Equipment residual life estimation method fusing grey prediction modeling and discrete random impact damage

Abstract

The invention discloses an equipment residual life estimation method fusing grey prediction modeling and discrete random impact damage, which comprises the following steps: collecting degradation data of equipment performance indexes, and analyzing the data to extract degradation data features; according to the degradation data characteristics, searching equipment performance degradation characteristics, constructing a grey prediction model fused with random impact, and estimating the residual life of the equipment; solving gray prediction model parameters, solving impact characteristics including the number of first-arrival impact times and an average degradation track, and then calculating the equipment failure probability; and according to the impact characteristics, solving the probability distribution of the residual life of the equipment, calculating the reliability of the equipment, and estimating the average residual life of the equipment. According to the method, the problems of residual life estimation and health management of the equipment in a small sample discrete impact scene are effectively solved, the performance degradation rule of the equipment under discrete random impact can be accurately described, the residual life estimation precision is remarkably improved, and a reliable decision basis is provided for scientific management of the equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of equipment health management, and particularly relates to a method for estimating the remaining life of equipment by integrating grey prediction modeling and discrete random shock damage. Background Art

[0002] During the service process, large and complex equipment is often subjected to the combined action of multi-source random shocks such as vibration, stress and temperature. Such external excitations will continuously cause the performance degradation of the equipment. In the time dimension, the occurrence time and action intensity of the random shocks suffered by the equipment show significant uncertainty characteristics. However, restricted by factors such as the high integration of the equipment system and the precise and complex internal structure, it is difficult to obtain the performance degradation data of the core components through real-time monitoring, resulting in the inapplicability of existing remaining life prediction methods. Building a scientific remaining life prediction model and accurately analyzing the life distribution and reliability of the equipment is not only the key basis for realizing the condition-based maintenance and dynamic management of the equipment, but also the core link for improving the reliability and effectiveness evaluation of the entire life cycle of the equipment, which is of great significance for ensuring the safe operation of the equipment and completing the production tasks.

[0003] The random shocks suffered by the equipment during operation are the result of the interweaving of multi-source uncertainty factors. This uncertainty includes both the randomness of the shock occurrence time and the volatility of the performance degradation caused by the shock. Each shock may cause the performance degradation of the equipment. As the number of shocks increases, it will lead to the failure of the equipment. It should be noted that the uncertainty of the shock arrival process has typical random characteristics, and more importantly, the number of shocks has the characteristics of "small sample and poor information" (some equipment is subjected to fewer shocks, such as ships being subjected to extreme wave shocks). Traditional probability statistical methods often face the dilemma of insufficient data when dealing with such problems. The grey prediction model, with its powerful processing ability for limited data, can effectively extract the degradation trend hidden in the discrete shock sequence and provide a scientific description tool for the quantification of equipment performance degradation. Compared with traditional probability statistical methods, the grey prediction model shows significant advantages under the conditions of "small sample and poor information". Traditional methods rely on a large amount of historical data to establish a probability distribution model. In actual engineering, many equipment are difficult to accumulate enough data samples due to short service time, low frequency of extreme working conditions, etc., resulting in large deviations in the prediction results. The grey prediction model transforms the discrete shock data into a regular sequence, and then realizes the dynamic tracking and prediction of the equipment degradation state. Combining deterministic trend prediction with random fluctuation analysis can comprehensively depict the degradation law of the equipment under random shocks.

[0004] To accurately characterize this complex uncertainty, the present invention combines a stochastic process with a grey prediction model to form a new hybrid modeling framework. First, the original discrete device performance degradation data is subjected to cumulative generation to weaken the data randomness and explore the potential trend of device performance degradation. At the same time, based on the theory of stochastic processes, a Poisson process or a renewal process is used to probabilistically model the shock arrival intervals to quantify the randomness of the shock occurrence times. After organically integrating the two, the model can not only capture the deterministic trend of device performance degradation but also describe the random characteristics of the shock process, realizing dynamic and accurate prediction of the remaining useful life of the device.

