Railway signal equipment fault prediction system based on edge computing

Abstract

The present application relates to the technical field of intelligent operation and maintenance of railway signal equipment, and particularly relates to a railway signal equipment fault prediction system based on edge computing, which comprises a data acquisition module, a basic degradation image module, an environmental stress feedforward module, a working condition drift correction module, a hot state optimization module and a fault warning execution module, to solve the problems of large communication delay and network failure of the existing cloud centralized architecture, inaccurate degradation benchmark caused by differences in equipment account of different devices, accumulation of resistance increase error into degradation index when working condition drift, misjudgment of temperature rise as insulation aging acceleration when hot state anomaly, and batch false alarm when environmental mutation, the present application realizes edge side autonomous availability of fault prediction, eliminates false alarm caused by individual differences of equipment and environmental mutation, and makes degradation index always reflect the real equipment health status.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology for railway signaling equipment, and in particular to a fault prediction system for railway signaling equipment based on edge computing. Background Technology

[0002] Existing railway signaling equipment fault prediction systems generally adopt a cloud-based centralized processing architecture, uploading field-collected data to a remote server for unified analysis via communication networks. This architecture suffers from problems such as large communication latency, high bandwidth consumption, and loss of prediction capabilities during network interruptions, making it difficult to meet the stringent real-time and reliability requirements of railway signaling equipment. Furthermore, existing prediction models often employ a uniform degradation threshold strategy, failing to consider the dispersion of baseline degradation caused by differences in service life, cumulative number of operations, and friction coefficients among different equipment. This leads to missed detections for new equipment and false detections for older equipment. Even systems that consider degradation still lack the ability to correct for operational drift and thermal anomalies online. They directly accumulate increased resistance due to insufficient lubrication into the degradation index and misjudge abnormal motor temperature rise as accelerated insulation aging, resulting in a serious waste of maintenance resources during track maintenance windows.

[0003] Chinese patent document with publication number CN121615023A discloses a method for predicting and diagnosing faults in railway signaling equipment, but this solution still leads to inaccuracy of references for different equipment and distortion of degradation index when operating conditions drift.

[0004] Chinese patent document with publication number CN120408330A discloses a method and system for predicting faults in rail transit equipment. However, this solution still adopts a centralized processing strategy based on time-series prediction models and classification models instead of dynamically generating degradation indices based on ledger parameters, does not perform online calibration of real-time resistance drift, and does not establish a feedforward compensation relationship between weather forecasts and degradation indices, resulting in inaccurate benchmarks when there are differences in ledgers and batch false alarms when there are sudden environmental changes. Summary of the Invention

[0005] To address these issues, this invention provides a railway signaling equipment fault prediction system based on edge computing. This system overcomes the problems in existing technologies, such as large communication delays and network outages in cloud-based centralized architectures, inaccurate degradation benchmarks due to differences in records between different devices, erroneous accumulation of degradation index due to increased resistance during operating condition drift, misjudgment of temperature rise as accelerated insulation aging during thermal anomalies, and batch false alarms during sudden environmental changes.

[0006] To achieve the above objectives, the present invention provides a railway signaling equipment fault prediction system based on edge computing, comprising: The data acquisition module is used to collect equipment parameter data, which includes: equipment ledger parameters, real-time operating current and real-time motor temperature rise; The basic degradation profile module is used to generate degradation indices based on equipment ledger parameters; The operating condition drift calibration module is used to calculate the current deviation based on the real-time operating current and to calibrate the degradation index based on the current deviation. The hot optimization module is used to optimize the degradation index based on the real-time motor temperature rise. The environmental stress feedforward module is used to acquire meteorological forecast data, calculate the environmental degradation coefficient based on the meteorological forecast data, and send the environmental degradation coefficient to the basic degradation profile module to perform feedforward compensation for the generation process of the degradation index. The fault warning execution module is used to calculate the fault warning probability based on the degradation index and output graded warnings according to the fault warning probability.

[0007] Preferably, the basic degradation profile module generates a degradation index based on equipment ledger parameters, specifically as follows: The service time t is obtained, and the equipment ledger parameters include the normalized value N of the cumulative number of operations and the normalized value M of the nominal friction coefficient. The degradation index is calculated based on the service time t, the normalized value N of the cumulative number of operations, and the normalized value M of the nominal friction coefficient. Perform calculations and set Where Tr is the reference lifetime, B is the time shape parameter, λ1 is the time degradation weight, λ2 is the cyclic degradation weight, λ3 is the frictional degradation weight, n is the cyclic shape parameter, and δ0 is the baseline degradation offset.

