Health degree evaluation method and system for turnout switch machine

Abstract

The invention relates to a health degree evaluation method for a turnout switch machine. The method specifically comprises the following steps: S1, respectively collecting historical health data and real-time monitoring data of the turnout switch machine; s2, constructing a health degree evaluation function by adopting normal distribution according to the historical health data; and S3, dividing different stages according to different preset time, and performing stage scoring according to the real-time monitoring data and the health degree evaluation function. According to the method, the self-adaptive health degree evaluation function is constructed, and evaluation is carried out in combination with the short-term stage, the middle-term stage and the long-term stage, so that the current equipment operation condition can be evaluated, and early warning can be carried out on the potential fault trend.

Description

Technical Field

[0001] The present invention relates to the technical field of health assessment of switch machines, and particularly to a health assessment method and system for switch machines. Background Art

[0002] As an indispensable part of the rail transit network, the normal and healthy operation of the switch machine is directly related to the driving safety, efficiency of the train, and the smooth operation of the entire urban rail network. The main responsibility of the switch machine is to ensure that the train can safely and accurately transfer from one track to another, completing complex track switching tasks, so as to ensure the efficient and orderly operation of the train in the busy urban rail transit network.

[0003] The invention patent with the publication number CN116205538A discloses a health assessment method for rail transit switch equipment, including the following steps: A. Collect switch equipment data, and generate health assessment source data for the switch equipment after processing the collected data; B. Establish a health assessment model for the switch equipment, input the health assessment source data for the switch equipment into the health assessment model for the switch equipment and generate a health score for the switch equipment; the data types used in this patent are numerous and the calculations are cumbersome, and it is impossible to conduct phased evaluations according to time.

[0004] Therefore, it is an urgent need to provide a health assessment method that uses fewer data types and can conduct phased evaluations. Summary of the Invention

[0005] The purpose of the present invention is to provide a health assessment method and system for switch machines to overcome the defects of the above-mentioned existing technologies.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] According to one aspect of the present invention, a health assessment method for a switch machine is provided, and the method specifically includes:

[0008] S1. Respectively collect the historical health data and real-time monitoring data of the switch machine;

[0009] S2. Construct a health assessment function according to the historical health data and using the normal distribution;

[0010] S3. Divide different stages according to different preset times, and conduct stage scoring according to the real-time monitoring data and the health assessment function.

[0011] As a preferred technical solution, both the historical health data and the real-time monitoring data include the current data, voltage data, and gap data of the switch machine.

[0012] As a preferred technical solution, the health assessment function is:

[0013]

[0014] where x is current data, voltage data, or notch data, and μ and σ are the mean and standard deviation of the current data, voltage data, or notch data, respectively.

[0015] As a preferred technical solution, the stages include a short-term stage, a medium-term stage, and a long-term stage.

[0016] As a preferred technical solution, the preset times include a first preset time, a second preset time, and a third preset time. The preset times are arranged in ascending order as the first preset time, the second preset time, and the third preset time. The short-term stage is the first preset time, the medium-term stage is the second preset structure, and the long-term stage is the third preset stage.

[0017] As a preferred technical solution, μ and σ are obtained by maximum likelihood estimation and follow a normal distribution:

[0018] X ∼ N(μ, σ 2 ), σ ≥ 0

[0019] X is current data, voltage data, or notch data.

[0020] According to the maximum likelihood function L(μ, σ 2 ) we can get:

[0021]

[0022] where:

[0023]

[0024] So:

[0025]

[0026] We can get:

[0027]

[0028] Take the log-likelihood function and continue to simplify to get:

[0029]

[0030] Set the partial derivative to zero to get the first equation:

[0031]

[0032] Set the partial derivative to zero to get the second equation:

[0033]

[0034] The solution of μ obtained from the first equation is:

[0035]

[0036] Substituting μ into the second equation gives:

[0037]

[0038] n is the number of samples of the same data type, μ is the mean value corresponding to the number of samples, and x i is the current data or voltage data or notch data corresponding to the number of samples, and σ 2 is the variance corresponding to the number of samples.

[0039] As a preferred technical solution, according to the number of samples of the current data, voltage data, and notch data in the short-term stage, medium-term stage, and long-term stage respectively, calculate the μ and σ corresponding to the current data in the short-term stage, medium-term stage, and long-term stage 2 , calculate the μ and σ corresponding to the voltage data in the short-term stage, medium-term stage, and long-term stage 2 , calculate the μ and σ corresponding to the notch data in the short-term stage, medium-term stage, and long-term stage 2 .

[0040] As a preferred technical solution, take the square root of the variance σ 2 to obtain the standard deviation σ.

