A method for evaluating the quality of rail transit signaling equipment

Abstract

This invention discloses a quality evaluation method for rail transit signaling equipment, belonging to the field of rail transit operation and maintenance. This invention incorporates each alarm type of every rail transit signaling device into the evaluation index. When an alarm is triggered, it not only provides the alarm level but also pinpoints a specific fault in a particular device, resulting in more accurate quality evaluation. This provides more professional guidance to maintenance personnel in the electrical engineering department, improving maintenance efficiency. Furthermore, it calculates the quality score of the rail transit signaling equipment using a nonlinear quadrature algorithm formula. This method not only evaluates the quality of the signaling equipment, its type, and the overall system but also produces a line graph that more closely approximates the equipment's degradation curve. Additionally, it further improves the prediction accuracy of the equipment's remaining lifespan through time-series autoregressive prediction.

Description

Technical Field

[0001] This invention relates to the field of rail transit operation and maintenance, and in particular to a method for evaluating the quality of rail transit signaling equipment. Background Technology

[0002] The centralized railway signal monitoring system is known as the "black box" of the electrical engineering system. The electrical engineering department can promptly detect various types of faults in the signal system and issue fault alarms. It plays an important role in monitoring the status of signal equipment, discovering potential problems with signal equipment, guiding on-site maintenance, improving the maintenance level of the electrical engineering department, and ensuring train operation safety.

[0003] Electrical maintenance personnel typically perform equipment maintenance in the following two situations:

[0004] (1) When the equipment alarms, the alarm should be handled according to the alarm information to restore the equipment to normal operation;

[0005] (2) Conduct regular inspections and maintenance of the equipment within the pipeline once a month to keep it in normal condition.

[0006] While regular scheduled inspections and maintenance can comprehensively identify problems with signal equipment, maintenance personnel need to test, inspect, and maintain each piece of signal equipment individually, resulting in a high workload and an inability to handle emergencies, leading to poor timeliness.

[0007] While processing signal equipment based on alarm information is more time-sensitive and less labor-intensive than scheduled routine inspections and maintenance, current industry software for evaluating the quality of signal equipment is inadequate. The only factor considered in the quality evaluation is the alarm itself. More precisely, existing quality evaluation methods only incorporate alarm-related information (including alarm level and alarm type), without detailed classification of alarm types. Furthermore, the deduction principle for alarm level quality evaluation is rather simplistic (usually deducting X1 points for each Level 1 alarm, X2 points for each Level 2 alarm, and so on), a simple linear subtraction algorithm. However, the degradation curve of signal equipment is non-linear.

[0008] Therefore, the evaluation methods used in existing centralized railway signal monitoring systems cannot obtain accurate alarm results or accurately reflect the quality of the equipment. They can only provide a general indication to the maintenance personnel of the electrical engineering department. After arriving at the site, the maintenance personnel still need to conduct a more precise quality assessment of the equipment based on their own experience and professional skills to ensure its normal operation. If the maintenance personnel lack experience and professional skills, they will spend a lot of time on quality assessment, resulting in low work efficiency. If the maintenance personnel only deal with superficial problems and overlook deeper hidden dangers during the quality assessment, not only will they have done useless work, but it will also lead to the eventual failure of the signal equipment and cause accidents. In addition, the evaluation results of existing signal equipment evaluation methods do not consider the detailed alarm types. Therefore, it is impossible to accurately obtain the degradation curve of the signal equipment based on the collected evaluation results, reasonably predict the remaining service life of the signal equipment, or provide relevant references for the electrical engineering department when an alarm is triggered. Summary of the Invention

[0009] The purpose of this invention is to provide a method for evaluating the quality of rail transit signaling equipment in order to solve the above-mentioned problems.

[0010] This invention is achieved through the following technical solution:

[0011] A method for evaluating the quality of rail transit signaling equipment includes the following steps:

[0012] S1. Set evaluation indicators for each of the n rail transit signaling devices. The evaluation indicators include the total score, alarm type, and the deduction ratio K corresponding to a single alarm type. t And the number of alarm signals A t ; where n is a positive integer;

[0013] S2. Monitor the alarms of n rail transit signal devices respectively, and collect the alarm signals when a rail transit signal device issues an alarm signal;

[0014] S3. Based on the collected alarm signals, evaluate the equipment quality of the corresponding rail transit signaling equipment;

[0015] Specifically, step S3 includes the following sub-steps:

[0016] S301. Collect alarm signals and extract evaluation indicators for the rail transit signaling equipment corresponding to the alarm signals;

[0017] S302. Based on the evaluation indicators set for the corresponding rail transit signaling equipment, substitute them into the nonlinear quadrature algorithm formula to evaluate the corresponding rail transit signaling equipment and obtain the quality score of the corresponding rail transit signaling equipment.