[0005] Based on the above modeling idea, by combining the Poisson process and the grey prediction model, a quantitative relationship is established between the device performance degradation index and the remaining useful life to predict the device performance degradation trend, and the failure probability of the device at different times is calculated using the theory, thereby realizing the dynamic prediction of the remaining useful life of the device and providing a reliable decision-making basis for the scientific management of the device. [[ID=##]] Summary of the Invention

[0006] Based on the problems existing in the prior art, the present invention proposes a method for estimating the remaining useful life of a device by integrating grey prediction modeling and discrete random shock damage. This method aims at the problem of estimating the remaining useful life of a device in the scenario of a small sample of discrete random shock times, constructs a discrete grey prediction model integrating the Poisson process; solves the expression of the first arrival shock times of device performance degradation, the average degradation trajectory, and calculates the device failure probability to obtain the probability distribution of the remaining useful life of the device, forming a method for estimating the remaining useful life of the device in the case of discrete random shock damage. Compared with traditional methods, the grey prediction model breaks through the shackles of "data dependence" and can still effectively model in extreme cases with a small number of samples, which is a new prediction idea and method for complex devices with scarce life data and high monitoring costs. Traditional statistical modeling methods (such as Weibull distribution) lead to parameter estimation deviations due to data sparsity, while life prediction based on physical models requires accurate characterization of processes such as material aging and thermal stress fatigue of satellite components, which is difficult to implement in practice. Compared with the prior art, the present invention aims to solve the problems of estimating the remaining useful life of a device and health management in the scenario of small sample discrete shocks, accurately characterize the performance degradation law of the device under discrete random shocks, significantly improve the accuracy of remaining useful life estimation, and provide a reliable decision-making basis for the scientific management of the device.

[0007] To achieve the above technical objectives, the present invention provides the following technical solutions: A method for estimating the remaining useful life of a device by integrating grey prediction modeling and discrete random shock damage, which specifically includes the following steps: S1. Collect the degradation data of the device performance indicators and analyze the degradation data to extract the degradation data characteristics; S2. According to the characteristics of degradation data, find the degradation characteristics of the equipment performance, and construct a grey prediction model integrating random shocks to estimate the remaining life of the equipment; S3. Solve the parameters of the grey prediction model constructed in step S2, and solve the shock characteristics based on the parameters of the grey prediction model, including: the number of first arrival shocks, the average degradation trajectory; then calculate the equipment failure probability; S4. According to the shock characteristics, solve the probability distribution of the remaining life of the equipment, calculate the equipment reliability, and estimate the average remaining life of the equipment.

[0008] Further, step S1 is specifically as follows: Collect and organize the degradation data of the equipment performance indicators. Use discrete time series to represent the performance degradation amount of the equipment after being subjected to discrete random shocks. Each discrete random shock is recorded as a moment; from the 1st moment to the n The sequence of the performance degradation amount of the equipment performance indicator at the moment is denoted as: ; At the same time, calculate the cumulative degradation amount of the equipment performance indicator after each shock, and the formula is expressed as: ; where represents the degradation amount of the equipment performance indicator at the i-th moment.

[0009] Further, step S2 is specifically as follows: According to the cumulative degradation amount of the equipment performance indicator obtained in step S1, analyze the growth trend of the performance degradation amount after discrete random shocks, and construct a discrete random grey prediction model, and the formula expression is: ; where is the number of shocks received by the equipment in the time period , follows a homogeneous Poisson process with a shock intensity parameter of ; , are two parameters to be solved.

[0010] Further, step S3 specifically includes: S31. According to the cumulative degradation amount of the equipment performance indicator obtained in step S1, solve the parameters of the random discrete grey prediction model, and its calculation formula is specifically: ; where ; ; S32. Solve the first hitting shock number \(T\) according to the discrete random grey prediction model and the failure threshold of the device performance index. The specific calculation formula is as follows: ; where is the critical value of the device performance index failure, is the initial value of the device performance index degradation, is the value-taking function; S33. Solve the average degradation trajectory of the device performance degradation index according to the first hitting shock number, the discrete random grey prediction model and the properties of the Poisson process, which is expressed as the mathematical expectation of the degradation amount. The formula is expressed as: ; S34. According to the first hitting shock number and the properties of the Poisson process, obtain the probability of being subjected to random shocks in the time period as: ; Then the probability of device failure in the time period is: ; where is the shock intensity parameter.

