Full-life-cycle refined maintenance method for automatic protection system of high-speed train

Through dynamic and variable maintenance strategies throughout the life cycle, the intermittent and economic problems of the maintenance of automatic protection systems of high-speed trains are solved, and the refined maintenance of equipment and cost reduction are achieved, ensuring the reliability and economic balance of the system.

CN118941270APending Publication Date: 2024-11-12CHINA STATE RAILWAY GRP CO LTD +4
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410964366.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2024-03-09
Filing Date
2024-07-18
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The maintenance of automatic protection systems of high-speed trains has problems such as intermittent work, strict maintenance timing and difficult to monitor equipment performance trends, resulting in the failure of existing maintenance methods to effectively cope with the characteristics of the ATP system and are of low economicality.

Method used

Using dynamic and variable maintenance strategies throughout the life cycle, through data acquisition, preprocessing, reliability test and maintenance model establishment, a two-dimensional fusion dynamic time-varying maintenance strategy based on application duration and operating mileage is generated to optimize the optimal life of the equipment. Distribution and maintenance cycles to achieve balance of equipment reliability and economics.

Benefits of technology

It realizes refined maintenance of different equipment, different stages and different time periods, reduces maintenance costs, and ensures the minimum reliability of the equipment, reducing the impact of equipment intermittent work on maintenance strategies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941270A_ABST
    Figure CN118941270A_ABST
Patent Text Reader

Abstract

The invention discloses a full-life-cycle refined maintenance method for an automatic protection system of a high-speed train, and relates to the technical field of rail transit. Comprising the following steps: S1, data acquisition; s2, a data preprocessing step; s3, a reliability test step; s4, establishing a maintenance model; s5, a maintenance strategy generation step; and S6, a maintenance implementation step. According to the invention, through a dynamically variable maintenance strategy in a full life cycle, different equipment and different maintenance methods are realized; the maintenance methods are different in different stages of the whole life cycle of the same equipment; the maintenance method is updated in time, and the maintenance cost is effectively reduced while the lowest reliability is ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of rail transit technology, and in particular to a refined maintenance method for a high-speed train automatic protection system throughout its life cycle. Background Art

[0002] The Automatic Train Protection (ATP) system is the nerve center that ensures the safe and efficient operation of high-speed trains. It directly controls the operation of trains by working closely with ground systems and vehicle systems. At present, there are five types of ATP systems used on a large scale, which are denoted as A, B, C, D, and E. Among them, three types are suitable for 350km / h high-speed railways, and two types are suitable for 250km / h high-speed railways. Regardless of the type, according to the functional division, it is composed of a security computer (Vital Computer, VC), a balise transmission module (BTM), a driver-machine interface (DMI), a train interface (Train Interface Unit, TIU), a track circuit reader (TCR), a speed and distance processing unit (Speed ​​& Distance Processing Unit, SDU) and other sub-equipment.

[0003] In order to ensure the safety and security capabilities of ATP, my country's railways implement strict maintenance strategies. Different from other systems, the daily maintenance of the ATP system has the following characteristics:

[0004] The system works intermittently. Since the ATP system is installed at both ends of the high-speed train, the system starts working as the train performs the transportation task and shuts down when the transportation task is completed. Therefore, the system does not work around the clock. Compared with the operation time, the correlation between operating mileage and reliability is closer.

[0005] Specific maintenance time and conditions. ATP's business responsibilities and installation location determine that after a piece of equipment fails, maintenance can only be carried out after the train has completed its transportation mission and stopped at a specific location. On the one hand, the maintenance time is limited and the conditions for timely maintenance are not met. On the other hand, there is no equipment downtime or railway transportation interruption due to maintenance.

[0006] Equipment performance trends are difficult to monitor. Most electronic equipment fails randomly, and it is impossible to characterize the degradation trends from normal to failure through monitoring methods. Therefore, condition-based maintenance based on equipment performance changes is difficult to achieve.

[0007] It is difficult to uniformly maintain multiple devices. Due to the particularity of transportation tasks, different ATP devices are scattered in different physical areas along with the trains. Only different sub-devices belonging to the same ATP or the same train can be uniformly maintained.

[0008] The existing ATP system mainly adopts the method of planned preventive maintenance, and performs fixed maintenance items at fixed periods. Although this maintenance method guarantees the quality of equipment to a certain extent, it is relatively extensive, has not been adaptively improved according to the characteristics of ATP maintenance, and has low economic efficiency, which is specifically reflected in the following aspects:

[0009] The equipment types are not differentiated, and all types of ATP equipment implement a unified maintenance strategy. However, different types of equipment have different production processes, source quality, etc., which means different reliability laws. The existing maintenance model fails to reflect the differences between different types.

