Method and apparatus for preventive maintenance with preventive strategy

Abstract

The application relates to a failure prediction maintenance method and device with a preventable strategy, and relates to the field of equipment prediction and maintenance technology. In view of the problems in the prior art that a traditional two-state reliability model is difficult to accurately describe the degradation degree, failure condition and maintenance state of a system in a service life, the calculation process is relatively single, and the result error is relatively large, the technical scheme is as follows: the failure prediction maintenance method with the preventable strategy comprises the following steps: collecting a probability distribution obeyed by the residual life of a system; collecting maintenance time and maintenance cost; establishing an optimization model; iteratively processing the model to obtain an optimal failure prediction interval and a residual life threshold value; when the optimal failure prediction interval is reached, the residual life of the system is predicted, the residual life is compared with the threshold value, and if the residual life is lower than the threshold value, the system is subjected to corrective maintenance, and if the residual life is higher than the threshold value, the system is subjected to preventive maintenance. The application is suitable for application in preventive maintenance work, and establishes a new research basis for the research of preventive maintenance work.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of predictive and maintenance technology, and in particular to a maintenance strategy with preventive strategy. BACKGROUND

[0002] With the rapid development of science and technology, complex, large, intelligent, and fine are the four characteristics of modern industrial production equipment. Modern enterprises are increasingly dependent on large-scale intelligent and precise industrial equipment, and even the equipment has become a decisive factor for the survival and development of enterprises. Since industrial equipment is directly related to product production, it will directly affect product quality, yield, and cost. For some complex structure and large-scale engineering systems, if the equipment suddenly fails during operation, it will seriously affect the production of the enterprise and cause great economic losses to the enterprise. In order to prevent such incidents from occurring, it is of great significance to develop a scientific equipment maintenance strategy to effectively avoid failures and ensure the continuous normal operation of the equipment for the survival and development of enterprises.

[0003] Currently, there are three main maintenance methods for maintainable systems: after-failure maintenance, periodic maintenance, and reliability-centered maintenance. After-failure maintenance refers to maintenance work after the system is shut down due to equipment failure, and the system is restored to normal working condition by repairing or replacing the faulty equipment. This maintenance method will have a serious impact on the service life of the system and will also cause economic losses. Generally, this method is used for maintenance of small consequences and short maintenance time. Periodic maintenance refers to maintenance of equipment within a specified period, which is the most commonly used preventive maintenance strategy in traditional industries before the information age. The characteristics are simple and convenient, and usually rely on the working experience of equipment managers to have a vague prediction of the failure law of the equipment and then perform periodic maintenance. However, this method is prone to "over-repair" and "under-repair" phenomena. After entering the information age, with the more complex and large-scale structure of modern industrial equipment, enterprises have expanded their maintenance work centered on reliability. Based on the principles of maintenance economy, reliability, and risk, the system working and maintenance time, reliability, and maintenance interval are determined to ensure that the system can achieve the lowest reliability condition, and to reduce the minimum maintenance risk and consume the least maintenance resources for preventive maintenance. Therefore, it can effectively reduce the number of failures and downtime.

[0004] Simple reliability maintenance analysis method can not meet the growing demand of people in terms of calculation accuracy or efficiency. In recent years, a reliability analysis method with prediction gradually rises and is widely concerned by domestic and foreign scholars. Failure prediction is to evaluate the future situation according to the current operation state, bearing capacity and the like of a system, including possible failure mode and residual life. Through failure prediction, accidents can be prevented, and the system maintenance personnel can take corresponding preparations in advance, which is an important measure to realize preventive maintenance. Meanwhile, according to the predicted situation, it can be judged whether the future result can reach the predetermined reliability, and if not, maintenance activities or scientific and effective preventive maintenance strategies need to be developed. Preventive maintenance optimization strategy includes reducing system risk, improving system reliability and reducing maintenance cost and the like.

