[0034] In a third aspect, the present invention provides an electronic device comprising a processor and a memory, wherein the memory is used to store a program, and the processor is used to implement a device reliability prediction method based on q-weibull distribution as described in any one of the aforementioned embodiments when executing the program.

[0035] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a device reliability prediction method based on q-weibull distribution as described in any one of the aforementioned embodiments.

[0036] Compared with the prior art, the present invention has the following beneficial effects: the present invention determines the model parameters of the model after iteratively optimizing the parameter distribution of the model parameters of the reliability prediction model constructed based on the q-weibull distribution based on the preset mean and the preset standard deviation, thereby quickly determining the model parameters with high precision, thereby improving the accuracy of the prediction results of the reliability prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0038] Figure 1 An example graph of the failure rate function of the bathtub feature provided for this embodiment.

[0039] Figure 2 This is an example flowchart of the equipment reliability prediction method based on q-weibull distribution provided in this embodiment.

[0040] Figure 3 This is an example diagram for comparing the fitting effects provided in this embodiment.

[0041] Figure 4 This is a block diagram of an example of a device reliability prediction apparatus based on q-weibull distribution provided in this embodiment.

[0042] Figure 5 An example block diagram of an electronic device provided in this embodiment.

[0043] Icons: 10 - electronic device; 11 - processor; 12 - memory; 13 - bus; 100 - device reliability prediction device based on Q-Weibull distribution; 110 - acquisition module; 120 - prediction module; 130 - determination module. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0045] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0047] In the description of the present invention, it should be noted that if the terms "upper", "lower", "inside", "outside", etc. appear to indicate an orientation or position relationship, they are based on the orientation or position relationship shown in the accompanying drawings, or are the orientation or position relationship in which the product of the invention is usually placed when used. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0048] In addition, the terms “first”, “second”, etc., if used, are merely used to distinguish between the descriptions and should not be understood as indicating or implying relative importance.

[0049] It should be noted that, in the absence of conflict, the features in the embodiments of the present invention may be combined with each other.

[0050] In the field of equipment reliability technology, reliability engineers usually use the bathtub characteristic failure rate function to describe the reliability degradation process of equipment. Please refer to Figure 1 , Figure 1An example diagram of the failure rate function of the bathtub feature provided in this embodiment, Figure 1 In the , the failure rate function of the bathtub characteristic includes three stages: early failure stage, random failure stage, and loss failure stage.

[0051] In the early failure stage, the failure rate of the equipment is high, but it will decrease rapidly over time. This is usually caused by defects in design, raw materials and manufacturing process.

[0052] In the random failure stage, the failure rate is low and relatively stable, and can usually be approximated as a constant. The equipment in this period operates stably, and it is a good use stage for the product. Occasional failures are mainly caused by factors such as quality defects, material weaknesses, environment and improper use.

[0053] During the wear and tear failure stage, the failure rate increases rapidly with time, mainly caused by wear, fatigue, aging and consumption.

[0054] In order to simultaneously describe all stages of the failure rate function with bathtub characteristics so as to accurately predict the reliability of the equipment, a reliability prediction model is usually constructed based on the q-weibull distribution. Since the q-weibull distribution can not only describe the monotonic situation that the weibull distribution can describe, but also describe the classic bathtub-shaped and unimodal failure rate curves, it can more accurately simulate the failure change trend of the equipment to a certain extent and can more accurately predict the failure rate of the equipment. However, the existing point estimation of the q-weibull distribution parameters calculated based on the contour error function may have a slow convergence speed in the solution process, and may even fall into a local optimal solution, and cannot find the global optimal solution, resulting in low parameter estimation efficiency and the fitting effect cannot meet expectations, affecting the accuracy of the prediction results.

[0055] In view of this, an embodiment of the present invention provides a device reliability prediction method based on q-weibull distribution, which determines the model parameters of the model after iteratively optimizing the parameter distribution of the model parameters of the reliability prediction model constructed based on q-weibull distribution based on a preset mean and a preset standard deviation, thereby quickly determining the model parameters with high precision, thereby improving the accuracy of the prediction results of the reliability prediction model, which will be described in detail below.