[0010] The relationship between reliability and time is not distinguished, and a fixed maintenance strategy is implemented throughout the life cycle of the equipment from production to extinction. The differences in different time periods are not reflected.

[0011] The maintenance items, maintenance intervals, and maintenance methods of all equipment remain fixed throughout their life cycle, failing to reflect the differences in maintenance models.

[0012] Appropriate maintenance is an effective solution to reduce equipment failures and keep the system running normally. Preventive maintenance is the best choice to keep the system running stably. Usually, preventive maintenance is divided into regular preventive maintenance and condition-based maintenance. Regular maintenance has a fixed time period and wastes resources, which is not the best option. Therefore, most domestic and foreign research focuses on condition-based maintenance based on equipment performance changes. Its key is condition monitoring, and there are two ways to achieve it technically:

[0013] (1) At the electronic component level, establish the correspondence between high-reliability characteristic parameters and device aging status, so as to achieve safe and effective electronic device status monitoring. For a complex system such as ATP composed of thousands of electronic components, it is difficult to establish an accurate mathematical model.

[0014] (2) Install various sensors at key locations or conduct various experiments, using the data collected by sensors or experiments as a breakthrough point to indirectly understand the service rules of the system from the outside. For the ATP system with extremely complex operating environment and strict safety requirements, there is currently no external system to monitor its performance changes. It can be seen that it is not feasible to implement condition-based maintenance for ATP.

[0015] The ATP system has independent characteristics such as intermittent operation and strict maintenance timing, which requires personalized maintenance strategies.

[0016] Therefore, it is an urgent problem for technical personnel in this field to propose a refined maintenance method for the entire life cycle of the high-speed train automatic protection system to solve the difficulties existing in the prior art. Summary of the invention

[0017] In view of this, the present invention provides a refined maintenance method for the entire life cycle of a high-speed train automatic protection system. Through a dynamically variable maintenance strategy throughout the entire life cycle, the maintenance cost is effectively reduced while ensuring the minimum reliability.

[0018] In order to achieve the above object, the present invention adopts the following technical solution:

[0019] A refined maintenance method for a high-speed train automatic protection system throughout its life cycle, comprising the following steps:

[0020] S1, data acquisition step: obtaining fault data sets of various devices of the high-speed train automatic protection system;

[0021] S2, data preprocessing step: perform abnormal data inspection on the fault data set obtained in S1 to obtain a qualified fault data set;

[0022] S3, reliability test step: perform reliability test based on the qualified failure data set obtained in S2 to obtain the optimal life distribution of different equipment;

[0023] S4, maintenance model establishment step: based on the optimal life distribution obtained in S3, establish an optimal maintenance model that balances equipment mission reliability and economy;

[0024] S5, maintenance strategy generation step: Based on the optimal maintenance model of equipment task reliability and economic balance obtained in S4, the mapping relationship between equipment reliability and operation time and running mileage is analyzed through the Pearson correlation coefficient, and a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operation time and running mileage is output;

[0025] S6. Maintenance implementation steps: Based on the full life cycle maintenance strategy obtained in S5, the high-speed train automatic protection system is maintained, and the maintenance data is saved and imported as a new fault data set.

[0026] In the above method, optionally, the specific contents in S1 include:

[0027] Within the selected date range, for the failures of each device in the high-speed train automatic protection system, the name, type, start date of use and date of failure of the device are determined, and the life data samples X of different devices are obtained as the failure data set.

[0028] In the above method, the specific content of S2 optionally includes: judging whether there is abnormal data in the fault data set based on the optimized Laida criterion, and removing the abnormal data to obtain a qualified fault data set.

[0029] In the above method, optionally, the specific steps in S3 include:

[0030] S301 estimates the qualified fault data set obtained in S2 using the maximum likelihood estimation method to obtain parameter values ​​of possible life distribution types;

[0031] S302, based on the parameter values ​​of the possible life distribution types obtained in S301, a KS test is performed to determine whether the equipment life obeys a certain distribution, and test results of different life distribution types are obtained;

[0032] S303 analyzes and compares the test results obtained in S302 to obtain the life distribution type with the highest degree of consistency with the actual data as the optimal life distribution.