[0008] Preferably, the environmental stress feedforward module acquires meteorological forecast data, calculates the environmental degradation coefficient based on the meteorological forecast data, and also sends the environmental degradation coefficient to the basic degradation profile module for feedforward compensation in the degradation index generation process, specifically: The predicted ambient temperature Tf and predicted precipitation Rf are obtained, and the environmental degradation coefficient Ke is calculated based on these parameters. Where Tref is the reference temperature, w1 is the temperature stress weight, and w2 is the precipitation stress weight; The environmental degradation coefficient Ke is sent to the basic degradation profile module for analysis of the degradation index. Feedforward compensation is performed to obtain the environmental compensation degradation index Dbe, and then a setting is made. and the degradation index The value is replaced with the value of the environmental compensation degradation index Dbe.

[0009] Preferably, the operating condition drift correction module calibrates the degradation index based on the real-time operating current and the current deviation, specifically as follows: The current deviation ΔI is calculated based on the real-time effective value of the operating current Ir and the nominal effective value of the operating current I0, and ΔI is set as (Ir-I0) / I0; The absolute value of the current deviation ΔI is compared with a preset first current deviation threshold I1 and a preset second current deviation threshold I2, where I1 < I2. Based on the comparison result, the resistance drift state is judged, and the degradation index is adjusted accordingly. Proofreading is performed, including a first proofreading and a second proofreading.

[0010] Preferably, the drag drift state is determined based on the comparison results, and the degradation index is determined based on the determination results. Proofreading, including: When |ΔI|≤I1, the operating condition drift correction module determines the resistance drift state as slight drift and does not affect the degradation index. Proofreading; When I1 < |ΔI| ≤ I2, the operating condition drift correction module determines the resistance drift state as moderate drift, and the degradation index... Perform the first proofreading; When |ΔI|>I2, the operating condition drift correction module determines that the resistance drift state is severe drift, and the degradation index is affected. Perform a second proofreading.

[0011] Preferably, the first calibration includes: adjusting the degradation index according to the drift compensation coefficient c1. Degradation compensation is performed to obtain the calibrated degradation index Dc, and then the following settings are made. and the degradation index Replace the value with the value of the calibrated degradation index Dc; The second verification includes: calculating the estimated friction coefficient Mest, setting Mest=M0×(1+β×ΔI), where M0 is the nominal friction coefficient and β is the friction estimation coefficient, and feeding the estimated friction coefficient Mest as the normalized value M of the nominal friction coefficient back to the basic degradation profile module to regenerate the degradation index.

[0012] Preferably, the thermal optimization module optimizes the degradation index based on the real-time motor temperature rise, specifically as follows: The temperature rise deviation rate Et is calculated based on the real-time motor temperature rise Tm and the designed motor temperature rise Td, and Et is set as (Tm-Td) / Td; The temperature rise deviation rate Et is compared with the preset temperature rise deviation rate threshold E1. The hot state deviation is judged based on the comparison result, and the post-calibration degradation index Dc is optimized based on the judgment result.

[0013] Preferably, the thermal deviation state is judged based on the comparison results, and the post-calibration degradation index Dc is optimized based on the judgment results, including: When |Et|≤E1, the hot-state optimization module determines that the hot-state deviation is slight and does not perform optimization on the post-calibration degradation index Dc. When |Et|>E1, the hot-state optimization module determines that the hot-state deviation is excessive and optimizes the post-calibration degradation index Dc.

[0014] Preferably, the optimization includes: optimizing and correcting the calibration degradation index Dc based on the temperature rise optimization gain coefficient g to obtain the optimized degradation index Do, setting Do = Dc × (1 + g × Et), and adjusting the degradation index in the basic degradation profile module. The value of is replaced with the value of the degradation index Do after optimization.

[0015] Preferably, the fault warning execution module calculates the fault warning probability based on the degradation index and outputs graded warnings according to the fault warning probability, specifically as follows: According to the degradation index The probability Pf of fault warning is calculated and set. ,in, The degradation threshold index, For Weibull shape parameters, is the base of the natural logarithm; The fault warning probability Pf is compared with preset first probability threshold P1, preset second probability threshold P2, and preset third probability threshold P3, where P1 < P2 < P3. Based on the comparison results, the warning level is determined and a handling strategy is output, wherein: When Pf < P1, the fault early warning execution module determines it to be at the green level and outputs a continuous monitoring command; When P1≤Pf<P2, the fault early warning execution module determines it to be at the yellow level and outputs an enhanced inspection command; When P2≤Pf<P3, the fault warning execution module determines it to be at the orange level and outputs a sunroof repair instruction. When Pf≥P3, the fault warning execution module determines it to be at the red level and outputs an immediate blocking command.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: the system generates a degradation index to establish a quantitative correlation between the inherent attributes of the equipment and the degradation benchmark, eliminating benchmark inaccuracies caused by differences in records between different equipment; the system also uses feedforward compensation to enable the degradation benchmark to adapt to environmental changes in advance, eliminating batch false alarms caused by cold waves and rainfall; the system also calculates current deviation and calibrates the degradation index to achieve adaptive identification and compensation of resistance drift, eliminating the impact of operating condition drift on the degradation index; the system also optimizes the degradation index to achieve automatic optimization and closed-loop correction when there is thermal deviation, ensuring that the degradation index always reflects the true health status of the equipment; and the system also calculates the probability of fault warning and outputs graded warnings to achieve direct availability of fault prediction. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the fault prediction system for railway signaling equipment based on edge computing in this embodiment; Figure 2 This is a schematic diagram illustrating the logic of the working condition drift correction module in this embodiment correcting the degradation index. Figure 3 This is a schematic diagram illustrating the logic of the hot-state optimization module in this embodiment optimizing the degradation index. Figure 4 This is a logical diagram illustrating how the fault warning execution module in this embodiment outputs graded warnings based on the fault warning probability. Detailed Implementation