[0056] Please refer to Figure 2 , Figure 2 The flowchart of the device reliability prediction method based on q-weibull distribution provided in this embodiment includes the following steps:

[0057] Step S101, obtaining the estimated operating time of the target device.

[0058] In this embodiment, the target device is a device for which a failure rate prediction is required, and the target device includes, but is not limited to, medical devices in the medical field, such as ventilators, monitors, blood gas analyzers, etc. The estimated operating time is the operating time of the target device to be estimated, and the estimated operating time can be the current operating time of the target device. At this time, the estimated probability of failure is the probability of the target device currently failing. For example, if the current device has been running for 50 hours, and 50 hours is taken as the estimated operating time, then the estimated probability of failure is the probability of failure corresponding to the target device running for 50 hours. The estimated operating time can also be the expected operating time of the target device, which is longer than the operating time. For example, the current target device has been running for 50 hours, and the estimated operating time is 100 hours, that is, it is necessary to estimate the probability of failure when its operating time reaches 100 hours, so as to determine a suitable strategy based on the probability to deal with the countermeasures when the target device fails, so that the impact of the target device when the failure occurs is within a controllable range.

[0059] Step S102, input the estimated operating time into the reliability prediction model pre-constructed based on the q-weibull distribution, and obtain the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined after iteratively optimizing the parameter distribution of the model parameters based on the preset mean and preset standard deviation.

[0060] In this embodiment, the probability density function of the q-weibull distribution can be expressed as:

[0061]

[0062] in, is the probability density function of the q-weibull distribution, It is the time from the preset starting point of the equipment (for example, the time when the equipment starts to be used or the system starts to run) to the time when the equipment fails. , , are the first shape parameter, the second shape parameter and the scale parameter respectively. 2 o'clock, The value range is ;when When represents the failure rate of the equipment at the end of its life is monotonically increasing, so There is a theoretical upper limit to the value: When q 1, the distribution simplifies to a Weibull distribution: ;when When , it can be transformed into q-exponential distribution.

[0063] According to the probability density function of q-weibull distribution, the distribution function of q-weibull distribution can be obtained:

[0064]

[0065] in, is the distribution function of the q-weibull distribution, , It is the time from the preset starting point of the equipment (for example, the time when the equipment starts to be used or the system starts to run) to the time when the equipment fails. , , are the first shape parameter, the second shape parameter and the scale parameter respectively.

[0066] By determining the value of the distribution parameter in the distribution function of the q-weibull distribution, we can obtain the reliability prediction model based on the q-weibull distribution.

[0067] This embodiment iteratively optimizes the parameter distribution of the model parameters (ie, the first shape parameter, the second shape parameter, and the scale parameter in the distribution function) based on a preset mean and a preset standard deviation, and finally determines the values ​​of the model parameters.

[0068] It should also be noted that after obtaining the reliability prediction model with determined model parameters, in addition to estimating the corresponding probability of failure based on the estimated operating time, a failure probability threshold can also be set in advance, and the failure probability threshold is substituted into the reliability prediction model to obtain the operating time of the target device when the failure probability of the target device reaches the failure probability threshold. For example, the failure probability threshold is set to 0.9, substituted into the reliability prediction model, and the operating time of the target device when the probability of failure of the target device is 0.9 is obtained, so as to evaluate and analyze the reliability of the target device. The failure probability threshold can also be set as needed, so that the target device can be inspected and maintained before the probability of failure occurs, and protective measures can be taken before the target device fails to prevent problems before they occur.

[0069] The above method provided in this embodiment determines the model parameters of the model after iteratively optimizing the parameter distribution of the model parameters of the reliability prediction model constructed based on the q-weibull distribution based on the preset mean and the preset standard deviation, thereby quickly determining the model parameters with high precision, thereby improving the accuracy of the prediction results of the reliability prediction model.