[0033] In the above method, optionally, the specific contents of S4 include:

[0034] Based on the optimal life distribution determined in S3, a dynamically changing failure rate function is introduced to characterize the impact of maintenance behavior on equipment performance, including analysis of equipment reliability and maintenance cost; an optimal maintenance model that balances equipment mission reliability and economy is established, and the maintenance model is solved through genetic algorithm or particle swarm algorithm to obtain the length of the maintenance cycle.

[0035] In the above method, optionally, the reliability analysis in S4 includes the following contents:

[0036] Reliability analysis mainly describes the changes in equipment performance before and after maintenance by introducing service age regression factors and failure rate increase factors, and obtains the dynamically variable failure rate function and reliability of the equipment.

[0037] In the above method, optionally, the economic analysis in S4 includes the following contents:

[0038] Based on the dynamically variable failure rate function, the costs of three maintenance activities, namely preventive maintenance, fault repair and complete replacement, are analyzed respectively. With the goal of minimizing the maintenance cost per unit time under the premise of meeting the minimum reliability requirements of the equipment throughout the entire life cycle, an optimal maintenance model that balances equipment mission reliability and economy is constructed.

[0039] In the above method, optionally, the specific steps in S5 include:

[0040] S501 obtains the usage time and mileage data of each device;

[0041] S502 calculates the Pearson correlation coefficient between the use time and the running mileage of different equipment, and determines the mapping relationship between the equipment reliability and the use time and the running mileage;

[0042] S503 solves the maintenance model in S4 to obtain the optimal preventive maintenance cycle, and integrates the mapping relationship obtained in S502 to obtain a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage.

[0043] In the above method, optionally, the specific contents in S6 include:

[0044] The dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage obtained in S503 is executed, and the life distribution and maintenance strategy are optimized according to the new fault data generated during the operation process.

[0045] It can be seen from the above technical solution that, compared with the prior art, the refined maintenance method of the high-speed train automatic protection system throughout the life cycle of the present invention has the following beneficial effects:

[0046] 1) A refined ATP maintenance method for the entire life cycle is proposed. The refinement is reflected in the following aspects: different equipment has different maintenance methods; different equipment has different maintenance methods at different stages of its life cycle; and maintenance methods are updated in a timely manner;

[0047] 2) Based on equipment failure data and reliability tests, the optimal life distribution of each equipment was quantitatively determined;

[0048] 3) A maintenance model based on equipment life distribution was established; a dynamically changing failure rate function was introduced to quantitatively derive a maintenance strategy that is both reliable and economical;

[0049] 4) The correlation between maintenance strategy and operating mileage and operation time is revealed, which reduces the impact of intermittent operation of equipment on maintenance strategy;

[0050] 5) Through dynamically variable maintenance strategies throughout the entire life cycle, maintenance costs are effectively reduced while ensuring minimum reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0052] Figure 1 A flow chart of a refined maintenance method for a high-speed train automatic protection system throughout its life cycle provided by the present invention;

[0053] Figure 2 A flowchart of a method for fine-grained maintenance of a high-speed train automatic protection system throughout its life cycle provided by an embodiment of the present invention;

[0054] Figure 3 A detailed maintenance diagram of a certain high-speed train automatic protection system sub-equipment during the entire life cycle provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. 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.

[0056] In this application, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "comprise one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0057] The present invention can be used in many general or special computing device environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multi-processor devices, distributed computing environments including any of the above devices or devices, etc.

[0058] Reference Figure 1-2 As shown, the present invention discloses a refined maintenance method for a high-speed train automatic protection system throughout its life cycle, comprising the following steps:

[0059] S1, data acquisition step: obtaining fault data sets of various devices of the high-speed train automatic protection system;

[0060] S2, data preprocessing step: perform abnormal data inspection on the fault data set obtained in S1 to obtain a qualified fault data set;

[0061] S3, reliability test step: perform reliability test based on the qualified failure data set obtained in S2 to obtain the optimal life distribution of different equipment;

[0062] S4, maintenance model establishment step: based on the optimal life distribution obtained in S3, establish an optimal maintenance model that balances equipment mission reliability and economy;

[0063] S5, maintenance strategy generation step: Based on the optimal maintenance model of equipment task reliability and economic balance obtained in S4, the mapping relationship between equipment reliability and operation time and running mileage is analyzed through the Pearson correlation coefficient, and a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operation time and running mileage is output;

[0064] S6. Maintenance implementation steps: Based on the full life cycle maintenance strategy obtained in S5, the high-speed train automatic protection system is maintained, and the maintenance data is saved and imported as a new fault data set.

[0065] Furthermore, the specific contents of S1 include:

[0066] Within the selected date range, for the failures of each device in the high-speed train automatic protection system, the name, type, start date of use and date of failure of the device are determined, and the life data samples X of different devices are obtained as the failure data set.