[0011] Furthermore, step S4 specifically includes: S41. Calculate the distribution of the remaining life of the device according to the random discrete grey prediction model followed by the device performance degradation law which follows a gamma distribution , and its probability density function is: ; Its reliability function is: ; S42. According to the distribution of the remaining life of the device, calculate the mathematical expectation of the distribution: ; S43. Estimate the shock intensity parameter according to the shock arrival interval time : ; is the time interval between the th shock and the th shock received by the device, is the total number of shocks received by the device; S44. Substitute the estimated value of the impact strength parameter calculated in step S43 and the number of first - reaching impacts calculated in step S32 into the expectation of the remaining life distribution of the device to obtain the estimated value of the average remaining life of the device.

[0012] The present invention also provides a device, including a memory and a processor, where: The memory is used to store a computer program that can run on the processor; The processor is used to execute a method for estimating the remaining life of a device that integrates grey prediction modeling and discrete random shock damage as described above when running the computer program.

[0013] In addition, according to the above - mentioned device, the present invention also provides a computer - readable storage medium, which stores computer instructions, and the computer instructions are used to enable the processor to implement a method for estimating the remaining life of a device that integrates grey prediction modeling and discrete random shock damage as described above when executed.

[0014] Based on the above - mentioned technical solutions, the present invention has at least the following beneficial effects: Aiming at the randomness and non - repeatability of impact events and the non - linear characteristics and multi - factor coupling effects of the device performance degradation process when the device is under discrete random shocks, the present invention proposes a method for estimating the remaining life of a device that integrates grey prediction modeling and discrete random shock damage. By quantifying the occurrence law of random shocks and the trend of device performance degradation, a prediction model for device performance degradation and remaining life suitable for discrete - time shock scenarios is constructed. It can derive the number of first - reaching impacts of device performance degradation and its failure probability, give the probability distribution of the remaining life of the device, and realize the prediction of the remaining life of the device, that is, effectively solve the problem of estimating the remaining life of the device and health management in the small - sample discrete shock scenario, and can accurately describe the performance degradation law of the device under discrete random shocks; thus significantly improving the estimation accuracy of the remaining life of the device, helping device managers better grasp the operation of the device, and providing a reliable decision - making basis for scientific device management. Description of the Drawings

[0015] Figure 1 It is a flow chart of the method proposed by the present invention; Figure 2 It is for calculating the device performance degradation prediction curve and the number of first - reaching impacts; Figure 3 It is a graph of the device failure probability under different impact strength levels; Figure 4 It is a graph of the probability density function of the remaining life distribution of the device under different impact strength levels. Detailed Embodiments

[0016] To make the objectives, technical solutions and advantages of the present invention more clearly understood, the following further describes the present invention in detail with reference to the accompanying Figures 1-4 drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not used to limit the present invention.

[0017] Although the steps in the present invention are arranged with reference numerals, they are not used to limit the order of the steps. Unless the order of the steps is clearly stated or the execution of a certain step requires other steps as a basis, the relative order of the steps can be adjusted. It can be understood that the term "and / or" used herein relates to and encompasses any and all possible combinations of one or more of the associated listed items.

[0018] As Figure 1 shown, the present invention proposes a method for estimating the remaining life of a device by integrating grey prediction modeling and discrete random shock damage. This method includes two parts, namely, drawing the device performance degradation prediction curve and calculating the remaining life distribution; specifically including the following steps: S1. Collect the degradation data of the device performance indicators, and analyze the degradation data to extract the characteristics of the degradation data; As a preferred implementation manner, step S1 is specifically: Collect and sort out the degradation data of the device performance indicators, and use discrete time series to represent the performance degradation amount of the device after being subjected to discrete random shocks. Each discrete random shock is recorded as a moment; from the 1st moment to the n th moment, the sequence of the device performance indicator degradation amounts is recorded as: ; At the same time, calculate the cumulative degradation amount of the device performance indicators after each shock, and the formula is expressed as: ; where represents the degradation amount of the device performance indicator at the i-th moment.