[0018] The alarm types include directly acquired alarm data types and indirectly acquired alarm data types. The directly acquired alarm data types include alarm information input from external data, and the indirectly acquired alarm data types include at least the equipment's track entry time, the number of times the turnout equipment was operated, the number of times the train was pressed, the number of times the track equipment was pressed, and the number of times the signal equipment was lit.

[0019] Furthermore, in step S302, the specific calculation process for the score evaluation is as follows:

[0020]

[0021] Wherein, Q i The quality score of the rail transit signaling equipment is represented by t, which is a variable representing the number of alarm types.

[0022] Furthermore, step S3 also includes adding / subtracting points for the equipment quality evaluation. The specific steps are as follows: extract the maintenance records of the rail transit signal equipment and add / subtract points based on the number of maintenance records.

[0023] Furthermore, the total score is initialized to 100 points each day.

[0024] Furthermore, the rail transit signaling equipment includes basic signaling equipment, intelligent equipment, and central equipment, wherein when calculating the quality score of a single basic signaling device, a corresponding deduction standard is assigned to the on-track time of that basic signaling device.

[0025] Furthermore, step S4 is included, which involves visualizing the alarm type, alarm frequency, and quality score.

[0026] Furthermore, the visualization process includes generating statistical tables, bar charts, line charts, and pie charts for each rail transit signaling device.

[0027] Furthermore, the types, frequencies, and quality scores of alarms generated by each signaling device daily are collected, and time-series autoregressive predictions are made for the quality scores of each rail transit signaling device in the future.

[0028] Furthermore, the specific times when alarms were generated by various rail transit signaling devices on the same day were collected, and correlation analysis was performed on different alarm types of the same signaling device and alarm types of different signaling devices within a certain time period.

[0029] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0030] 1. In this invention, each alarm type of each rail transit signaling device is included in the evaluation index. When an alarm is triggered, not only is the alarm level given, but it can also be specified to a specific fault of a particular device. This makes the quality evaluation more accurate, provides more professional guidance to the maintenance personnel of the electrical department, and improves maintenance efficiency.

[0031] 2. In this invention, the quality score of rail transit signaling equipment is calculated using a nonlinear quadrature algorithm formula. This not only evaluates the quality of the signaling equipment, the type of equipment to which the signaling equipment belongs, and the system as a whole, but also produces a line graph that is closer to the degradation curve of the equipment itself. Furthermore, the prediction accuracy of the remaining lifespan of the equipment is further improved through time-series autoregressive prediction.

[0032] 3. In this invention, correlation analysis is performed on different alarm types to reasonably predict the correlation between different faults. When maintenance personnel in the power department are handling a fault, they are prompted to indicate other faults that may accompany the fault, which helps to eliminate hidden dangers and avoid secondary operations. Attached Figure Description

[0033] The above and other objects, features, and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings. In the exemplary embodiments, the same reference numerals generally represent the same components. The accompanying drawings, which are included to provide a further understanding of embodiments of the present invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0034] Figure 1 This is a flowchart of Embodiment 1 of the present invention;

[0035] Figure 2 This is a flowchart of Embodiment 2 of the present invention;

[0036] Figure 3 This is the external power grid evaluation index table proposed in Embodiment 1 of the present invention;

[0037] Figure 4 The turnout evaluation index table proposed in Embodiment 1 of this invention;

[0038] Figure 5 This refers to the track evaluation index table proposed in Embodiment 1 of the present invention;

[0039] Figure 6 The signal evaluation index table proposed in Embodiment 1 of this invention;

[0040] Figure 7 The table of evaluation indicators for turnout gaps proposed in Embodiment 1 of this invention;

[0041] Figure 8The uninterruptible power supply (UPS) evaluation index table proposed in Embodiment 1 of the present invention;