[0067] Specifically, the sub-equipment of the ATP system is mainly divided into 8 categories according to its function, including VC, BTM and DMI. ATP is a repairable system. After the equipment fails, it is repaired and put into use again. Its life is defined as the working time before the first failure or the working time between two adjacent failures. Therefore, the test data collected from the field includes maintenance activities, reflecting the reliability level of the equipment under the actual use environment and maintenance conditions. Within the selected date range, for each failure, the name, type, start date of use and failure date of the equipment are determined to obtain the life data sample X of different equipment.

[0068] Furthermore, the specific contents of S2 include: judging whether there is abnormal data in the fault data set based on the optimized Laida criterion, and removing the abnormal data to obtain a qualified fault data set.

[0069] Specifically, the basic logic for determining the distribution of life data is to use samples to infer the population. At present, the samples of ATP failure information are recorded manually, and important information such as the failure date may contain errors. Therefore, abnormal data in the sample should be removed first. The reasons for the existence of abnormal data include but are not limited to:

[0070] The manually recorded equipment start date is earlier than the train production date.

[0071] The device start date is empty.

[0072] The device has no operation record, but has fault record.

[0073] Fault record data is duplicated.

[0074] Equipment is replaced due to technical transformation, upgrading, etc.

[0075] The mean of the life sample X of a certain equipment is μ, and the standard deviation is σ. The Laida criterion states that if a certain x j satisfy

[0076] x j -μ≥3σ (1)

[0077] Then, x j are considered outliers.

[0078] Formula (1) can be expressed as

[0079] x j ≤a1=μ-3σ (2)

[0080] or

[0081] x j ≥a2=μ+3σ (3)

[0082] When using the Laida criterion, if there are too many abnormal points in the life data of the ATP equipment, the threshold range for determining the abnormal points will be very large, so that only extremely abnormal data can be eliminated, and sub-abnormal values ​​cannot be eliminated. To this end, the coefficient in front of the standard deviation is optimized. In formula (1), μ and σ are calculated after removing the maximum and minimum values ​​of the data set, and b = σ / μ. Similarly, the ratio calculated before removing the extreme value is recorded as b1. When the b value changes, a1 and a2 are calculated as follows:

[0083]

[0084] In the embodiment of the present invention, the optimized Laida criterion is used to process the abnormal life data of different sub-devices of different types of ATP, so as to provide a relatively accurate data source for the subsequent reliability mathematical distribution test.

[0085] Furthermore, the specific steps in S3 include:

[0086] S301 estimates the qualified fault data set obtained in S2 using the maximum likelihood estimation method to obtain parameter values ​​of possible life distribution types;

[0087] S302, based on the parameter values ​​of the possible life distribution types obtained in S301, a KS test is performed to determine whether the equipment life obeys a certain distribution, and test results of different life distribution types are obtained;

[0088] S303 analyzes and compares the test results obtained in S302 to obtain the life distribution type with the highest degree of consistency with the actual data as the optimal life distribution.

[0089] Specifically, in S301, the ATP equipment life sample set is X = (x1, x2, ..., x n) , n represents the sample size. The probability density function of the overall distribution is expressed as f(x; θ). Since the ATP equipment life distribution is continuous, the probability of the random variable taking a specific value is zero. Assume that the sample x i Falling on (x i ,x i +dx i ) is approximately f(x;θ)dx i , i = 1, 2, ..., n, then, any sample observation x i The probability of occurrence is approximately:

[0090]

[0091] Among them, dx1, dx2, …, dx n It can be regarded as fixed. To maximize the probability of the observed value, we only need to maximize formula (7).

[0092]

[0093] Solving equations

[0094]

[0095] The estimated value of the maximum likelihood estimate can be obtained is the parameter value of the assumed distribution.

[0096] In S302, hypothesis testing is performed on the life distribution type. ATP equipment is an electronic product, and failure is random. The failure of the entire electronic product due to corrosion, wear, or failure of a weak link may be a Weibull distribution. Fatigue caused by electronic components may be a lognormal distribution or a Weibull distribution. Equipment performance degradation or hardware failure caused by critical parameter failure generally conforms to the normal distribution. The failure of electronic devices caused by electromagnetic interference has nothing to do with the length of working time and is suitable for exponential distribution. Therefore, four typical distributions, namely Weibull distribution, lognormal distribution, normal distribution and exponential distribution, are selected as candidate distributions. These are also the most common distributions of electronic product life. In addition, Weibull distribution has a wide range of applications. Depending on the value of the shape parameter, it can be approximated to exponential distribution, Rayleigh distribution, normal distribution, etc. The life distribution of ATP is inseparable from the common properties of electronic products and must obey a known common distribution. Assuming that the distribution is a normal distribution, the method shown in formula (7) is used to estimate the mean and variance of the parameter values ​​of the normal distribution. Similarly, the parameters of other distributions can be estimated. After obtaining the specific parameter values ​​of the distribution, a specific theoretical distribution is actually determined. Next, test whether there is a significant difference between the population from which the sample comes and the specified theoretical distribution.