[0018] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0019] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0020] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0021] Please see Figure 1As shown, this is a schematic diagram of the fault prediction system for railway signaling equipment based on edge computing in this embodiment. The system includes: The data acquisition module is used to collect equipment parameter data, which includes: equipment ledger parameters, real-time operating current and real-time motor temperature rise; The basic degradation profile module is used to generate a degradation index based on equipment ledger parameters. The basic degradation profile module is connected to the data acquisition module. An environmental stress feedforward module is used to acquire meteorological forecast data, calculate the environmental degradation coefficient based on the meteorological forecast data, and send the environmental degradation coefficient to the basic degradation profile module to perform feedforward compensation on the generation process of the degradation index. The environmental stress feedforward module is connected to the basic degradation profile module. The operating condition drift correction module is used to calculate the current deviation based on the real-time operating current and to correct the degradation index based on the current deviation. The operating condition drift correction module is connected to the basic degradation profile module. A hot-state optimization module is used to optimize the degradation index based on the real-time motor temperature rise. The hot-state optimization module is connected to the basic degradation profile module and the operating condition drift correction module. The fault warning execution module is used to calculate the fault warning probability based on the degradation index and output graded warnings based on the fault warning probability. The fault warning execution module is connected to the basic degradation profile module.

[0022] Specifically, the edge computing-based railway signaling equipment fault prediction system is applied to the fault prediction system of the ZD6 electric switch machine equipped with edge computing nodes, Hall current sensors, and temperature sensors. The system generates a degradation index to establish a quantitative correlation between the inherent attributes of the equipment and the degradation benchmark, eliminating benchmark inaccuracies caused by differences in records between different equipment. Furthermore, through feedforward compensation, the system allows the degradation benchmark to adapt to environmental changes in advance, eliminating batch false alarms caused by cold waves and rainfall. The system also calculates current deviations and calibrates the degradation index to achieve adaptive identification and compensation of resistance drift, eliminating the impact of operating condition drift on the degradation index. Additionally, the system optimizes the degradation index to achieve automatic optimization and closed-loop correction when there is thermal deviation, ensuring that the degradation index always reflects the true health status of the equipment. Finally, the system calculates the fault warning probability and outputs graded warnings, achieving direct usability of fault prediction.

[0023] Specifically, the data acquisition module collects equipment parameter data, including: equipment ledger parameters, real-time operating current, and real-time motor temperature rise. The equipment ledger parameters include the normalized value of the cumulative number of operations and the normalized value of the nominal friction coefficient. The data acquisition module collects the equipment ledger parameters through the equipment management information system. In this embodiment, the normalized value of the cumulative number of operations N=0.25 and the normalized value of the nominal friction coefficient M=1.17 are obtained by consulting the equipment's factory calibration record. The real-time operating current refers to the effective value of the motor winding current during a single operation of the switch machine. The data acquisition module collects the real-time operating current through a Hall current sensor. In this embodiment, the Hall current sensor installed in the power supply circuit of the switch machine motor collects the time-domain signal of the operating current and calculates the effective value Ir of the real-time operating current through the root mean square. The real-time motor temperature rise refers to the difference between the motor housing temperature and the ambient temperature after the switch machine operation is completed. The data acquisition module collects the real-time motor temperature rise through a temperature sensor. In this embodiment, the temperature sensor installed at the motor housing collects the motor housing temperature and calculates the real-time motor temperature rise Tm in combination with the ambient temperature.

[0024] Specifically, the basic degradation profile module generates a degradation index based on equipment ledger parameters, as follows: The service time t is obtained, and the equipment ledger parameters include the normalized value N of the cumulative number of operations and the normalized value M of the nominal friction coefficient. The degradation index is calculated based on the service time t, the normalized value N of the cumulative number of operations, and the normalized value M of the nominal friction coefficient. Perform calculations and set Where Tr is the reference lifetime, B is the time shape parameter, λ1 is the time degradation weight, λ2 is the cyclic degradation weight, λ3 is the frictional degradation weight, n is the cyclic shape parameter, and δ0 is the baseline degradation offset.