[0005] In the traditional analysis method, the working state of the system is generally divided into two cases of 'normal operation' and 'failure'. However, in actual engineering, there are a series of other intermediate states between the two states of 'normal operation' and 'failure', such as repair state, test state and installation state. The model established by the above two system working states is called two-state model and multi-state model. Since the traditional two-state reliability model is difficult to accurately describe the degradation degree, failure condition and maintenance state of the system in the life cycle, it only describes the two states of system working and failure, without considering other intermediate factors, so the calculation process is relatively simple, and the result error is large. SUMMARY

[0006] In view of the problem in the prior art that the traditional two-state reliability model is difficult to accurately describe the degradation degree, failure condition and maintenance state of the system in the life cycle, it only describes the two states of system working and failure, without considering other intermediate factors, so the calculation process is relatively simple, and the result error is large, the technical scheme provided by the present application is as follows:

[0007] The failure prediction maintenance method with preventable strategy comprises the following steps:

[0008] Step 1: collecting the probability distribution of the residual life of the system to be maintained;

[0009] Step 2: collecting the maintenance time and maintenance cost of the system to be maintained;

[0010] Step 3: establishing a maintenance strategy optimization model with failure prevention according to the maintenance time and maintenance cost;

[0011] Step 4: collecting a preset optimization algorithm and initializing the optimization algorithm;

[0012] Step 5: according to the optimization algorithm, the model is iterated to obtain the optimal fault prediction interval and the residual life threshold;

[0013] Step 6: when reaching each of the optimal fault prediction intervals, the residual life of the system to be maintained is predicted, and the residual life of the system to be maintained is compared with the residual life threshold; if it is lower than the preset residual life threshold, preventive maintenance or repair maintenance is performed; if it is higher than the preset residual life threshold, preventive maintenance is performed.

[0014] Further, a preferred embodiment is provided, wherein the step 1 further comprises the step of collecting the residual life distribution and the reliability function of the system to be maintained.

[0015] Further, a preferred embodiment is provided, wherein the step 2, the maintenance cost comprises: preventive maintenance cost, preventive repair cost and repair maintenance cost; the maintenance time comprises: preventive maintenance time, preventive repair time and repair maintenance time.

[0016] Further, a preferred embodiment is provided, wherein the step 2 further comprises the step of collecting the maintenance risk value of the system to be maintained.

[0017] Further, a preferred embodiment is provided, wherein the maintenance strategy optimization model with fault prevention is specifically: taking the reliability index as the input of the model, and taking the optimization variable as the output of the model.

[0018] The reliability index is: the reliability of the system, the risk of the system and the average maintenance cost.

[0019] The optimization variable is: the optimal fault prediction interval and the residual life threshold.

[0020] Further, a preferred embodiment is provided, wherein the step 4, initializing the optimization algorithm specifically comprises: setting the population size, the space dimension, the maximum iteration number and the search upper and lower bound of the optimization algorithm.

[0021] Further, a preferred embodiment is provided, wherein the step 5, the number of iterations of the model is the preset maximum iteration number.

[0022] Based on the same inventive concept, the present application also provides a fault prediction maintenance device with preventable strategy, which comprises:

[0023] Module 1: used for collecting the probability distribution to which the residual life of the system to be maintained is subjected;

[0024] Module 2: used for collecting the maintenance time and maintenance cost of the system to be maintained;

[0025] Module 3: for establishing a maintenance strategy optimization model with fault prevention according to the maintenance time and maintenance cost;

[0026] Module 4: for collecting a preset optimization algorithm and initializing the optimization algorithm;

[0027] Module 5: for iterating the model according to the optimization algorithm;

[0028] Module 6: for predicting the remaining life of the system to be maintained when each optimal fault prediction interval is reached, comparing the remaining life of the system to be maintained with the remaining life threshold, and if the remaining life of the system to be maintained is lower than the preset remaining life threshold, performing preventive maintenance or repair maintenance, and if the remaining life of the system to be maintained is higher than the preset remaining life threshold, performing preventive maintenance.

[0029] Based on the same inventive concept, the application also provides a computer storage medium for storing a computer program, and when a processor of a computer processes the computer program stored in the computer storage medium, the computer executes the fault prediction maintenance method with a predictable strategy.