[0070] In order to specifically illustrate the process of iteratively optimizing parameter distribution, this embodiment provides a method for determining model parameters:

[0071] Step 1: Obtaining failure data of each sample product of a first preset number of sample products when the failure occurs for the first time, where all sample products are of the same model as the target device;

[0072] In this embodiment, in order to enable the determined model parameters to more accurately simulate the failure distribution of the target device, thereby more accurately predicting the failure rate of the target device, a device of the same model as the target device can be selected as a sample product, or a device of the same model and the same batch as the target device can be selected as a sample product.

[0073] In this embodiment, the first preset number can be set according to actual needs. Generally speaking, the more the first preset number is, the more accurately the model parameters can simulate the fault distribution of the target device.

[0074] In this embodiment, the fault data includes but is not limited to the operation time and the corresponding fault probability. For example, if the operation time is 10 hours, the corresponding fault probability is the probability of the sample product failing for the first time when the operation time reaches 10 hours. According to the needs of the actual scenario, the fault data may also include information such as the cause of the fault and the faulty components in the equipment, so as to predict the fault more specifically.

[0075] In this embodiment, as an implementation method, the failure probability of each sample product can be calculated using the following formula:

[0076] , i= 1,2,…, n

[0077] in, is the failure probability of the i-th sample product, n The first preset number.

[0078] Step 2: Use the preset mean and the preset standard deviation as the current mean and the current standard deviation respectively;

[0079] In this embodiment, when there are multiple model parameters, there are also multiple corresponding preset means and preset standard deviations. For example, the model parameters include , , , , , are the first shape parameter, the second shape parameter and the scale parameter respectively, then the preset mean also includes , , The mean and , , As an implementation method, multiple model parameters can be expressed in the form of parameter vectors: ,in, is the parameter vector.

[0080] Step 3: Determine the parameter distribution of the model parameters according to the current mean and the current standard deviation and generate a second preset number of current parameters that conform to the parameter distribution;

[0081] In this embodiment, as an implementation method, a normal distribution determined according to the current mean and the current standard deviation can be used as a parameter distribution, and then a second preset number of current parameters are generated according to the determined parameter distribution, which is expressed as: , j= 1,2,…, N ,in, is the current mean, µ , The parameters are q , β , λ The mean of is the current standard deviation, σ , The parameters are q , β , λ The standard deviation of For the j Current parameters, represents normal distribution, and N is the second preset number. The second preset number can also be set as needed.

[0082] Step 4: Calculate the total error of the predicted value obtained based on each current parameter prediction relative to all fault data;

[0083] In this embodiment, for each current parameter, the method of calculating the total error is the same. The total error of each current parameter is calculated one by one to obtain the total error of each current parameter. The specific implementation method is:

[0084] (1) For any target current parameter, calculate the intermediate error between each fault data and the target current parameter;

[0085] In this embodiment, since each sample product has its own fault data, for the target current parameter, it is necessary to calculate the intermediate error between it and the fault data of each sample product, and then sum up all the intermediate errors to obtain the total error of the target current parameter.

[0086] In this embodiment, for the target current parameter, the method of calculating the intermediate error between it and the fault data of each sample product is the same. This embodiment takes the target current parameter and any target fault data among all the fault data as an example to illustrate the method of calculating the intermediate error:

[0087] Using the formula Calculate the intermediate error between the target fault data and the target current parameters, where is the intermediate error between the target fault data and the target current parameters, is the failure probability of the target failure data, is the running time of the target fault data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is a scale parameter of the target current parameter, and the first shape parameter and the second shape parameter are used to determine the shape of the failure rate function.

[0088] (2) Sum all the intermediate errors to obtain the total error of the predicted value based on the target current parameter prediction relative to all fault data;

[0089] In this embodiment, the total error of the predicted value obtained by the target current parameter prediction relative to all fault data can be expressed as:

[0090]

[0091] in, is the total error of the predicted value obtained based on the target current parameter prediction relative to all fault data, is the intermediate error between the i-th fault data and the target current parameter, n is the first preset number, is the failure probability of the i-th failure data, is the running time of the i-th fault data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is a scale parameter of the target current parameter, and j indicates that the target current parameter is the jth current parameter among the second preset number of current parameters.