[0025] Specifically, the service time refers to the cumulative operating time of the switch machine since its installation and commissioning. In this embodiment, the service time is obtained through the equipment management information system. This embodiment sets t=2.5 years, corresponding to the equipment being in a medium service period. The normalized value of the cumulative number of actions refers to the ratio of the cumulative number of conversion actions to the standard action life, with a value of N=0.25. The reason for this value is that it is calculated using N=Nc / Ndesign, where Nc is the cumulative number of actions (5000). 0 times, Ndesign represents the design life of 200,000 cycles. When N=0.25, it indicates that the mechanical components of the equipment have consumed approximately 25% of their operating life, indicating the early stage of wear. The normalized nominal friction coefficient is the ratio of the factory-calibrated friction coefficient to the reference friction coefficient, with a value of M=1.17. The rationale for this normalized nominal friction coefficient value is that it is calculated using M=M0 / Mref, where M0 is the nominal friction coefficient of 0.35 and Mref is the reference friction coefficient of 0.30. When M= At 1.17, it was reported that the factory adjustment of the friction coupler of the equipment was slightly higher than the baseline state. The reference life refers to the baseline life used for normalizing service time, with a value of Tr = 1 year. The reason for this reference life value is that it is determined through system calibration experiments, providing a baseline scale for normalizing service time. When Tr = 1 year, the service time is directly involved in the power law calculation in years, simplifying the edge node calculation. The time shape parameter refers to the power law exponent between service time and degradation index, with a value of B = 1.2. The rationale for this value is as follows: This value was determined through least-squares fitting of fault data from similar equipment. When B=1.2, the degradation index exhibits a superlinear growth with service time, consistent with the bathtub curve wear period characteristics of mechanical wear equipment. The time degradation weight refers to the change in the degradation index caused by a unit normalized service time change, and its value is λ1=0.05. The rationale for this time degradation weight is as follows: This value was determined through degradation calibration experiments during the system debugging phase. When λ1=0.05, the degradation index increases by 1.0 for every 1.0 increase in normalized service time. This reflects the power-law contribution of time aging to degradation. The cyclic degradation weight refers to the change in the degradation index caused by a change in the number of normalized actions, and its value is λ2=0.02. The reason for choosing this value is that it was determined through degradation calibration experiments during the system debugging phase. When λ2=0.02, the degradation index increases by 1.0 for every 1.0 increase in the number of normalized actions. The friction degradation weight, which reflects the power-law contribution of mechanical cyclic fatigue to degradation, refers to the change in the degradation index caused by a change in the normalized value of the nominal friction coefficient. Its value is λ3 = 0.03. The rationale for this value is that it was determined through degradation calibration experiments during the system debugging phase. When λ3 = 0.03, for every 1.0 increase in the normalized friction coefficient, the degradation index increases by 0.03, reflecting the linear contribution of the friction state to mechanical degradation. The cyclic shape parameter refers to the power-law exponent between the normalized number of actions and the degradation index, with a value of n = 0.8. The rationale for this value is as follows: This value was determined through a power-law fitting experiment during the system debugging phase. When n=0.8, the degradation index has a sublinear relationship with the number of cycles, which is consistent with the running-in characteristics in the early stage of wear. The reference degradation offset refers to the theoretical degradation reference value when the service time, number of operations, and friction coefficient are all zero, and the value is δ0=0.01. The rationale for this reference degradation offset is as follows: This value was determined through a calibration experiment during the system debugging phase. It reflects the reference degradation caused by the inherent design margin of the equipment and unexplained random factors. Even if all the ledger parameters are zero, the system still has a minimum degradation index of 0.01.

[0026] Specifically, the environmental stress feedforward module acquires meteorological forecast data, calculates the environmental degradation coefficient based on the meteorological forecast data, and also sends the environmental degradation coefficient to the basic degradation profile module for feedforward compensation in the degradation index generation process. Specifically: The forecast ambient temperature Tf and forecast precipitation Rf are obtained, and the environmental degradation coefficient Ke is calculated based on the above parameters. Ke is set as w1×(Tref-Tf) / Tref+w2×Rf, where Tref is the reference temperature, w1 is the temperature stress weight, and w2 is the precipitation stress weight. The environmental degradation coefficient Ke is sent to the basic degradation profile module for analysis of the degradation index. Feedforward compensation is performed to obtain the environmental compensation degradation index Dbe, and then a setting is made. and the degradation index The value is replaced with the value of the environmental compensation degradation index Dbe.