[0030] Based on the same inventive concept, the application also provides a computer including a processor and a storage medium, and the storage medium is used to store a computer program, and when the processor of the computer processes the computer program stored in the computer storage medium, the computer executes the fault prediction maintenance method with a predictable strategy.

[0031] The fault prediction maintenance method with a preventable strategy provided by the application establishes a maintenance strategy optimization model with fault prevention, which considers maintenance factors, so that the reliability of the system is higher.

[0032] Compared with the calculation result of a single-objective optimization model, the maintenance strategy optimization model with fault prevention established by the fault prediction maintenance method with a preventable strategy provided by the application fully considers various influencing factors, so that there is a mutual restriction relationship among the three objective functions, the best combination mode of the system is comprehensively considered, and the result is more accurate.

[0033] The fault prediction maintenance method with a preventable strategy provided by the application obtains the best maintenance scheme through the prediction of preventive maintenance behavior, that is, the time interval of each fault prediction is obtained, and guidance for preventive maintenance work is provided.

[0034] The fault prediction maintenance method with a preventable strategy provided by the application is suitable for application in preventive maintenance work, and establishes a new research basis for the research of preventive maintenance work. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1The flowchart of the failure prediction maintenance method with preventable strategy provided for the embodiment one;

[0036] Figure 2 The schematic diagram of the relationship between the preventive maintenance frequency mentioned for the embodiment five and the system maintenance cost change;

[0037] Figure 3 The schematic diagram of the relationship between the preventive maintenance frequency mentioned for the embodiment five and the system reliability change;

[0038] Figure 4 The schematic diagram of the relationship between the preventive maintenance frequency mentioned for the embodiment five and the system risk change. DETAILED DESCRIPTION

[0039] In order to make the advantages and benefits of the technical solutions provided by the present application more specific, the technical solutions provided by the present application will be further described in detail in combination with the drawings, and the specific embodiments are as follows:

[0040] Embodiment one, in combination Figure 1 It is illustrated that the embodiment provides a failure prediction maintenance method with preventable strategy, which comprises:

[0041] Step 1: Collecting the probability distribution obeyed by the residual life of the system to be maintained;

[0042] Step 2: Collecting the maintenance time and maintenance cost of the system to be maintained;

[0043] Step 3: Establishing a maintenance strategy optimization model with failure prevention according to the maintenance time and maintenance cost;

[0044] Step 4: Collecting a preset optimization algorithm and initializing the optimization algorithm;

[0045] Step 5: Iterating the model according to the optimization algorithm to obtain optimal failure prediction intervals and residual life thresholds;

[0046] Step 6: When each optimal failure prediction interval is reached, the residual life of the system to be maintained is predicted, and the residual life of the system to be maintained is compared with the residual life threshold value. If it is lower than the preset residual life threshold value, preventive maintenance or repair maintenance is performed, and if it is higher than the preset residual life threshold value, preventive maintenance is performed.

[0047] The failure prediction maintenance method with preventable strategy provided by the embodiment is based on a multi-state reliability model, which has repair, installation and test states, contains many factors and considers more fully. The reliability method with prediction has better applicability in analyzing multi-state systems at present.

[0048] The failure prediction maintenance method with preventive strategy provided by the embodiment applies the failure prediction method to the maintainable engineering structure, establishes a preventive maintenance strategy model based on the failure prediction, and solves the time interval and the minimum residual life of the failure prediction through an intelligent optimization algorithm. The system maintenance cost, reliability and risk reach an optimal combination state. By changing the preventive maintenance frequency and other variables in the objective function, the change law of the system average maintenance cost, reliability and risk value is summarized, and the sensitivity of each variable can be analyzed. The optimization model can improve the accuracy and efficiency of the calculation result, evaluate and predict the system risk, and strengthen the safety of the engineering structure.

[0049] Embodiment two, the embodiment is a further limitation of the failure prediction maintenance method with preventive strategy provided by embodiment one, and the step 1 further includes the steps of collecting the residual life distribution and the reliability function of the system to be maintained.