[0042] Figure 9 The battery evaluation index table proposed in Embodiment 1 of this invention;

[0043] Figure 10 The power supply screen evaluation index table proposed in Embodiment 1 of this invention;

[0044] Figure 11 The axle counting evaluation index table proposed in Embodiment 1 of this invention;

[0045] Figure 12 The computer interlocking evaluation index table proposed in Embodiment 1 of this invention;

[0046] Figure 13 The evaluation index table for the train control center proposed in Embodiment 1 of this invention;

[0047] Figure 14 The evaluation index table for the Centralized Train Dispatch Control System (CTC) / Dispatch and Control System (TDCS) covering the entire railway network proposed in Embodiment 1 of this invention;

[0048] Figure 15 The following is the performance table of the non-insulated frequency-shift automatic block (ZPW2000) system proposed in Embodiment 1 of the present invention;

[0049] Figure 16 The environmental evaluation index table proposed in Embodiment 1 of this invention;

[0050] Figure 17 This refers to the comprehensive monitoring and evaluation index table for the interval proposed in Embodiment 1 of the present invention;

[0051] Figure 18 The evaluation index table for Temporary Rate Limiting Server (TSRS) proposed in Embodiment 1 of this invention;

[0052] Figure 19 The evaluation index table for Radio Block Center (RBC) proposed in Embodiment 1 of this invention;

[0053] Figure 20 This is the evaluation index table for the nationwide dispatching and command management system (TDCS) / train dispatching centralized command and control system (CTC) proposed in Embodiment 1 of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are for illustrative purposes only and are not intended to limit the invention. It should be noted that this invention is already in the actual research and development stage.

[0055] Example 1

[0056] like Figure 1 As shown in this embodiment, a method for evaluating the quality of rail transit signaling equipment is provided, specifically as follows:

[0057] S1 specifies the type of rail transit signaling equipment;

[0058] S2 sets the alarm types and corresponding deduction standards for rail transit signaling equipment;

[0059] S3 monitors the types and frequency of alarms from rail transit signaling equipment;

[0060] S4 calculates the quality score for each signal device, the device type of each signal device, and the overall system.

[0061] S5 generates statistical tables, bar charts, line charts, and pie charts for alarm types, frequencies, and quality scores.

[0062] Preferably, the types of rail transit signaling equipment in S1 include basic signaling equipment, intelligent equipment, and central equipment. Specifically, the basic signaling equipment includes external power grid, turnouts, tracks, and signal lights.

[0063] The alarm types of the external power grid are: external power grid input power phase failure / power failure, external power grid three-phase power supply sequence error, external power grid input power instantaneous power failure, external power grid dual-path power failure alarm, electrical characteristic over-limit, analog quantity change trend, sudden change, abnormal fluctuation and intelligent analysis fault diagnosis.

[0064] The alarm types for the turnout are turnout derailment, electrical characteristic over-limit, turnout usage time over-limit, turnout no indication, analog quantity change trend, sudden change, abnormal fluctuation, intelligent analysis early warning and intelligent analysis fault diagnosis.

[0065] The alarm types for the track are: electrical characteristic over-limit, analog quantity change trend, sudden change, abnormal fluctuation, intelligent analysis and early warning, intelligent analysis and fault diagnosis, safety supervision alarm and section equipment alarm.

[0066] The alarm types of the signal controller are electrical characteristic over-limit, safety supervision alarm, analog quantity change trend, sudden change, abnormal fluctuation, and intelligent analysis fault diagnosis.

[0067] The intelligent equipment includes turnout gaps, uninterruptible power supplies (UPS), batteries, power supply panels, axle counters, computer interlocking, train control center, centralized train dispatching and control system (CTC) / dispatching and command management system (TDCS) covering the entire railway, non-insulated frequency shift automatic block system (ZPW2000), dynamic environment and section integrated monitoring;

[0068] The alarm type for the turnout gap is turnout gap alarm;

[0069] The alarm types of the uninterruptible power supply (UPS) are electrical characteristic over-limit, intelligent power panel alarm, analog quantity change trend, sudden change, abnormal fluctuation and intelligent analysis fault diagnosis.

[0070] The alarm types for the battery are electrical characteristic over-limit, intelligent power supply panel alarm, and analog quantity change trend, sudden change, and abnormal fluctuation.