[0027] Specifically, the forecast ambient temperature refers to the predicted ambient temperature of the section where the equipment is located within the next 72 hours. In this embodiment, the forecast ambient temperature is obtained by accessing the railway meteorological service interface. The reference temperature refers to the baseline temperature under standard environmental conditions, with a value of Tref=20℃. The reason for choosing the reference temperature is that this value is the standard design ambient temperature of railway signaling equipment, reflecting the thermal balance benchmark point of the switch machine under standard lubrication conditions. The forecast precipitation refers to the cumulative precipitation forecast value within the next 72 hours. In this embodiment, the forecast precipitation is obtained by accessing the railway meteorological service interface. The temperature stress weight refers to the environmental degradation coefficient caused by a unit temperature deviation. The change in temperature stress weight is set at w1=0.04. The reason for this value is that it was determined through environmental stress calibration experiments. When w1=0.04, a temperature drop from 20℃ to -20℃ can increase the environmental degradation coefficient by 1.6, reflecting the dominant influence of low temperature on the viscosity of lubricating grease. The change in precipitation stress weight is set at w2=0.005 / mm. The reason for this value is that it was determined through environmental stress calibration experiments. When w2=0.005 / mm, an increase of 10mm in precipitation can increase the environmental degradation coefficient by 0.05, reflecting the secondary influence of rainfall on the lubrication state of the slide plate.

[0028] Please see Figure 2 As shown, this is a logic diagram of the operating condition drift calibration module calibrating the degradation index. The operating condition drift calibration module calibrates the degradation index based on the real-time operating current deviation, specifically as follows: The current deviation ΔI is calculated based on the real-time effective value of the operating current Ir and the nominal effective value of the operating current I0, and ΔI is set as (Ir-I0) / I0; The absolute value of the current deviation ΔI is compared with a preset first current deviation threshold I1 and a preset second current deviation threshold I2, where I1 < I2. Based on the comparison result, the resistance drift state is judged, and the degradation index is adjusted accordingly. Proofreading is performed, which includes a first proofreading and a second proofreading, wherein: When |ΔI|≤I1, the operating condition drift correction module determines the resistance drift state as slight drift and does not affect the degradation index. Proofreading; When I1 < |ΔI| ≤ I2, the operating condition drift correction module determines the resistance drift state as moderate drift, and the degradation index... Perform the first proofreading; When |ΔI|>I2, the operating condition drift correction module determines that the resistance drift state is severe drift, and the degradation index is affected. Perform a second proofreading; The first calibration includes: adjusting the degradation index according to the drift compensation coefficient c1. Degradation compensation is performed to obtain the calibrated degradation index Dc, and then the following settings are made. and the degradation index Replace the value with the value of the calibrated degradation index Dc; The second verification includes: calculating the estimated friction coefficient Mest, setting Mest=M0×(1+β×ΔI), where M0 is the nominal friction coefficient and β is the friction estimation coefficient, and feeding the estimated friction coefficient Mest as the normalized value M of the nominal friction coefficient back to the basic degradation profile module to regenerate the degradation index.

[0029] Specifically, the nominal effective value of the operating current refers to the benchmark effective value of the operating current of the switch machine under standard operating conditions, with a value of I0 = 1.84A. The rationale for this nominal effective value is as follows: This value is determined through the factory electrical calibration test of the switch machine. Under standard switching resistance and rated voltage conditions, the current of 100 operating processes is collected and the effective value is calculated. The statistical average value is 1.84A, reflecting the benchmark current level of the switch machine in a healthy state. The preset first current deviation threshold is the boundary value for judging the resistance drift state as slight drift, with a value of I1 = 0.10. The rationale for this preset first current deviation threshold is as follows: This value is determined through current deviation tests. When the current deviation does not exceed 10%, the switching resistance of the switch machine is within the normal fluctuation range, and the environmental compensation degradation index can be used directly, avoiding system oscillations caused by over-calibration. The preset second current deviation threshold is the boundary value for judging the resistance drift state as moderate drift, with a value of I2 = 0.25. The rationale for this preset second current deviation threshold is as follows: This value is determined through current deviation tests. It is determined that when the current deviation exceeds 25%, the switching resistance of the switch machine deviates significantly from the design value, requiring degradation compensation or friction coefficient reconstruction to restore prediction accuracy. The resistance drift state refers to the classification of the degree of deviation of the turnout resistance from the design value based on the current deviation. The resistance drift state includes slight drift, moderate drift, and severe drift. The drift compensation coefficient is the gain coefficient when performing degradation compensation for resistance drift, with a value of c1=0.30. The reason for the value of the drift compensation coefficient is that this value was determined through drift compensation experiments. When c1=0.30, every 10% increase in current deviation can increase the post-calibration degradation index by 3%, achieving a balance between compensation effect and stability. The friction estimation coefficient is the proportionality coefficient between the current deviation and the estimated friction coefficient, with a value of β=0.20. The reason for the value of the friction estimation coefficient is that this value was determined through friction coefficient estimation experiments. When the current deviation increases by 1.0, the estimated friction coefficient increases by 20%, reflecting the linear estimation relationship between current deviation and friction coefficient drift, and is used for friction coefficient reconstruction in cases of severe drift.