[0050] Step 1 is specifically: collect the probability distribution to which the system residual life distribution is subjected, and respectively represent the residual life distribution F m (t) and the reliability function R m (t).

[0051] Embodiment three, the embodiment is a further limitation of the failure prediction maintenance method with preventive strategy provided by embodiment one, and in the step 2, the maintenance cost includes: preventive maintenance cost, preventive maintenance cost and repair maintenance cost; and the maintenance time includes: preventive maintenance time, preventive maintenance time and repair maintenance time.

[0052] Step 2 is specifically: according to the maintainability characteristics of the system, determine the system maintenance time and cost parameters. Among them, the cost of preventive maintenance and maintenance of the equipment is C k , C p , the maintenance cost of the equipment after failure is C f ; the time required for preventive maintenance and maintenance is T k , T p , and the repair maintenance time is T f ; the risk values of preventive maintenance, maintenance and repair maintenance are R k , R p , and R f .

[0053] Embodiment four, the embodiment is a further limitation of the failure prediction maintenance method with preventive strategy provided by embodiment one, and the step 2 further includes the step of collecting the maintenance risk value of the system to be maintained.

[0054] Embodiment five, in combinationFigures 2-4 The embodiment is a further limitation of the failure prediction maintenance method with preventive strategy provided in Embodiment One, and the maintenance strategy optimization model with failure prevention is specifically: taking reliability indicators as inputs of the model, and taking optimization variables as outputs of the model.

[0055] The reliability indicators are: reliability of the system, risk of the system, and average maintenance cost.

[0056] The optimization variables are: optimal failure prediction interval and residual life threshold.

[0057] The step 3 is specifically:

[0058] The maintenance strategy optimization mathematical model of the system with failure prediction is established, that is, the relationship between the reliability indicators and the optimization variables is established. In the life time of the system, the average residual life at time kh is predicted as u(kh), when u(kh)>u0, the system needs to be maintained; when u(kh)≤u0, the system needs to be repaired; if the system suddenly fails in [(k-1)h, kh] time, immediate repair maintenance is needed. Wherein h is the prediction interval, u0 is the threshold of preventive maintenance, and k is the number of times of preventive maintenance of the equipment in the range of (0, t).

[0059] When the average residual life u(kh) of the system at time kh is predicted to be ≤u0, the probability P of preventive maintenance is pk which can be expressed as:

[0060]

[0061] Wherein t represents time, P represents probability, and R represents reliability.

[0062] The probability P of immediate repair maintenance of the system suddenly failing in [(k-1)h, kh] time is fk which can be expressed as:

[0063]

[0064] Wherein T represents the service life of the system.

[0065] Since the maintenance of the system is a cumulative function, the probability of the Nth preventive maintenance of the system is The probability of the Nth repair maintenance of the system is

[0066] The cost generated by the maintenance of the system mainly includes preventive maintenance cost C k , preventive repair cost C pand the repair cost C f The average cost EC of the system in the life cycle can be expressed as:

[0067]

[0068] The average life cycle E(T f ) of the system in [(k-1)h, kh] at time t when the system suddenly fails can be expressed as:

[0069]

[0070] Where F Tf (t) is the distribution function of the system failure at time t in [(k-1)h, kh]:

[0071]

[0072] Assuming the model, the preventive maintenance at time kh or the repair maintenance at time t in [(k-1)h, kh] when the system fails are considered as the end of a life cycle, so the average life cycle ET of the system can be calculated as:

[0073]

[0074] The average cost per unit time C of the system in the life cycle is equal to the ratio of the average cost EC of the system in the life cycle and the average life cycle ET of the system, that is, the maintenance cost model of the system is:

[0075]

[0076] The system will be shut down when the equipment is in three maintenance or repair states, and the downtime of the system in preventive maintenance, preventive repair and repair repair is T k , T p and T f , respectively. The average downtime ED of the system in the life cycle can be expressed as:

[0077]

[0078] The model of the reliability A of the system can be expressed as:

[0079]

[0080] The risk of the system can be understood as the safety problem of equipment loss risk and personal environmental risk. The system will have risks when the equipment is in three maintenance or repair states. The risk values of the system in preventive maintenance, preventive repair and repair repair are Rk , R p and R f , the average risk value ER in the system life time can be expressed as:

[0081]

[0082] The model of the risk B of the system can be expressed as:

[0083]

[0084] Under the premise that the system can be safely operated, a preventive maintenance strategy method model with fault prediction is established to meet the requirements of minimum average maintenance cost, maximum reliability and minimum risk degree of the system. The reliability of the system meets the minimum requirement, the preventive maintenance threshold is not lower than the failure threshold, and the comprehensive optimization model of the system is:

[0085]

[0086] Wherein, A 0 represents the minimum required reliability of the system, u s represents the failure threshold.

[0087] From the above formula, the two important parameters affecting the average cost, reliability and risk of the system maintenance are the preventive maintenance threshold u0 and the preventive maintenance cycle h, therefore, according to the objective function in the model, the two parameters are optimized and analyzed.

[0088] The fault prediction maintenance method with preventive strategy provided by the application, in the process of establishing the model, if a single objective optimization model of the system is established, the constraint relationship between multiple objectives is not considered, the set parameters are respectively brought into the maintenance cost model, the reliability model and the risk three optimization models, and the results of each objective function can be calculated. If a multi-objective optimization model of the system is established, since there is a mutual constraint between the three objectives in the comprehensive optimization model of the system, the average maintenance cost and risk degree of the system are the lowest, and the reliability is the highest, at this time, it should be the best combination of the system, and the practicability is the strongest.

[0089] The main purpose of the fault prediction maintenance method with preventive strategy provided by the application is to obtain the time interval h of each fault prediction and judge the current state of the system. If the remaining life of the system is greater than u0, the system is in a relatively safe state, and appropriate preventive maintenance can be performed, if the remaining life of the system is less than u0 at this time, the system is dangerous, and preventive maintenance or shutdown for repair maintenance is required.

[0090] In actual work, the reliability of the system can be improved by increasing the number of preventive maintenance k. In the initial value case, as the number of preventive maintenance increases, the changes of the system maintenance cost, reliability and risk of the three objectives can be obtained, as shown in Figures 2-4 The results show that increasing the number of preventive maintenance can reduce the prediction interval and timely discover potential failures, and take appropriate measures to avoid further deterioration of the failure, thereby leading to an increase in the reliability value of the system and a decrease in the risk value. However, the proportion of preventive maintenance cost and repair cost is relatively large, and excessive number of preventive maintenance makes the maintenance cost also increase significantly, while the reliability and risk value increase or decrease is not obvious, thereby causing adverse effects on the economic benefits of the system. Therefore, it is necessary to select a reasonable number of preventive maintenance to ensure the safety of the system, so that the reliability, maintenance cost and risk value of the system can reach the optimal state;

[0091] Embodiment six, this embodiment is a further limitation of the failure prediction maintenance method with preventive strategy provided by embodiment one, in step 4, initializing the optimization algorithm specifically includes: setting the population size, space dimension, maximum iteration number and search upper and lower bound of the optimization algorithm.

[0092] The step 4 specifically includes:

[0093] Selecting appropriate intelligent optimization algorithm and setting part of the parameters in the algorithm, according to the multi-objective optimization model, the relationship between the maintenance average cost, system reliability and risk and optimization variable can be established. Since it is a multi-objective multi-variable nonlinear continuous optimization problem, the parameters set in step two are brought into the optimization model formula, and appropriate intelligent optimization algorithm is selected, under the condition of meeting the minimum reliability and failure threshold of the system, the optimal preventive maintenance period and preventive maintenance threshold are obtained, so that the average maintenance cost and risk degree of the system are the lowest, and the reliability is the highest. Setting part of the parameters in the algorithm, setting the population size S, the space dimension D, the maximum iteration number T max , the search upper and lower bound U max , U min .