[0071] The alarm types of the power supply panel are: electrical characteristic over-limit, intelligent power supply panel alarm, power supply panel output power failure alarm, analog quantity change trend, sudden change, abnormal fluctuation, power supply panel output phase loss, and power supply panel output error.

[0072] The alarm type for the axle counting system is the axle counting system alarm;

[0073] The alarm types of the computer interlocking are computer interlocking system alarm, safety supervision alarm, and intelligent analysis fault diagnosis;

[0074] The alarm types of the train control center are train control system alarm and safety supervision alarm;

[0075] The alarm type of the Centralized Train Control System (CTC) / Dispatching and Control System (TDCS) covering the entire railway network is the alarm of the Centralized Train Control System (CTC) covering the entire railway network.

[0076] The alarm type of the non-insulated frequency shift automatic block (ZPW2000) system is non-insulated frequency shift automatic block (ZPW2000) system alarm;

[0077] The alarm types for the dynamic environment are environmental monitoring alarms, electrical characteristic over-limit alarms, and analog quantity change trends, sudden changes, and abnormal fluctuations.

[0078] The alarm type for the integrated monitoring of the interval is the integrated monitoring system alarm.

[0079] The central equipment includes a Temporary Speed ​​Limiting Server (TSRS), a Radio Block Center (RBC), and a Train Dispatch and Control System (TDCS) / Centralized Train Dispatch Control System (CTC) covering the entire railway line.

[0080] The alarm type of the Temporary Rate Limiting Server (TSRS) is Temporary Rate Limiting Server (TSRS) Device Alarm;

[0081] The alarm type of the Radio Block Center (RBC) is Radio Block Center (RBC) alarm;

[0082] The alarm type of the nationwide dispatching and command management system (TDCS) / train dispatching centralized command and control system (CTC) is the nationwide dispatching and command management system (TDCS) / train dispatching centralized command and control system (CTC) system alarm.

[0083] It should be noted that this invention can incorporate more equipment-related parameters to classify alarm types. For example, it can indirectly acquire alarm data types, representing alarm parameters recorded through sensors, such as: equipment track entry time, number of turnout operations and train rollovers, track rollovers, and signal light activation times. These alarm types can acquire alarm parameters through sensor collection, counter counting, etc., and the system will issue an alarm based on these parameters. Alternatively, it can directly acquire alarm types: external data directly input alarm information, representing alarm parameters directly input from external sources, such as notification information (non-alarm) obtained through real-time online analysis of equipment data. Based on the alarm parameters provided by indirect and direct alarms, a more comprehensive quality evaluation of the equipment can be performed.

[0084] The corresponding alarm types and deduction criteria in S2 are shown in the following table:

[0085] (1) Evaluation indicators for external power grids, such as Figure 3 ;

[0086] (2) Turnout evaluation indicators, such as Figure 4 ;

[0087] (3) Track evaluation indicators, such as Figure 5 ;

[0088] (4) Signal evaluation indicators, such as Figure 6 ;

[0089] (5) Evaluation indicators for intelligent devices

[0090] Smart devices include the following signaling devices

[0091] Switch gaps, such as Figure 7 ;

[0092] Uninterruptible power supply (UPS), such as Figure 8 ;

[0093] Storage batteries, such as Figure 9 ;

[0094] Power screen, such as Figure 10 ;

[0095] Axle counting, such as Figure 11 ;

[0096] Computer interlocking, such as Figure 12 ;

[0097] Train control center, such as Figure 13 ;

[0098] Centralized Train Control (CTC) / Railway-wide Dispatching and Control System (TDCS), such as Figure 14 ;

[0099] Non-insulated frequency-shift automatic block system (ZPW2000), such as Figure 15 ;

[0100] Dynamic ring, such as Figure 16 ;

[0101] Integrated monitoring of the interval, such as Figure 17 ;

[0102] (6) Evaluation indicators of central equipment

[0103] The central equipment includes the following signaling devices.

[0104] Temporary rate limiting servers (TSRS), such as Figure 18 ;

[0105] Radio Block Center (RBC), such as Figure 19 ;

[0106] A nationwide dispatching and control system (TDCS) / centralized train dispatching and control system (CTC), such as Figure 20 .