[0030] Please see Figure 3 As shown, this is a logical diagram illustrating how the hot-state optimization module optimizes the degradation index in this embodiment. Specifically, the hot-state optimization module optimizes the degradation index based on the real-time motor temperature rise. The temperature rise deviation rate Et is calculated based on the real-time motor temperature rise Tm and the designed motor temperature rise Td, and Et is set as (Tm-Td) / Td; The temperature rise deviation rate Et is compared with the preset temperature rise deviation rate threshold E1. Based on the comparison result, the hot-state deviation is judged, and the post-calibration degradation index Dc is optimized based on the judgment result. When |Et|≤E1, the hot-state optimization module determines that the hot-state deviation is slight and does not perform optimization on the post-calibration degradation index Dc. When |Et|>E1, the hot-state optimization module determines that the hot-state deviation is excessive and optimizes the post-calibration degradation index Dc. The optimization includes: optimizing and correcting the calibration degradation index Dc based on the temperature rise optimization gain coefficient g to obtain the optimized degradation index Do, setting Do = Dc × (1 + g × Et), and adjusting the degradation index in the basic degradation profile module. The value of is replaced with the value of the degradation index Do after optimization.

[0031] Specifically, the real-time motor temperature rise refers to the difference between the motor housing temperature and the ambient temperature after the switch machine operation is completed. In this embodiment, the real-time motor temperature rise is obtained through a temperature sensor. The process of obtaining the real-time motor temperature rise is as follows: the motor housing temperature Tmt is collected by a temperature sensor installed on the motor housing, and the ambient temperature Tamb is collected by a temperature sensor installed on the equipment housing. The real-time motor temperature rise Tm is calculated based on the motor housing temperature Tmt and the ambient temperature Tamb, and Tm = Tmt - Tamb is set. The design motor temperature rise refers to the target value of the switch machine motor housing temperature rise under rated design conditions. In this embodiment, the design motor temperature rise is obtained through the technical manual provided by the switch machine manufacturer. The rationale for the value of the design motor temperature rise is as follows: this value is the design thermal index of the ZD6 type switch machine under standard ambient temperature of 20℃, rated load, and rated operating frequency, and the value is taken as Td = 45K. This temperature rise value reflects the Class B insulation of the motor under standard operating conditions. The design incorporates thermal reserves. When the actual temperature rise is lower than this value, it indicates that the motor's thermal state is normal. The preset temperature rise deviation rate is the boundary value for determining a slight deviation in the thermal state, with a value of E1=0.15. The rationale for this preset temperature rise deviation rate is that it is determined through temperature rise deviation experiments. When the temperature rise deviation rate does not exceed 15%, the motor's thermal state is within the normal fluctuation range, and the degradation index can be directly used after calibration, avoiding system oscillations caused by over-optimization. The thermal state deviation refers to the classification of the degree of deviation of the motor's thermal state from the design value based on the temperature rise deviation rate. The thermal state deviation includes slight deviation and excessive deviation. The temperature rise optimization gain coefficient is the gain coefficient used to optimize and correct the degradation index when the thermal state deviates excessively, with a value of g=0.25. The rationale for this temperature rise optimization gain coefficient is that it is determined through thermal state optimization experiments. When the thermal state deviates excessively, the degradation index is increased by 25%, which can advance the fault probability calculation to a higher warning range by about 10%, effectively suppressing the risk of missed alarms caused by sudden changes in the thermal state.

[0032] Please see Figure 4 As shown, this is a logical diagram of the fault warning execution module in this embodiment performing graded warning output based on the fault warning probability. The fault warning execution module calculates the fault warning probability based on the degradation index and performs graded warning output based on the fault warning probability, specifically as follows: According to the degradation index The probability Pf of fault warning is calculated and set. ,in, The degradation threshold index, For Weibull shape parameters, is the base of the natural logarithm; The fault warning probability Pf is compared with preset first probability threshold P1, preset second probability threshold P2, and preset third probability threshold P3, where P1 < P2 < P3. Based on the comparison results, the warning level is determined and a handling strategy is output, wherein: When Pf < P1, the fault early warning execution module determines it to be at the green level and outputs a continuous monitoring command; When P1≤Pf<P2, the fault early warning execution module determines it to be at the yellow level and outputs an enhanced inspection command; When P2≤Pf<P3, the fault warning execution module determines it to be at the orange level and outputs a sunroof repair instruction. When Pf≥P3, the fault warning execution module determines it to be at the red level and outputs an immediate blocking command.