[0019] S2. According to the characteristics of the degradation data, find the device performance degradation characteristics, and construct a grey prediction model integrating random shocks to estimate the remaining life of the device; As a preferred implementation manner, step S2 is specifically: According to the cumulative degradation amount of the device performance indicators obtained in step S1, analyze the trend of the growth of the performance indicator degradation amount after discrete random shocks, and construct a discrete random grey prediction model, and the formula expression is: ; where is the number of shocks received by the device within the time period , obeys the shock intensity parameter of Homogeneous Poisson process; , are two parameters to be solved.

[0020] In the present invention, a new hybrid modeling idea is formed by combining a stochastic process with a grey prediction model. First, the original discrete device performance degradation data is subjected to cumulative generation to weaken the randomness of the data and mine the potential trend of the device performance degradation. At the same time, based on the stochastic process theory, a Poisson process or a renewal process is used to probabilistically model the shock arrival intervals to quantify the randomness of the shock occurrence times. After the two are organically integrated, the model can not only capture the deterministic trend of the device performance degradation but also describe the random characteristics of the shock process, realizing dynamic and accurate prediction of the remaining life of the device. Through the above steps S1 - S2, a device performance degradation prediction curve as shown in Figure 2 can be obtained. The device performance degradation prediction curve passes through discrete time nodes and can systematically depict the evolution law of the performance degradation amount under different shock numbers.

[0021] S3. Solve the parameters of the grey prediction model constructed in step S2, and solve the shock characteristics based on the parameters of the grey prediction model, including: the first - arrival shock number, the average degradation trajectory; and then calculate the device failure probability; As a preferred embodiment, step S3 specifically includes: S31. According to the cumulative degradation amount of the device performance index obtained in step S1 , solve the parameters of the stochastic discrete grey prediction model, and its specific calculation formula is: ; where, ; ; S32. According to the discrete stochastic grey prediction model and the device performance index failure threshold, solve the first - arrival shock number T, and its specific calculation formula is: ; where, is the device performance index failure critical value, is the initial value of the device performance index degradation, is the value - taking function; in this application, the first - arrival shock number T is defined as the minimum shock number corresponding to the device failure, and the remaining life distribution parameters of the device are calculated through the first - arrival shock number.

[0022] S33. According to the first - arrival shock number, the discrete stochastic grey prediction model, and the properties of the Poisson process, solve the average degradation trajectory of the device performance degradation index, which is expressed as the mathematical expectation of the degradation amount, and the formula is expressed as: ; In this embodiment, by constructing the average degradation trajectory, the degradation degree of the performance index at any time can be accurately simulated; S34. According to the number of first arrival shocks and the properties of the Poisson process, obtain the probability of being subjected to random shocks in the time period as: Then the probability of equipment failure in the time period is: ; wherein, is the shock intensity parameter; meanwhile, a time-varying graph of the failure probability under different shock intensities as shown in Figure 3 can be plotted.

[0023] S4. According to the shock characteristics, solve the probability distribution of the remaining life of the equipment, calculate the reliability of the equipment, and estimate the average remaining life of the equipment; As a preferred implementation manner, step S4 specifically includes: S41. According to the random discrete grey prediction model followed by the equipment performance degradation law, calculate the distribution of the remaining life of the equipment obeying the gamma distribution , and its probability density function is: ; Similarly, a time-varying graph of the probability density function under different shock intensities as shown in Figure 4 can be plotted; Its reliability function is: ; S42. According to the distribution of the remaining life of the equipment , calculate the mathematical expectation of the distribution: ; S43. According to the shock arrival interval time , estimate the shock intensity parameter: ; is the time interval between the th shock and the th shock received by the equipment, is the total number of shocks received by the equipment; S44. Substitute the estimated value of the shock intensity parameter calculated in step S43 and the number of first arrival shocks calculated in step S32 into the expectation of the remaining life distribution of the equipment to obtain the estimated value of the average remaining life of the equipment.