[0094] Among them, the optimization algorithm adopts multi-objective genetic algorithm;

[0095] Wherein, when the optimal fault prediction interval is reached, the remaining life of the system to be maintained is predicted by using the existing grey prediction or Markov prediction method. The grey prediction is mainly used to predict the system containing part of the known information and uncertain information, that is, to predict the process related to time and changing within a certain range. For this kind of system, it is usually called a grey system. The Markov prediction method is to analyze the state of the current event to predict the probability of various states occurring at the next time, so as to predict the state change of each stage. This feature is also called non-aftereffect. The neural network prediction method constructs a sample set through training to obtain a prediction model, which can be applied to the field of predicting the remaining life of the event, and has strong environmental adaptability and self-learning ability.

[0096] Embodiment seven, this embodiment is a further limitation of the fault prediction maintenance method with preventable strategy provided by embodiment one, wherein the number of iterations of the model in step 5 is the preset maximum iteration number.

[0097] Specifically, step five is specifically: using programming software to calculate the optimization algorithm, iteratively optimize, constantly update the position of the searcher, obtain the result of the objective function, determine the time interval of each fault prediction, compare the prediction result with the remaining life threshold, and judge whether the system is in a safe and reliable state. If the system is relatively safe, appropriate preventive maintenance can be performed, and if the system is in a dangerous state, preventive maintenance or shutdown for repair maintenance is required.

[0098] Embodiment eight, this embodiment provides a fault prediction maintenance device with a preventable strategy, the device comprises:

[0099] Module 1: for collecting the probability distribution of the remaining life of the system to be maintained;

[0100] Module 2: for collecting the maintenance time and maintenance cost of the system to be maintained;

[0101] Module 3: for establishing a maintenance strategy optimization model with fault prevention according to the maintenance time and maintenance cost;

[0102] Module 4: for collecting a preset optimization algorithm and initializing the optimization algorithm;

[0103] Module 5: for iterating the model according to the optimization algorithm;

[0104] Module 6: for predicting the remaining life of the system to be maintained when reaching each of the optimal failure prediction intervals, comparing the remaining life of the system to be maintained with the remaining life threshold, and if the remaining life is lower than the preset remaining life threshold, performing preventive maintenance or corrective maintenance, and if the remaining life is higher than the preset remaining life threshold, performing preventive maintenance.

[0105] Embodiment nine, the embodiment provides a computer storage medium for storing a computer program, when a processor of a computer processes the computer program stored in the computer storage medium, the computer executes the failure prediction maintenance method with predictable strategy provided by any one of embodiments one to seven.

[0106] Embodiment ten, the embodiment provides a computer including a processor and a storage medium, the storage medium is used to store a computer program, when the processor of the computer processes the computer program stored in the computer storage medium, the computer executes the failure prediction maintenance method with predictable strategy provided by any one of embodiments one to seven.

[0107] The above further describes the technical solutions provided by the present application in several specific embodiments, in order to highlight the advantages and benefits of the technical solutions provided by the present application. However, the above several specific embodiments are not used as a limitation of the present application, any modification and improvement of the present application, combination and equivalent replacement of the embodiments, etc. based on the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method of failure prognostic maintenance with preventive strategy, characterized in that, The method comprises: Step 1: collecting a probability distribution obeyed by the residual life of a system to be maintained; Step 2: collecting maintenance time and maintenance cost of the system to be maintained; Step 3: establishing a maintenance strategy optimization model with failure prevention according to the maintenance time and maintenance cost; Step 4: collecting a preset optimization algorithm and initializing the optimization algorithm; Step 5: iterating the model according to the optimization algorithm to obtain optimal failure prediction intervals and residual life thresholds; Step 6: predicting the residual life of the system to be maintained when each optimal failure prediction interval is reached, comparing the residual life of the system to be maintained with the residual life threshold, and if the residual life is lower than the preset residual life threshold, performing preventive maintenance or corrective maintenance, and if the residual life is higher than the preset residual life threshold, performing preventive maintenance. Specifically, step 4 is specifically: Selecting a suitable intelligent optimization algorithm and setting part of the parameters in the algorithm, according to the multi-objective optimization model, the relationship between the maintenance average cost, system reliability and risk and optimization variables is established; bringing the parameters set in step 2 into the optimization model formula, through the intelligent optimization algorithm, the optimal preventive maintenance period and preventive maintenance threshold are obtained under the condition of meeting the minimum system reliability and failure threshold, setting part of the parameters in the algorithm, setting the population size , the spatial dimension , the maximum iteration number is , the search upper and lower bounds are , ; The optimization algorithm uses a multi-objective genetic algorithm; In the process of predicting the residual life of the system to be maintained when the optimal failure prediction interval is reached, a grey prediction or Markov prediction method is used.