[0033] Specifically, the degradation threshold index refers to the critical value at which the probability of a fault warning reaches 63.2%, with a value of Dth = 0.25. The rationale for this value is as follows: This value is determined through statistical analysis of historical fault data. A degradation index sequence of the 72 hours preceding a fault in a ZD6 type switch machine over the past 5 years was collected, and the Db distribution at the time of fault occurrence was statistically analyzed. The 63.2% quantile is 0.25. When Db ≥ Dth, the fault probability increases rapidly according to the Weibull model. The Weibull shape parameter refers to the fault warning probability. The shape control parameter, which varies with the degradation index, is set to m=2.0. The reason for this Weibull shape parameter value is that it was determined through reliability engineering experiments. When m=2.0, the failure probability curve exhibits a linear increasing characteristic, consistent with the cumulative degradation law of mechanical wear failures. The preset first probability threshold is the boundary value for determining the green and yellow levels, set to P1=0.15. The reason for this preset first probability threshold value is that it was determined through historical failure data statistics. When Pf<0.15, the equipment is expected to fail within the next 72 hours. The probability is less than 15%, which is within the normal fluctuation range and requires no intervention. The preset second probability threshold is the dividing value between the yellow and orange levels, with a value of P2=0.45. The reason for this preset second probability threshold is that this value is determined through historical fault data statistics. When Pf≥0.45, the probability of the equipment failing within the next 72 hours exceeds 45%, requiring scheduled maintenance. The preset third probability threshold is the dividing value between the orange and red levels, with a value of P3=0.75. The reason for this preset third probability threshold is that... This value is determined through historical fault data statistics. When Pf≥0.75, the probability of equipment failure within the next 72 hours exceeds 75%, and the risk of continued operation is extremely high. The line must be immediately blocked and workers dispatched for emergency repairs. The continuous monitoring instruction refers to the instruction to maintain the existing monitoring frequency without special handling. The enhanced inspection instruction refers to the instruction to increase the monitoring frequency and arrange manual on-site inspections. The inclusion in the track maintenance window instruction refers to the instruction to include the equipment in the planned next track maintenance operation. The immediate blocking instruction refers to the instruction to send a command to the interlocking system to block the turnout section.

[0034] Specifically, this embodiment does not limit the specific communication protocol of the output processing strategy. Those skilled in the art can set it according to the actual situation, as long as it meets the output requirements of the hierarchical processing strategy. For example, it can be output through the railway signal safety communication protocol RSSP-II.

[0035] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A fault prediction system for railway signaling equipment based on edge computing, characterized in that, include: The data acquisition module is used to collect equipment parameter data, which includes: equipment ledger parameters, real-time operating current and real-time motor temperature rise; The basic degradation profile module is used to generate degradation indices based on equipment ledger parameters; The operating condition drift calibration module is used to calculate the current deviation based on the real-time operating current and to calibrate the degradation index based on the current deviation. The hot optimization module is used to optimize the degradation index based on the real-time motor temperature rise. The environmental stress feedforward module is used to acquire meteorological forecast data, calculate the environmental degradation coefficient based on the meteorological forecast data, and send the environmental degradation coefficient to the basic degradation profile module to perform feedforward compensation for the generation process of the degradation index. The fault warning execution module is used to calculate the fault warning probability based on the degradation index and output graded warnings according to the fault warning probability.

2. The railway signaling equipment fault prediction system based on edge computing according to claim 1, characterized in that, The basic degradation profile module generates a degradation index based on equipment ledger parameters, specifically: The service time t is obtained, and the equipment ledger parameters include the normalized value N of the cumulative number of operations and the normalized value M of the nominal friction coefficient. The degradation index is calculated based on the service time t, the normalized value N of the cumulative number of operations, and the normalized value M of the nominal friction coefficient. Perform calculations and set Where Tr is the reference lifetime, B is the time shape parameter, λ1 is the time degradation weight, λ2 is the cyclic degradation weight, λ3 is the frictional degradation weight, n is the cyclic shape parameter, and δ0 is the baseline degradation offset.

3. The railway signaling equipment fault prediction system based on edge computing according to claim 2, characterized in that, The environmental stress feedforward module acquires meteorological forecast data and calculates the environmental degradation coefficient based on the meteorological forecast data. It also sends the environmental degradation coefficient to the basic degradation profile module for feedforward compensation during the generation process of the degradation index. Specifically: The predicted ambient temperature Tf and predicted precipitation Rf are obtained, and the environmental degradation coefficient Ke is calculated based on these parameters. Where Tref is the reference temperature, w1 is the temperature stress weight, and w2 is the precipitation stress weight; The environmental degradation coefficient Ke is sent to the basic degradation profile module for analysis of the degradation index. Feedforward compensation is performed to obtain the environmental compensation degradation index Dbe, and then a setting is made. and the degradation index The value is replaced with the value of the environmental compensation degradation index Dbe.