[0092] (3) Taking each current parameter as the target current parameter in turn, the total error of the predicted value obtained based on each current parameter relative to all fault data is obtained.

[0093] Step 5: determining a plurality of target parameters from a second preset number of current parameters according to the total errors of all current parameters;

[0094] In this embodiment, in order to continuously iterate so that the model parameters determined in each iteration can be closer to the actual fault distribution, each iteration selects the current parameters with smaller errors from the local iteration as the target parameters, and then determines the current mean and current standard deviation required for the next iteration based on the target parameters. This embodiment provides a method for determining the target parameters:

[0095] (1) Sort all total errors in ascending order, and take the total errors before the preset position in the sequence as the target total errors;

[0096] In this embodiment, the preset position can be determined according to the preset ratio, and the determination formula is: ,in, is the preset position, is the preset ratio, N is the second preset number, that is, The current parameter corresponding to the total error is taken as the target parameter.

[0097] (2) The current parameter corresponding to each total error in the target total error is taken as a target parameter to obtain multiple target parameters.

[0098] Step 6: Calculate the parameter mean and parameter standard deviation of all target parameters;

[0099] In this embodiment, the calculation formulas for the parameter mean and parameter standard deviation of the target parameter are respectively:

[0100]

[0101]

[0102] in, is the parameter mean of the target parameter, is the parameter standard deviation of the target parameter, is the jth target parameter, Represents a collection of target parameters.

[0103] Step 7: Reset the parameter mean as the current mean and the parameter standard deviation as the current standard deviation, and return to step 3 until the preset conditions are met;

[0104] In this embodiment, the preset condition is that the number of iterations reaches a preset number or the current mean and the current standard deviation both reach their respective error convergence thresholds.

[0105] In this embodiment, the preset number of times can be set according to actual needs, for example, the number of iterations is set to 100.

[0106] In this embodiment, the error convergence threshold may include a first error convergence threshold corresponding to the current mean value and a second error convergence threshold corresponding to the current standard deviation. The current mean value reaching the first error convergence threshold can be expressed as: < , k is the number of iterations that have been completed. is the mean value of the parameters obtained in this iteration, is the mean value of the parameters obtained in the previous iteration, is the first error convergence threshold. The current standard deviation reaches the second error convergence threshold and can be expressed as < , k is the number of iterations that have been completed. is the standard deviation of the parameters obtained in this iteration, is the standard deviation of the parameters obtained in the previous iteration, is the second error convergence threshold.

[0107] The first error convergence threshold and the second error convergence threshold can be set to be the same or different. For example, both can be set to .

[0108] Finally, the model parameters are determined according to the multiple target parameters obtained for the last time.

[0109] In this embodiment, the model parameters may be determined by taking the target parameter with the smallest total error among the multiple target parameters obtained last time as the model parameter. The model parameters may also be determined by taking the average value among the multiple target parameters obtained last time as the model parameter.

[0110] In order to more objectively compare the method provided in this embodiment with other methods, this embodiment uses the following indicators for evaluation: Akaike information criterion, modified Akaike information criterion, Bayesian information criterion, maximum determination coefficient, mean trouble-free working time, mean square error and P value. By comparing these indicators, the performance and applicability of different methods can be systematically evaluated. The meaning of each indicator is detailed in Table 1.

[0111] Table 1

[0112]

[0113] This example collects the running time data of 36 devices:

[0114] 0.058, 0.070, 0.090, 0.105, 0.113, 0.121, 0.153, 0.159, 0.224, 0.421, 0.570, 0.596, 0.618, 0.834, 1.019, 1.104, 1.497, 2.027, 2.234, 2.372, 2.433, 2.505, 2.690, 2.877, 2.879, 3.166, 3.455, 3.551, 4.378, 4.872, 5.085, 5.272, 5.341, 8.952, 9.188, 11.399.

[0115] Arrange the above 36 running times in ascending order according to the formula ,i=1,2,…,n calculate the corresponding failure probability, n=36, set the number of iterations to 100, generate 2000 candidate solutions in each iteration, and preset the ratio Set to 0.2, the error convergence threshold is , the point estimate of the q-weibull distribution is approximately .