2. The method of claim 1, wherein, In step 1, the step of collecting the residual life distribution and reliability function of the system to be maintained is further included.

3. The method of claim 1, wherein, In step 2, the maintenance cost includes preventive maintenance cost, preventive maintenance cost and corrective maintenance cost; and the maintenance time includes preventive maintenance time, preventive maintenance time and corrective maintenance time.

4. The method of claim 1, wherein the method further comprises: In step 2, the step of collecting the maintenance risk value of the system to be maintained is further included.

5. The method of claim 1, wherein the method further comprises: The maintenance strategy optimization model with failure prevention is specifically: taking reliability indicators as inputs of the model and taking optimization variables as outputs of the model; The reliability indicators are: system reliability, system risk and average maintenance cost; The optimization variables are: optimal failure prediction intervals and residual life thresholds.

6. The method of claim 1, wherein the method further comprises: In step 4, initializing the optimization algorithm specifically includes: setting the population size, spatial dimension, maximum iteration number and search upper and lower bounds of the optimization algorithm.

7. The method of claim 1, wherein the method further comprises: In step 5, the number of iterations of the model is the preset maximum iteration number.

8. A failure prognostic maintenance device with preventive strategy, characterized by, The device comprises: Module 1: for collecting a probability distribution obeyed by the residual life of a system to be maintained; Module 2: for collecting maintenance time and maintenance cost of the system to be maintained; Module 3: for establishing a maintenance strategy optimization model with failure prevention according to the maintenance time and maintenance cost; Module 4: for collecting a preset optimization algorithm and initializing the optimization algorithm; Module 5: for iterating the model according to the optimization algorithm; Module 6: for predicting the residual life of the system to be maintained when each optimal failure prediction interval is reached, comparing the residual life of the system to be maintained with the residual life threshold, and if the residual life is lower than the preset residual life threshold, performing preventive maintenance or corrective maintenance, and if the residual life is higher than the preset residual life threshold, performing preventive maintenance. Specifically, step 4 is specifically: Selecting a suitable intelligent optimization algorithm and setting part of the parameters in the algorithm, according to the multi-objective optimization model, the relationship between the maintenance average cost, system reliability and risk and optimization variables is established; bringing the parameters set in step 2 into the optimization model formula, through the intelligent optimization algorithm, the optimal preventive maintenance period and preventive maintenance threshold are obtained under the condition of meeting the minimum system reliability and failure threshold, setting part of the parameters in the algorithm, setting the population size , the spatial dimension , the maximum iteration number is , the search upper and lower bounds are , ; The optimization algorithm uses a multi-objective genetic algorithm; In the process of predicting the residual life of the system to be maintained when the optimal failure prediction interval is reached, a grey prediction or Markov prediction method is used. When the optimal fault prediction interval is reached, the remaining life process of the system to be maintained is predicted by using grey prediction or Markov prediction method.

9. Computer storage medium for storing a computer program, characterized in that When the processor of the computer processes the computer program stored in the computer storage medium, the computer executes the fault prediction maintenance method with predictable strategy according to any one of claims 1-7.

10. Computer comprising a processor and a storage medium for storing a computer program, characterized in that When the processor of the computer processes the computer program stored in the computer storage medium, the computer executes the fault prediction maintenance method with predictable strategy according to any one of claims 1-7.

Images