4. The railway signaling equipment fault prediction system based on edge computing according to claim 3, characterized in that, The operating condition drift correction module calibrates the degradation index based on the real-time operating current and the current deviation, specifically as follows: The current deviation ΔI is calculated based on the real-time effective value of the operating current Ir and the nominal effective value of the operating current I0, and ΔI is set as (Ir-I0) / I0; The absolute value of the current deviation ΔI is compared with a preset first current deviation threshold I1 and a preset second current deviation threshold I2, where I1 < I2. Based on the comparison result, the resistance drift state is judged, and the degradation index is adjusted accordingly. Proofreading is performed, including a first proofreading and a second proofreading.

5. The railway signaling equipment fault prediction system based on edge computing according to claim 4, characterized in that, The drag drift state is judged based on the comparison results, and the degradation index Db is calibrated based on the judgment results, including: When |ΔI|≤I1, the operating condition drift correction module determines the resistance drift state as slight drift and does not affect the degradation index. Proofreading; When I1 < |ΔI| ≤ I2, the operating condition drift correction module determines the resistance drift state as moderate drift, and the degradation index... Perform the first proofreading; When |ΔI|>I2, the operating condition drift correction module determines that the resistance drift state is severe drift, and the degradation index is affected. Perform a second proofreading.

6. The railway signaling equipment fault prediction system based on edge computing according to claim 4, characterized in that, The first calibration includes: adjusting the degradation index according to the drift compensation coefficient c1. Degradation compensation is performed to obtain the calibrated degradation index Dc, and then the following settings are made. and the degradation index Replace the value with the value of the calibrated degradation index Dc; The second verification includes: calculating the estimated friction coefficient Mest, setting Mest=M0×(1+β×ΔI), where M0 is the nominal friction coefficient and β is the friction estimation coefficient, and feeding the estimated friction coefficient Mest as the normalized value M of the nominal friction coefficient back to the basic degradation profile module to regenerate the degradation index.

7. The railway signaling equipment fault prediction system based on edge computing according to claim 4, characterized in that, The thermal optimization module optimizes the degradation index based on the real-time motor temperature rise, specifically as follows: The temperature rise deviation rate Et is calculated based on the real-time motor temperature rise Tm and the designed motor temperature rise Td, and Et is set as (Tm-Td) / Td; The temperature rise deviation rate Et is compared with the preset temperature rise deviation rate threshold E1. The hot state deviation is judged based on the comparison result, and the post-calibration degradation index Dc is optimized based on the judgment result.

8. The railway signaling equipment fault prediction system based on edge computing according to claim 7, characterized in that, Based on the comparison results, the thermal deviation state is judged, and the post-calibration degradation index Dc is optimized based on the judgment results, including: When |Et|≤E1, the hot-state optimization module determines that the hot-state deviation is slight and does not perform optimization on the post-calibration degradation index Dc. When |Et|>E1, the hot-state optimization module determines that the hot-state deviation is excessive and optimizes the post-calibration degradation index Dc.

9. The railway signaling equipment fault prediction system based on edge computing according to claim 7, characterized in that, The optimization includes: optimizing and correcting the calibration degradation index Dc based on the temperature rise optimization gain coefficient g to obtain the optimized degradation index Do, setting Do = Dc × (1 + g × Et), and adjusting the degradation index in the basic degradation profile module. The value of is replaced with the value of the degradation index Do after optimization.

10. The railway signaling equipment fault prediction system based on edge computing according to claim 2, characterized in that, The fault warning execution module calculates the fault warning probability based on the degradation index and outputs tiered warnings according to the fault warning probability, specifically: According to the degradation index The probability Pf of fault warning is calculated and set. ,in, The degradation threshold index, For Weibull shape parameters, is the base of the natural logarithm; The fault warning probability Pf is compared with preset first probability threshold P1, preset second probability threshold P2, and preset third probability threshold P3, where P1 < P2 < P3. Based on the comparison results, the warning level is determined and a handling strategy is output, wherein: When Pf < P1, the fault early warning execution module determines it to be at the green level and outputs a continuous monitoring command; When P1≤Pf<P2, the fault early warning execution module determines it to be at the yellow level and outputs an enhanced inspection command; When P2≤Pf<P3, the fault warning execution module determines it to be at the orange level and outputs a sunroof repair instruction. When Pf≥P3, the fault warning execution module determines it to be at the red level and outputs an immediate blocking command.

Images