[0024] So far, the method proposed by the present invention addresses the problem of estimating the remaining life of equipment in the small-sample scenario of discrete random shock times. A discrete grey prediction model integrating the Poisson process is constructed. By solving the expression for the first hitting shock times of equipment performance degradation, the average degradation trajectory, and calculating the equipment failure probability, the probability distribution of the remaining life of the equipment can be obtained, and then the remaining life of the equipment can be estimated.

[0025] This embodiment also provides a device, including a memory and a processor, where: The memory is used to store a computer program that can run on the processor; The processor is used to execute a method for estimating the remaining life of equipment integrating grey prediction modeling and discrete random shock damage as described above when running the computer program.

[0026] In addition, according to the above device, the present invention also provides a computer-readable storage medium storing computer instructions for causing the processor to execute a method for estimating the remaining life of equipment integrating grey prediction modeling and discrete random shock damage as described above when executed.

[0027] So far, the overall process of the method proposed by the present invention has been introduced. Next, the feasibility and effectiveness of the method proposed by the present invention are verified through example simulations. In this embodiment, combined with the ultimate strength simulation data of the hull stiffened panel under extreme cyclic loads, the model constructed by the present invention is applied to predict the ultimate strength and remaining life of the stiffened panel with a single stiffener (hereinafter referred to as the stiffened panel). The data in the example is from a published literature. In this literature, the Abaqus software is used to simulate the cyclic load of the stiffened panel made of S355 steel. The pressure amplitude in the cyclic load is 5100 kN, and the tensile amplitude is 4950 kN. The prediction curve result of the ultimate strength decrease of the stiffened panel is as [[ID=I4]] Figure 2 shown. The numerical experimental results show that the prediction curve based on the ultimate strength decrease of the stiffened panel is very close to the true value, which reflects the effectiveness and rationality of the method proposed by the present invention.

[0028] In summary, the method proposed by the present invention addresses the randomness and non-repeatability of shock events and the non-linear characteristics and multi-factor coupling effects of the equipment performance degradation process when the equipment is subjected to discrete random shocks. A method for estimating the remaining life of equipment integrating grey prediction modeling and discrete random shock damage is proposed. By quantifying the occurrence law of random shocks and the trend of equipment performance degradation, a prediction model for equipment performance degradation and remaining life applicable to discrete-time shock scenarios is constructed. The first hitting shock times of equipment performance degradation and its failure probability are derived, and the probability distribution of the remaining life of the equipment is given to achieve the prediction of the remaining life of the equipment.

[0029] Based on the remaining life distribution of the equipment under discrete random shocks, it can help equipment managers better understand the operation of the equipment and provide a reliable decision-making basis for the scientific management of the equipment.

[0030] In this specification, descriptions such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" refer to at least one embodiment or example described in combination with specific features, structures, materials, or characteristics. These specific features, structures, materials, or characteristics can be combined in one or more embodiments or examples in an appropriate manner. In addition, without contradiction, those skilled in the art can combine and combine different embodiments or examples and their features described in this specification.

[0031] The logic and / or steps shown in the flowchart or otherwise described can be regarded as a sequence of executable instructions for implementing logical functions. These instructions can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device. These systems, apparatuses, or devices include a processor system or other systems capable of receiving and executing instructions.

[0032] The above embodiments have introduced the principles and embodiments of the present invention in detail and illustrated its working principles with specific examples. These examples are only used to help understand the method and its core idea of the present invention. At the same time, according to the idea of the present invention, the actual embodiments and application scopes may vary. Therefore, the content of this specification should not be construed as a limitation on the present invention.