[0097] The KS test (Kolmogorov-Smirnov test) is a method based on the cumulative distribution function to test whether a sample obeys a given distribution.

[0098] The KS method defines the following statistics,

[0099]

[0100] in, represents the empirical distribution function of the sample, F(x) represents the empirical distribution function of the population, and sup represents the maximum value.

[0101] Test hypothesis H0: F(x) = F0(x), where F0(x) is a known continuous distribution, such as the Weibull distribution.

[0102] The elements in the ATP equipment life sample set X may be repeated. Arrange all elements in order as follows:

[0103] x (1) <x (2) <…<x (k) ,1≤k≤n (10)

[0104] Among them, duplicate data are merged into one.

[0105] Assume n i is the frequency of x(i) in the sample, then:

[0106] n i ≥1,n1+n2+…+n k =n (11)

[0107]

[0108] make

[0109]

[0110] At this time, formula (9) can be transformed into:

[0111] D n =max{d1,d2,...,d k} (13)

[0112] The KS method proves that The limiting distribution of is:

[0113]

[0114] For a given significance level α, the probability of equation (15) is recorded as the p-value.

[0115]

[0116] From equations (14) and (15), we can see that the p-value is an extreme value. If p>α, then the hypothesis H0 is accepted, otherwise, H0 is rejected. It can be seen that the p-value is the probability of the sample observation result or a more extreme result when the null hypothesis is true. The larger the p-value, the greater the statistic D n The smaller it is, the less significant the difference between the probability distribution of the sample and the hypothesised probability distribution is, so the reason for accepting the hypothesis H0 is more sufficient.

[0117] In S303, for a specific device, the processing logic is as follows:

[0118] (1) When the p values ​​of multiple candidate distributions of a certain device are all greater than α, the larger the p value, the greater the possibility of obeying the corresponding candidate distribution. Therefore, the candidate distribution corresponding to the maximum p value is taken as the life distribution of the device.

[0119] (2) When only one candidate distribution has a p value greater than α, this candidate distribution is selected as the equipment life distribution.

[0120] (3) When the p-values ​​of all candidate distributions are less than α, a probability graph of the possible candidate distributions is drawn, and the distribution is judged based on whether the data points are more concentrated on the diagonal. The smaller the goodness-of-fit statistic, the more concentrated the data points are on the diagonal, indicating that the expected cumulative probability is more consistent with the measured cumulative probability.

[0121] In summary, first assume that the lifespan of the ATP device follows a certain probability distribution, then estimate the parameter values of the assumed distribution through the maximum likelihood method, determine the specific form of the assumed distribution, and finally calculate the statistic D and p-value using the K-S method to test whether the hypothesis holds and select the optimal lifespan distribution of the device. n Furthermore, the specific content in S4 includes:

[0122] Based on the optimal lifespan distribution determined in S3, introduce a dynamically changing failure rate function to describe the impact of maintenance behavior on device performance, including the analysis of device reliability and maintenance costs; establish an optimal maintenance model that balances the reliability and economy of device tasks, and solve the maintenance model through genetic algorithms or particle swarm algorithms to obtain the length of the maintenance cycle.

[0123] Furthermore, the reliability analysis in S4 includes the following content:

[0124] Reliability analysis mainly describes the changes in device performance before and after maintenance by introducing an age reduction factor and a failure rate increase factor, and obtains the dynamically variable failure rate function and reliability of the device.