[0116] Table 2 is a comparison between the method provided in this embodiment and the maximum likelihood method based on the above indicators.

[0117] Table 2

[0118]

[0119] In order to more intuitively demonstrate the effect of the method for predicting equipment reliability based on q-weibull distribution provided in this embodiment, the method provided in this embodiment is compared with the empirical distribution function (calculated based on Blom's method). Figure 3 , Figure 3 This is a comparison diagram of the fitting effect provided in this embodiment. Figure 3 It can be seen that the q-weibull distribution cumulative distribution curve fitted by the method provided in this embodiment is closer to the empirical distribution function, which means that the model fits the data better and the difference between the model prediction and the actual data is smaller. Since the empirical distribution function is directly calculated based on the actual data, it can reflect the actual distribution of the data, so being closer to the empirical distribution function indicates that the model has a higher fitting accuracy.

[0120] In order to execute the corresponding steps in the above embodiment and each possible implementation method, the following provides a method for implementing the device reliability prediction device 100 based on q-weibull distribution. Figure 4 , Figure 4The block diagram of the equipment reliability prediction device based on q-weibull distribution provided in this embodiment. It should be noted that the basic principle and technical effect of the equipment reliability prediction device 100 based on q-weibull distribution provided by the present invention are the same as those of the corresponding above-mentioned embodiments, and are not mentioned in this embodiment for the sake of brief description.

[0121] The device reliability prediction apparatus 100 based on q-weibull distribution includes an acquisition module 110 and a prediction module 120 .

[0122] An acquisition module 110 is used to acquire an estimated operating time of a target device;

[0123] The prediction module 120 is used to input the estimated operating time into a reliability prediction model pre-constructed based on the q-weibull distribution to obtain the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined by iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation.

[0124] In an optional implementation, the device 100 for predicting equipment reliability based on q-weibull distribution may further include a determination module 130, and the determination module 130 is used to determine the model parameters in the following manner:

[0125] Obtaining failure data of each sample product of a first preset number of sample products when the failure occurs for the first time, where all the sample products are of the same model as the target device;

[0126] The preset mean and the preset standard deviation are used as the current mean and the current standard deviation respectively;

[0127] Determine a parameter distribution of the model parameters according to the current mean and the current standard deviation and generate a second preset number of current parameters that conform to the parameter distribution, the second preset number being greater than the first preset number;

[0128] Calculate the total error of the predicted value obtained based on each current parameter prediction relative to all fault data;

[0129] Determining a plurality of target parameters from a second preset number of current parameters according to a total error of all current parameters;

[0130] Calculate the parameter mean and parameter standard deviation of all target parameters;

[0131] The parameter mean is reset as the current mean, and the parameter standard deviation is reset as the current standard deviation, and the step of returning to determining the parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution is returned until a preset condition is met;

[0132] The model parameters are determined according to the multiple target parameters obtained last time.

[0133] In an optional implementation manner, when the determination module 130 is used to calculate the total error of the predicted value obtained based on each current parameter prediction relative to all fault data, it is specifically used to:

[0134] For any target current parameter, calculate the intermediate error between each fault data and the target current parameter; sum all the intermediate errors to obtain the total error of the predicted value based on the target current parameter relative to all fault data; take each current parameter as the target current parameter in turn, and obtain the total error of the predicted value based on each current parameter relative to all fault data.

[0135] In an optional embodiment, the fault data includes the running time and the fault probability, the current parameter includes the scale parameter and the first shape parameter and the second shape parameter for determining the shape of the fault rate function, and for any target fault data among all the fault data, the determination module 130 calculates the intermediate error between the target fault data and the target current parameter in the following manner:

[0136] Using the formula Calculate the intermediate error between the target fault data and the target current parameters, where is the intermediate error between the target fault data and the target current parameters, is the failure probability of the target failure data, is the running time of the target fault data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is the scale parameter of the target's current parameter.