[0107] Furthermore, in step S4, the nonlinear quadrature algorithm formula used to calculate the mass fraction is as follows: The initial quality score is initialized to 100 points per day. Of course, the initial value can be adjusted according to the time situation after the signal equipment has been used for a period of time. In addition, when calculating the quality score of a single basic signal equipment, a corresponding deduction standard is assigned to the on-line time of the basic signal equipment. In this embodiment, 0.03% of the current score is deducted for each additional day of on-line time of the signal equipment.

[0108] For example, a preferred implementation scheme is further proposed, specifically as follows: assuming that on a certain day the system experiences 2 instances of external power grid input phase loss / power outage, 3 instances of instantaneous external power grid input power outage, 1 instance of turnout derailment, 1 instance of turnout gap alarm, and 1 instance of Temporary Speed ​​Limit Server (TSRS) equipment alarm, then the system's total score is Q. i =100*(1-5%) 2 *(1-1%) 3 *(1-20%)*(1-1%)*(1-1%)=68.66, the equipment rating for the external power grid (on-line time 100 days) is Q. i =100*(1-5%) 2 *(1-1%) 3 *(1-0.03%*100)=84.94, the score for the basic signal equipment is Q. i =100*(1-5%) 2 *(1-1%) 3 *(1-20%)=70.05.

[0109] Furthermore, step S3 also includes adding / subtracting points for equipment quality evaluation. The specific steps are as follows: extract the maintenance records of the rail transit signaling equipment and add / subtract points based on the number of maintenance records. For example, if the maintenance record for a certain rail transit signaling equipment for a given day is 1, then points are added to that equipment. All point additions / subtractions mentioned here are determined based on predefined standard values. For instance, if a rail transit signaling equipment requires daily inspection, the minimum maintenance record value for that equipment is set to 1. Therefore, if the maintenance record for a particular day is 0, points are subtracted. The specific point addition / subtraction values ​​can be set based on the equipment's maintenance demand, maintenance complexity, and maintenance personnel requirements.

[0110] Using Example 1, the precise quality score of a certain signal device / the type of device to which the signal device belongs / the overall system for the day can be obtained, as well as precise alarm data such as alarm type and alarm frequency. This not only helps maintenance personnel to quickly handle signal device faults, but also provides a clear understanding of the signal device degradation trend based on statistical tables, bar charts, line charts, and pie charts of alarm type, frequency, and quality score.

[0111] Example 2

[0112] like Figures 1 to 2 As shown, this embodiment, based on embodiment 1, further includes the following steps:

[0113] S6 performs time-series autoregressive prediction of future quality scores;

[0114] S7 uses the Apriori algorithm to perform correlation analysis on different alarm types.

[0115] S6 specifically involves collecting the types, frequencies, and quality scores of alarms generated by each signaling device daily, and then performing time-series autoregressive predictions on the future quality scores of each signaling device, the device type to which each signaling device belongs, and the overall system. The steps are as follows:

[0116] S61 collects data;

[0117] S62 establishes the model;

[0118] S63 divides the data into training and testing sets;

[0119] S64 training and testing models;

[0120] S65 predicts future quality scores.

[0121] In this embodiment, a DeepAR deep autoregressive network is used to perform time-series autoregressive prediction of the quality score, and a conditional distribution model obtained by the product of likelihood functions is established as follows:

[0122]

[0123] The above formula is the output h of the hidden layer. i,t A parameterized DeepAR deep autoregressive network model, where L Θ Let i be the likelihood function, t be the time series ID, t be the time point, T be the current time of the covariate, t0 be the prediction start time, and Z be the likelihood function. i,t0:T The output at time T of the i-th sequence is obtained using the output before time t0 and the covariate sequence from time 1 to time T as prior information. Let X be the output sequence from time 1 to time t0-1. i,1:T Let Z be the sequence of covariates from time 1 to time T. i,t Z is the output of the i-th sequence at time t. i,1:t-1为从 The output sequence from time 1 to the current time, where p represents the conditional probability, θ(h) i,t ,Θ) is a function of the hidden layer output, h i,t The hidden layer output is Θ, which represents the neural network parameters, p(Z) i,t |θ(h i,t ,Θ)) follows a predetermined distribution, and the parameters of this distribution are given by the function θ(h) output by the hidden layer. i,tIn this embodiment, the Gaussian likelihood function is used as the conditional distribution model. Of the collected data, 80% is used as the training set and 20% is used as the test set. The DeepAR deep autoregressive network model associates the time series with multiple groups, which can handle nonlinear problems and is more suitable for the nonlinear quality score calculation curve in this invention. It can more accurately predict the degradation curve of signal equipment.