[0125] Specifically, after each preventive maintenance of the ATP system or device, the device performance cannot be restored to as good as new. To accurately describe the dynamic changes in device performance before and after maintenance, an age reduction factor and a failure rate increase factor are introduced. The relationship between the failure rate functions of the device before and after preventive maintenance is as follows:

[0126]

[0127]

[0128]

[0129] i where λ(t) represents the failure rate function before the i-th maintenance. i represents the number of maintenance times, i = 1, 2,..., N. λ1(t) is the failure rate function of the device in the initial state, and its specific expression is obtained from the results of the hypothesis test in S3. T i represents the time interval between the (i - 1)-th and i-th maintenance of the device, and any two T i are not necessarily equal, indicating that the time interval of device maintenance changes dynamically. r represents the age reduction factor, 0 < r < 1, c represents the failure rate increase factor, c > 1. It should be particularly noted that the maintenance here refers to preventive maintenance. Furthermore, the economic analysis in S4 includes the following content:

[0130] Based on the dynamically variable failure rate function, the costs of three maintenance activities, namely preventive maintenance, fault repair and complete replacement, are analyzed respectively. With the goal of minimizing the maintenance cost per unit time under the premise of meeting the minimum reliability requirements of the equipment throughout the entire life cycle, an optimal maintenance model that balances equipment mission reliability and economy is constructed.

[0131] Specifically, at present, ATP implements a maintenance model that combines preventive maintenance and post-fault maintenance. Among them, the former is a proactive arrangement planned in advance, and the latter is an uncontrollable passive behavior. Based on the actual situation of ATP maintenance, the core of the refined maintenance of each ATP equipment is to determine the planned preventive maintenance cycle. If the cycle is too long, the failure of ATP equipment cannot meet the safe and reliable transportation mission requirements of the train; if the cycle is too short, there will be excess repairs, resulting in excessive maintenance costs. Therefore, it is necessary to balance the length of the maintenance cycle and comprehensively consider the relationship between equipment reliability, mission and economy.

[0132] First of all, ensuring the safe and reliable operation of trains is the basic responsibility of ATP equipment. All operation and maintenance personnel should strictly limit the number of events that affect train safety. Therefore, equipment reliability is the primary factor in considering the length of the maintenance cycle. Secondly, trains need to run on time and efficiently. The transportation tasks undertaken by ATP equipment are an important factor in considering the length of the maintenance cycle. In addition, maintenance costs affect the allocation of maintenance resources and the frequency of maintenance. Spending the least on maintenance costs is the goal pursued by railway companies. Therefore, maintenance economy is also one of the factors that need to be considered. Reliability and mission are reflected through the reliability of equipment, and economy is achieved by controlling maintenance costs. In summary, the decision-making of the refined maintenance strategy of ATP equipment mainly considers the following factors: the reliability of the equipment is not lower than a certain threshold, and the maintenance cost of the equipment is as low as possible.

[0133] Knowing the failure rate function, in any time interval [t i-1 ,t i ]The number of failures is:

[0134]

[0135] Similarly, the number of failures during the entire life cycle is:

[0136]

[0137] Among them, t0=0 represents the initial time, t i represents the time of the ith maintenance. Then, T i =t i -t i-1 .

[0138] During the entire life cycle of ATP equipment, in addition to preventive maintenance and breakdown maintenance, there is also a complete replacement of the equipment at the end of its life. Corresponding to this are three aspects of maintenance costs, which are as follows.

[0139] (1) Preventive maintenance costs

[0140] Suppose the cost of a single preventive maintenance of a certain equipment is C p , in the N maintenances in the whole life cycle, the last maintenance is replacement, which means the end of life. Therefore, the number of preventive maintenance is N-1. Then, the total cost of preventive maintenance is C sp for:

[0141] C sp =(N-1)*C p (19)

[0142] (2) Fault repair costs

[0143] Suppose the cost of repairing a single failure of a certain equipment is C f , and the number of failures is H. Then, the total cost of failure repair is C sf for:

[0144]

[0145] (3) Complete replacement cost

[0146] Suppose the cost of completely replacing a certain equipment is C r .

[0147] In summary, the total maintenance cost of the equipment during its entire life cycle is:

[0148] C s =C sp +C sf +C r (twenty one)

[0149] The maintenance of ATP equipment is carried out during a specific period of time when the train is not performing transportation tasks. This period of time is planned and specially arranged. The maintenance does not affect the normal tasks of the equipment. Therefore, the time spent on preventive maintenance and equipment replacement can be ignored. The maintenance time of the entire life cycle is the sum of the equipment service time and the fault repair time, that is:

[0150]

[0151] Among them, t f Indicates the time it takes to repair a single equipment failure.

[0152] Thus, the maintenance cost per unit time during the entire life cycle is the maintenance cost rate C. es for:

[0153]

[0154] Based on the requirement of minimum equipment reliability, the ATP life cycle refined maintenance model is shown in formula (24). Its essence is a nonlinear programming problem to minimize maintenance costs under the constraint of specific reliability.