[0137] In an optional implementation manner, when the determination module 130 is used to determine multiple target parameters from a second preset number of current parameters according to the total errors of all current parameters, it is specifically used to:

[0138] All total errors are sorted in ascending order, and multiple total errors before a preset position in the sequence are taken as target total errors; the current parameter corresponding to each total error in the target total error is taken as a target parameter to obtain multiple target parameters.

[0139] In an optional implementation, when the determination module 130 is used to determine the model parameter according to the multiple target parameters obtained last time, it is specifically used to: determine the target parameter with the smallest total error among the multiple target parameters obtained last time as the model parameter.

[0140] In an optional implementation, the preset condition in the determination module 130 is that the number of iterations reaches a preset number or the current mean and the current standard deviation both reach their respective error convergence thresholds.

[0141] The embodiment of the present invention also provides a block diagram of an electronic device 10, and the electronic device 10 implements the device reliability prediction method based on q-weibull distribution of the above embodiment, please refer to Figure 5 , Figure 5 This is a block diagram of an electronic device 10 provided in this embodiment. The electronic device 10 includes a processor 11, a memory 12 and a bus 13. The processor 11 and the memory 12 are connected via the bus 13.

[0142] The processor 11 may be an integrated circuit chip with signal processing capability. In the implementation process, each step of the device reliability prediction method based on q-weibull distribution in the above embodiment may be completed by hardware integrated logic circuits or software instructions in the processor 11. The above processor 11 may be a general-purpose processor, including a CPU (Central Processing Unit), NP (Network Processor), etc.; it may also be a DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Logic Gate Array) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0143] The memory 12 is used to store a program for implementing a device reliability prediction method based on q-weibull distribution. The program may be a software function module stored in the memory 12 in the form of software or firmware or solidified in the OS (Operating System) of the electronic device 10.

[0144] After receiving the execution instruction, the processor 11 executes the program to implement the device reliability prediction method based on q-weibull distribution of the above embodiment.

[0145] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the device reliability prediction method based on q-weibull distribution as described in any one of the aforementioned embodiments is implemented.

[0146] In summary, the embodiments of the present invention provide a method, apparatus, electronic device and storage medium for predicting equipment reliability based on q-weibull distribution, the method comprising: obtaining an estimated operating time of a target device; inputting the estimated operating time into a reliability prediction model pre-constructed based on q-weibull distribution, and obtaining the probability of failure when the operating time of the target device reaches the estimated operating time; the model parameters of the reliability prediction model are determined after iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation. Compared with the prior art, the present embodiment determines the model parameters of the model after iteratively optimizing the parameter distribution of the model parameters of the reliability prediction model constructed based on q-weibull distribution based on a preset mean and a preset standard deviation, thereby quickly determining high-precision model parameters, thereby improving the accuracy of the prediction results of the reliability prediction model.

[0147] The above are only various embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A device reliability prediction method based on q-weibull distribution, characterized in that: The method comprises: Get the estimated runtime of the target device; The estimated operating time is input into a reliability prediction model pre-constructed based on q-weibull distribution to obtain the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined by iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation; The model parameters are determined as follows: Obtaining failure data of each of the sample products of a first preset number when a failure occurs for the first time, wherein all the sample products are of the same model as the target device, and the failure data includes operation time and failure probability; Using the preset mean and the preset standard deviation as the current mean and the current standard deviation respectively; Determining a parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution, wherein the second preset number is greater than the first preset number, and the current parameters include a scale parameter and a first shape parameter and a second shape parameter that determine a shape of a failure rate function; Calculate the total error of the predicted value obtained based on each of the current parameters relative to all fault data; Determining a plurality of target parameters from the second preset number of current parameters according to the total errors of all the current parameters; Calculate the parameter mean and parameter standard deviation of all the target parameters; Resetting the parameter mean as the current mean and resetting the parameter standard deviation as the current standard deviation, returning to the step of determining the parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution, until a preset condition is met; Determine the model parameters according to the multiple target parameters obtained last time; The step of calculating the total error of the predicted value obtained based on each of the current parameters relative to all fault data comprises: For any target current parameter, calculating an intermediate error between each of the fault data and the target current parameter; Sum all the intermediate errors to obtain a total error of the predicted value obtained based on the target current parameter prediction relative to all the fault data; Taking each of the current parameters as the target current parameter in turn, obtaining a total error of a predicted value obtained based on each of the current parameters relative to all fault data; For any target fault data among all the fault data, the method of calculating the intermediate error between the target fault data and the target current parameter is: Using the formula Calculate the intermediate error between the target fault data and the target current parameter, where: is the intermediate error between the target fault data and the target current parameter, is the failure probability of the target failure data, t is the running time of the target failure data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is the scale parameter of the current parameter of the target.