[0124] S7 specifically involves collecting the exact times when alarms are generated by each signal device each day, and then using the Apriori algorithm to perform correlation analysis on different alarm types of the same signal device and alarm types of different signal devices within a certain time period. The steps are as follows:

[0125] S71 Data processing and itemset creation; In this embodiment, data processing is first performed, and the alarm data is sorted according to the specific time of alarm generation by each signal device each day. Then, when creating the itemset, the following two rules are satisfied simultaneously:

[0126] a) Starting from the first alarm data W1, set W1 to W (n-1) Include in itemset, W n For the first one with W1~W (n-1) If any alarm data in W is of the same type, then continue to follow the above rules. (n-1) First, the subsequent alarm data is divided;

[0127] b) Starting from the first alarm data W1, set W1 to W( n-1) Include in itemset, W n and W (n-1) If the interval between them is greater than the set threshold T0, then continue to follow the above rules from W. (n-1) The subsequent alarm data is divided according to rule a). Rule a) ensures that there are no duplicate alarm data types in each item set, which facilitates further calculation. Rule b) ensures that each item set contains alarm data with short intervals, and the alarm data in the item set has stronger correlation, resulting in more accurate calculation results.

[0128] For example, a preferred specific implementation scheme is further proposed as follows:

[0129] Example 1: On a certain day, the following alarms, all with intervals less than T0, occur sequentially in the external power grid: external power grid input phase loss / power failure, external power grid three-phase power supply sequence error, external power grid input momentary power failure, external power grid dual-path power failure alarm, electrical characteristic over-limit, intelligent analysis fault diagnosis, external power grid three-phase power supply sequence error, and external power grid dual-path power failure alarm. Then, {external power grid input phase loss / power failure, external power grid three-phase power supply sequence error, external power grid input momentary power failure, external power grid dual-path power failure alarm, electrical characteristic over-limit, and intelligent analysis fault diagnosis} are classified into the first itemset, and {external power grid three-phase power supply sequence error and external power grid dual-path power failure alarm} are classified into the second itemset.

[0130] Example 2: On a certain day, the following alarms occur sequentially in the external power grid: external power grid input phase loss / power failure, external power grid three-phase power supply sequence error, external power grid input momentary power failure, external power grid dual-path power failure alarm, electrical characteristic over-limit, intelligent analysis fault diagnosis, external power grid three-phase power supply sequence error, and external power grid dual-path power failure alarm. Among them, the time interval between electrical characteristic over-limit and intelligent analysis fault diagnosis is greater than T0. Then, {external power grid input phase loss / power failure, external power grid three-phase power supply sequence error, external power grid input momentary power failure, external power grid dual-path power failure alarm, and electrical characteristic over-limit} is divided into the first itemset, and {intelligent analysis fault diagnosis, external power grid three-phase power supply sequence error, and external power grid dual-path power failure alarm} is divided into the second itemset.

[0131] S72 sets the minimum support and minimum confidence. In the Apriori algorithm, if an itemset is frequent (support greater than the minimum support), then all its subsets are also frequent. That is, if an itemset is infrequent, then all its supersets (the sets containing it) must be infrequent. Setting the minimum support can eliminate infrequent itemsets, improving the computation speed. The higher the final confidence, the higher the correlation between two different types of alarm data. In this embodiment, the minimum support is 30% and the minimum confidence is 50%.

[0132] S73 performs joins and pruning on itemsets to obtain frequent itemsets. The essence of the Apriori algorithm is to find the most frequent K itemsets in the dataset. It uses an iterative method, first searching for candidate 1-itemsets and their corresponding support, then pruning to remove 1-itemsets with support below the minimum, resulting in frequent 1-itemsets. Then, it joins the remaining frequent 1-itemsets to obtain candidate frequent 2-itemsets, and filters out candidate frequent 2-itemsets with support below the minimum, resulting in frequent 2-itemsets. This process continues iteratively until no more frequent k+1 itemsets can be found. The set of corresponding frequent k-itemsets is the output of the algorithm. In this embodiment, all alarm types are processed to remove infrequent itemsets, which can improve the computation speed.