[0155]

[0156] Among them, C es is the objective function, the number of maintenance N and the maintenance interval T i is the decision variable, R i (t) represents the reliability of the equipment, R min is the minimum reliability requirement, N * is a set of positive integers. The minimum reliability requirement, the relationship between the maintenance interval and the failure rate together constitute the constraint conditions.

[0157] By solving equation (24) using genetic algorithms, particle swarm optimization algorithms, etc., we can obtain the optimal number of preventive maintenance times N and maintenance interval T. i This is the most important manifestation of the refined maintenance strategy.

[0158] Furthermore, the specific steps in S5 include:

[0159] S501 obtains the usage time and mileage data of each device;

[0160] S502 calculates the Pearson correlation coefficient between the use time and the running mileage of different equipment, and determines the mapping relationship between the equipment reliability and the use time and the running mileage;

[0161] S503 solves the maintenance model in S4 to obtain the optimal preventive maintenance cycle, and integrates the mapping relationship obtained in S502 to obtain a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage.

[0162] Specifically, due to the characteristics of railway transportation, ATP equipment does not operate around the clock. From a time perspective, the working time described in the reliability distribution is calculated as the sum of natural days between adjacent fault intervals, rather than being strictly limited to the sum of the time spent on transport tasks. From a mileage perspective, the running mileage of ATP each time it undertakes a transport task can be accurately calculated. The size of the running mileage is a better measure of the use of ATP than the size of the operating time. In fact, daily maintenance work on the railway site is also arranged based on the operating time or operating mileage. As long as the correlation between the operating time and the operating mileage can be obtained, the relationship between reliability and time can be mapped to the relationship between reliability and mileage.

[0163] Two variables r i and rj The correlation is expressed by the Pearson correlation coefficient, which is calculated as:

[0164]

[0165] The range of values ​​for operating time and mileage varies greatly, and they have different dimensions. Using the Pearson correlation coefficient can eliminate the influence of the range and dimension on the correlation coefficient.

[0166] The calculation process of ATP application time and running mileage is as follows.

[0167] The running mileage of a certain ATP system on day j is represented by um j , then the cumulative running mileage of the ATP up to the i-th day is am i , expressed as:

[0168]

[0169] Where um1 represents the initial value of the mileage of an ATP system, that is, the mileage of the first transportation task. If there is no transportation task on the jth day, then um j =0.

[0170] For the same day i, there are na i The same type of ATP system is operating normally, and the average cumulative running mileage is calculated as mm i :

[0171]

[0172] Among them, am li represents the cumulative running mileage of the lth ATP system as of the i-th day, and the calculation method is shown in formula (26).

[0173] When processing raw data, special attention should be paid to the setting of initial values, because ATP equipment is put into use in batches gradually, not all at once.

[0174] Different types of ATP equipment have different operating mileages. Equipment types are represented as:

[0175] ta={A,B,C,D,E} (28)

[0176] The average cumulative running mileage of the five types of ATP can be obtained from formula (27). Similarly, the corresponding operating time can be obtained, calculated in days.

[0177] The data granularity is refined to different sub-equipment of each ATP type, and the operating time and mileage of different sub-equipment when failure occurs are extracted. Then, according to formula (25), the Pearson correlation coefficient of the operating time and mileage of different equipment can be calculated. If the smallest Pearson correlation coefficient is greater than 0.8, it means that the operating time and mileage of each type of sub-equipment are strongly correlated. This proves that there is indeed a certain mathematical correlation between the two, and the single-dimensional relationship between reliability and operating time can be mapped to a two-dimensional relationship between reliability and operating time and operating mileage. Similarly, for any type of sub-equipment, the corresponding two-dimensional relationship can be obtained. Furthermore, the maintenance strategy formulated based on the equipment reliability shown in formula (24) can be implemented in two dimensions according to the operating time or mileage. It emphasizes the dominant factor of mileage in the maintenance strategy, which conforms to the characteristics of intermittent operation of ATP equipment and is feasible.

[0178] After obtaining the optimal preventive maintenance cycle, the mapping relationship between time and mileage can be combined to obtain a refined maintenance strategy for the entire life cycle, such as Figure 3 shown.

[0179] Furthermore, the specific contents of S6 include:

[0180] The dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage obtained in S503 is executed, and the life distribution and maintenance strategy are optimized according to the new fault data generated during the operation process.

[0181] Specifically, for each type of equipment, after implementing a refined maintenance strategy, the life distribution is optimized based on the new fault data generated during operation, and at the same time, the maintenance strategy is optimized. It can be seen that the maintenance strategy is not static, but is continuously optimized and updated throughout the life cycle.