2. The equipment reliability prediction method based on q-weibull distribution according to claim 1 is characterized in that: The step of determining a plurality of target parameters from the second preset number of current parameters according to the total error of all the current parameters comprises: All total errors are sorted in ascending order, and the total errors before the preset position in the sequence are taken as the target total errors; The current parameter corresponding to each total error in the target total error is taken as a target parameter to obtain multiple target parameters.

3. The equipment reliability prediction method based on q-weibull distribution according to claim 1 is characterized in that: The step of determining the model parameters according to the multiple target parameters obtained last time comprises: The target parameter with the smallest total error among the multiple target parameters obtained for the last time is determined as the model parameter.

4. The equipment reliability prediction method based on q-weibull distribution according to claim 1, characterized in that: The preset condition is that the number of iterations reaches a preset number or the current mean and the current standard deviation both reach their respective error convergence thresholds.

5. A device for predicting equipment reliability based on q-weibull distribution, characterized in that: The device comprises: An acquisition module is used to obtain the estimated running time of the target device; A prediction module, used for inputting the estimated operating time into a reliability prediction model pre-constructed based on q-weibull distribution, and obtaining the probability of failure of the target device when the operating time reaches the estimated operating time; the model parameters of the reliability prediction model are determined by iteratively optimizing the parameter distribution of the model parameters based on a preset mean and a preset standard deviation; The model parameters are determined as follows: Obtaining failure data of each of the sample products of a first preset number when a failure occurs for the first time, wherein all the sample products are of the same model as the target device, and the failure data includes operation time and failure probability; Using the preset mean and the preset standard deviation as the current mean and the current standard deviation respectively; Determining a parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution, wherein the second preset number is greater than the first preset number, and the current parameters include a scale parameter and a first shape parameter and a second shape parameter that determine a shape of a failure rate function; Calculate the total error of the predicted value obtained based on each of the current parameters relative to all fault data; Determining a plurality of target parameters from the second preset number of current parameters according to the total errors of all the current parameters; Calculate the parameter mean and parameter standard deviation of all the target parameters; Resetting the parameter mean as the current mean and resetting the parameter standard deviation as the current standard deviation, returning to the step of determining the parameter distribution of the model parameter according to the current mean and the current standard deviation and generating a second preset number of current parameters that conform to the parameter distribution, until a preset condition is met; Determine the model parameters according to the multiple target parameters obtained last time; The method of calculating the total error of the predicted value obtained based on each current parameter prediction relative to all fault data is as follows: For any target current parameter, calculating an intermediate error between each of the fault data and the target current parameter; Sum all the intermediate errors to obtain a total error of the predicted value obtained based on the target current parameter prediction relative to all the fault data; Taking each of the current parameters as the target current parameter in turn, obtaining a total error of a predicted value obtained based on each of the current parameters relative to all fault data; For any target fault data among all the fault data, the method of calculating the intermediate error between the target fault data and the target current parameter is: Using the formula Calculate the intermediate error between the target fault data and the target current parameter, where: is the intermediate error between the target fault data and the target current parameter, is the failure probability of the target failure data, t is the running time of the target failure data, and are the first shape parameter and the second shape parameter of the current parameter of the target, is the scale parameter of the current parameter of the target.

6. An electronic device, characterized in that: The invention comprises a processor and a memory, wherein the memory is used to store a program, and the processor is used to implement the device reliability prediction method based on q-weibull distribution as described in any one of claims 1 to 4 when executing the program.

7. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the device reliability prediction method based on q-weibull distribution as described in any one of claims 1 to 4 is implemented.

Images