Claims

1. A method for estimating the remaining life of equipment by integrating grey prediction modeling and discrete random shock damage, characterized in that Specifically, it includes the following steps: S1. Collect the degradation data of the device performance indicators, and analyze the degradation data to extract the characteristics of the degradation data; S2. According to the characteristics of the degradation data, find the degradation characteristics of the device performance, and construct a grey prediction model integrating random shocks to estimate the remaining life of the device; S3. Solve the parameters of the grey prediction model constructed in step S2, and solve the shock characteristics based on the parameters of the grey prediction model, including: the number of first arrival shocks, the average degradation trajectory; then calculate the device failure probability; S4. According to the shock characteristics, solve the probability distribution of the remaining life of the device, calculate the reliability of the device, and estimate the average remaining life of the device.

2. The method for estimating the remaining life of a device by integrating grey prediction modeling and discrete random shock damage according to claim 1, characterized in that, Step S1 is specifically as follows: Collect and organize the degradation data of the device performance indicators. Use discrete time series to represent the performance degradation amount of the device after being subjected to discrete random shocks. Each discrete random shock is recorded as a moment; from the 1st moment to the n moment, the sequence of the device performance indicator degradation amounts is denoted as: ; At the same time, calculate the cumulative degradation amount of the device performance indicators after each shock, and the formula is expressed as: ； Among them, represents the degradation amount of the device performance index at time i.

3. The method for estimating the remaining life of a device by integrating grey prediction modeling and discrete random shock damage according to claim 2, wherein Step S2 is specifically as follows: The cumulative degradation amount of the device performance index obtained according to step S1 , analyze the growth trend of the degradation amount of the performance index after discrete random shocks, and construct a discrete random grey prediction model, which is expressed by the formula: ； wherein, is the number of shocks received by the device during the time period , obeys a homogeneous Poisson process with a shock intensity parameter of ; , are two parameters to be solved.

4. A method for estimating the remaining life of a device that combines grey prediction modeling and discrete random shock damage, as claimed in claim 3, wherein Step S3 specifically includes: S31. Cumulative degradation amount of device performance indicators obtained according to step S1 , solve the parameters of the stochastic discrete grey prediction model, and its specific calculation formula is as follows: ； Among them, ; ; S32. According to the discrete random grey prediction model and the device performance indicator failure threshold, solve the number of first arrival shocks T, and its calculation formula is specifically: ； Among them, is the critical failure value of the device performance index, is the initial value of the degradation of the device performance index, is the value-taking function; S33. According to the number of first arrival shocks, the discrete random grey prediction model and the properties of the Poisson process, solve the average degradation trajectory of the device performance degradation index, which is expressed as the mathematical expectation of the degradation amount, and the formula is expressed as: ； S34. According to the number of first - arrival shocks and the properties of the Poisson process, the probability of receiving random shocks times within the time period is as follows: ； Then in the time period The probability of equipment failure is: ； Among them, is the impact strength parameter.

5. A method for estimating the remaining life of a device by integrating grey prediction modeling and discrete random shock damage, as claimed in claim 4, wherein Step S4 specifically includes: S41. Calculate the distribution of the remaining life of the equipment according to the random discrete grey prediction model that the equipment performance degradation law follows. Follow a gamma distribution , and its probability density function is: ； Its reliability function is: ； S42. Calculate the mathematical expectation of the distribution according to the remaining life distribution of the device , and calculate the mathematical expectation of the distribution: ； S43. Estimate the impact strength parameter based on the impact arrival interval time , and ； is the time interval between the th impact and the th impact on the device; is the total number of impacts on the device; S44. Substitute the estimated value of the shock intensity parameter calculated in step S43 and the number of first arrival shocks calculated in step S32 into the expectation of the remaining life distribution of the device to obtain the estimated value of the average remaining life of the device.

6. A device, characterized in that, The device includes a memory and a processor, where: The memory is used to store a computer program that can run on the processor; The processor is used to execute a method for estimating the remaining life of a device integrating grey prediction modeling and discrete random shock damage as described in any one of claims 1-5 when running the computer program.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and the computer instructions are used to cause the processor to execute a method for estimating the remaining life of a device integrating grey prediction modeling and discrete random shock damage as described in any one of claims 1-5 when executed.

Images