[0182] It can be seen that the refined maintenance method of the present invention is firstly embodied in differentiating the types of ATP. In addition, it is also embodied in the following aspects.

[0183] 1) The maintenance strategies for different sub-equipment are different.

[0184] 2) The maintenance strategies are different in different time periods during the entire life cycle.

[0185] 3) Prevent dynamic changes in maintenance cycles.

[0186] 4) The maintenance strategy for the same equipment is continuously updated.

[0187] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0188] Those skilled in the art may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, computer software, or a combination of both.

[0189] In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0190] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A refined maintenance method for the full life cycle of a high-speed train automatic protection system, characterized in that: The following steps are involved: S1, data acquisition step: obtaining fault data sets of various devices of the high-speed train automatic protection system; S2, data preprocessing step: perform abnormal data inspection on the fault data set obtained in S1 to obtain a qualified fault data set; S3, reliability test step: perform reliability test based on the qualified failure data set obtained in S2 to obtain the optimal life distribution of different equipment; S4, maintenance model establishment step: based on the optimal life distribution obtained in S3, establish an optimal maintenance model that balances equipment mission reliability and economy; S5, maintenance strategy generation step: Based on the optimal maintenance model of equipment task reliability and economic balance obtained in S4, the mapping relationship between equipment reliability and operation time and running mileage is analyzed through the Pearson correlation coefficient, and a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operation time and running mileage is output; S6. Maintenance implementation steps: Based on the full life cycle maintenance strategy obtained in S5, the high-speed train automatic protection system is maintained, and the maintenance data is saved and imported as a new fault data set.

2. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 1, characterized in that: The specific contents of S1 include: Within the selected date range, for the failures of each device in the high-speed train automatic protection system, the name, type, start date of use and date of failure of the device are determined, and the life data samples X of different devices are obtained as the failure data set.

3. A method for the refined maintenance of a high-speed train automatic protection system throughout its life cycle according to claim 2, characterized in that: The specific contents of S2 include: judging whether there is abnormal data in the fault data set based on the optimized Laida criterion, and removing the abnormal data to obtain a qualified fault data set.

4. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 3, characterized in that: The specific steps in S3 include: S301 estimates the qualified fault data set obtained in S2 using the maximum likelihood estimation method to obtain parameter values ​​of possible life distribution types; S302, based on the parameter values ​​of the possible life distribution types obtained in S301, a KS test is performed to determine whether the equipment life obeys a certain distribution, and test results of different life distribution types are obtained; S303 analyzes and compares the test results obtained in S302 to obtain the life distribution type with the highest degree of consistency with the actual data as the optimal life distribution.

5. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 4, characterized in that: The specific contents of S4 include: Based on the optimal life distribution determined in S3, a dynamically changing failure rate function is introduced to characterize the impact of maintenance behavior on equipment performance, including analysis of equipment reliability and maintenance cost; an optimal maintenance model that balances equipment mission reliability and economy is established, and the maintenance model is solved through genetic algorithm or particle swarm algorithm to obtain the length of the maintenance cycle.

6. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 5, characterized in that: The reliability analysis in S4 includes the following: reliability The analysis mainly introduces the service age regression factor and the failure rate increasing factor to describe the changes in equipment performance before and after maintenance, and obtains the dynamically variable failure rate function and reliability of the equipment.

7. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 5, characterized in that: The economic analysis in S4 includes the following: Based on the dynamically variable failure rate function, the costs of maintenance activities in three aspects, namely, preventive maintenance, breakdown maintenance and complete replacement, are analyzed respectively. With the goal of minimizing the maintenance cost per unit time under the premise of meeting the minimum reliability requirements of the equipment throughout the entire life cycle, an optimal maintenance model that balances equipment mission reliability and economy is constructed.

8. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 7, characterized in that: The specific steps in S5 include: S501 obtains the usage time and mileage data of each device; S502 calculates the Pearson correlation coefficient between the use time and the running mileage of different equipment, and determines the mapping relationship between the equipment reliability and the use time and the running mileage; S503 solves the maintenance model in S4 to obtain the optimal preventive maintenance cycle, and integrates the mapping relationship obtained in S502 to obtain a dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage.

9. A refined maintenance method for the entire life cycle of a high-speed train automatic protection system according to claim 8, characterized in that: The specific contents of S6 include: The dynamic time-varying maintenance strategy based on the dual-dimensional fusion of operating time and running mileage obtained in S503 is executed, and the life distribution and maintenance strategy are optimized according to the new fault data generated during the operation process.