[0133] S74 generates association rules from frequent itemsets and calculates the lift of strong association rules. Lift represents the ratio of the probability that X is also present given Y to the overall probability of X occurring. In other words, it is the ratio of the lift of X to Y to the overall probability of X occurring. If the lift is greater than 1, then X←Y is a valid strong association rule. That is, the confidence of the association rule {X}→{Y} = the support of {X, Y} / the support of {X}. For example, based on the data collected in the previous three months, the lift of {external grid electrical characteristic over-limit}→{external grid input power phase loss / power outage} between external grid electrical characteristic over-limit and external grid input power phase loss / power outage is calculated to be 1.5. This indicates that external grid electrical characteristic over-limit will promote external grid input power phase loss / power outage. However, this alarm only occurred when external grid electrical characteristic over-limit. This also reminds maintenance personnel to check and maintain related issues of external grid input power phase loss / power outage when dealing with external grid electrical characteristic over-limit.

[0134] Using Example 2, a more accurate trend of signal equipment degradation can be obtained, and the correlation between different faults can be extracted from the data to help maintenance personnel find the risks of signal equipment.

[0135] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for evaluating the quality of rail transit signaling equipment, characterized in that, Includes the following steps: S1. Evaluation indicators are set for each of the n rail transit signaling devices, wherein the evaluation indicators include a total score. Alarm type, and the deduction ratio corresponding to a single alarm type. and number of alarm signals ; where n is a positive integer; S2. Monitor the alarms of n rail transit signal devices respectively, and collect the alarm signals when a rail transit signal device issues an alarm signal; S3. Based on the collected alarm signals, evaluate the equipment quality of the corresponding rail transit signaling equipment; Specifically, step S3 includes the following sub-steps: S301. Collect alarm signals and extract evaluation indicators for the rail transit signaling equipment corresponding to the alarm signals; S302. Based on the evaluation indicators set for the corresponding rail transit signaling equipment, substitute them into the nonlinear quadrature algorithm formula to evaluate the corresponding rail transit signaling equipment and obtain the quality score of the corresponding rail transit signaling equipment. The alarm types include directly acquired alarm data types and indirectly acquired alarm data types. The directly acquired alarm data types include alarm information input from external data, and the indirectly acquired alarm data types include at least the equipment's track entry time, the number of times the turnout equipment is operated, the number of times the train is pressed, the number of times the track equipment is pressed, and the number of times the signal equipment is lit. In step S302, the specific calculation process for the score evaluation is as follows: ； Among them, the The quality score of rail transit signaling equipment is represented by the following: Let be a variable representing the number of alarm types.

2. The method for evaluating the quality of rail transit signaling equipment as described in claim 1, characterized in that, Step S3 also includes adding / subtracting points for the equipment quality evaluation, and the specific steps are as follows: Extract maintenance records of rail transit signaling equipment and add / deduct points based on the number of maintenance records.

3. The method for evaluating the quality of rail transit signaling equipment as described in claim 1, characterized in that, The total score The daily score is initialized to 100.

4. The method for evaluating the quality of rail transit signaling equipment as described in claim 1, characterized in that, The rail transit signaling equipment includes basic signaling equipment, intelligent equipment, and central equipment. When calculating the quality score of a single basic signaling device, a corresponding deduction standard is assigned to the on-track time of that basic signaling device.

5. The method for evaluating the quality of rail transit signaling equipment as described in claim 1, characterized in that, It also includes step S4, which specifically involves visualizing the alarm type, alarm frequency, and quality score.

6. The method for evaluating the quality of rail transit signaling equipment as described in claim 5, characterized in that, The visualization process includes generating statistical tables, bar charts, line charts, and pie charts for each rail transit signaling device.

7. The method for evaluating the quality of rail transit signaling equipment as described in claim 1, characterized in that, Collect the types, frequencies, and quality scores of alarms generated by each signaling device daily, and perform time-series autoregressive predictions on the quality scores of each rail transit signaling device in the future.

8. The method for evaluating the quality of rail transit signaling equipment as described in claim 7, characterized in that, Collect the specific times when alarms are generated by various rail transit signaling devices on the same day, and conduct correlation analysis on different alarm types of the same signaling device and alarm types of different signaling devices within a certain time period.

Images