[0041] As a preferred technical solution, substitute the current data, voltage data, and notch data, as well as the corresponding μ and σ, into the health assessment function to calculate the corresponding scores of the current data in the short-term stage, medium-term stage, and long-term stage, calculate the corresponding scores of the voltage data in the short-term stage, medium-term stage, and long-term stage, and calculate the corresponding scores of the notch data in the short-term stage, medium-term stage, and long-term stage.

[0042] According to another aspect of the present invention, a health assessment system for a switch machine is provided. The system includes a current sensor, a voltage sensor, a notch sensor, and a control module. The current sensor, voltage sensor, and notch sensor are respectively communicatively connected to the control module. The current sensor, voltage sensor, and notch sensor respectively collect current data, voltage data, and notch data. The control module implements any of the above methods according to the current data, voltage data, and notch data.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention evaluates by constructing an adaptive health assessment function and combining short-term, medium-term, and long-term stages, which can not only evaluate the current operating condition of the device but also give early warnings of potential fault trends.

[0045] 2. The present invention uses the normal distribution method to collect the current, voltage, and gap data of the switch machine in real time and conducts multi-dimensional evaluations to capture the operating conditions of the switch under different monitoring methods.

[0046] 3. The present invention creates a health assessment function in combination with normal distribution knowledge. Assuming that under normal working conditions, the distributions of measured values such as voltage and current are close to normal distribution, the average value represents the voltage in the healthy state, and the standard deviation represents the normal fluctuation range of the voltage, and quantifies and calculates the health score of the device.

[0047] 4. The present invention sets short-term, medium-term, and long-term stages according to the length of time. This hierarchical time frame makes the evaluation process more refined and systematic, and can provide corresponding data support and insights for different operation and maintenance needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a schematic diagram of the working process of the present invention;

[0049] Figure 2 is a comparison diagram of a single data type and actual data of the present invention;

[0050] Figure 3 is a schematic diagram of the score of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] The switch device plays a crucial role in the operation of urban rail transit. As an essential part of the rail transit network, its normal and healthy operating condition is directly related to the driving safety and efficiency of trains and the smooth operation of the entire urban rail network. The main responsibility of the switch device is to ensure that trains can safely and accurately transfer from one track to another and complete complex track switching tasks, thus ensuring the efficient and orderly operation of trains in the busy urban rail transit network.

[0053] With the continuous expansion of the urban rail transit network and the increase in train operation frequencies, the working load borne by turnout equipment is becoming increasingly heavy. Long-term and high-frequency use makes turnout equipment prone to problems such as wear and aging, which may in turn trigger various faults and safety hazards. Therefore, continuous and accurate health monitoring of turnout equipment is carried out to detect and solve potential problems in a timely manner.

[0054] As an important means of health monitoring, the equipment health assessment method plays a crucial role in the maintenance of turnout equipment. By real-time monitoring the operating status of turnout equipment and analyzing the data, the equipment health assessment method can accurately evaluate the health status of the equipment and detect potential faults and hidden dangers in a timely manner. Compared with traditional regular inspections and fault troubleshooting, this technology has higher real-time performance and accuracy, can detect abnormal conditions of the equipment earlier, and thus take maintenance measures in advance to avoid the occurrence of faults.

[0055] The concept of the present invention is as follows: Monitor data of turnout equipment, such as current, voltage, and notch indication, are collected through sensors, and these data are analyzed and processed with the help of advanced algorithms and models to construct a health assessment function. When new monitoring data are generated, this function can calculate the health status of the turnout in real time and generate a corresponding health score. Maintenance personnel can reasonably judge the maintenance priority of the equipment based on the health index and formulate a targeted maintenance plan, thereby effectively improving the efficiency and quality of maintenance work.

[0056] The present invention provides a health assessment method and system for a turnout machine; by constructing an adaptive health assessment function and combining short-term, medium-term, and long-term stages for assessment, the present invention can not only evaluate the current operating status of the equipment but also give an early warning of potential fault trends. The present invention uses the normal distribution method to evaluate the current, voltage, and notch data of the turnout machine collected in real time in multiple dimensions, so as to capture the working conditions of the turnout under different monitoring methods. The present invention combines normal distribution knowledge to create a health assessment function. Assuming that under normal working conditions, the distributions of measured values such as voltage and current are close to the normal distribution, the mean value represents the voltage in the healthy state, and the standard deviation represents the normal fluctuation range of the voltage, and the health score of the equipment is quantified and calculated. The present invention sets short-term, medium-term, and long-term stages according to the length of time. This hierarchical time framework makes the assessment process more refined and systematic, and can provide corresponding data support and insights for different operation and maintenance needs.

[0057] Embodiment 1

[0058] As Figures 1-3 shown, a health assessment method for a turnout machine, the method specifically includes:

[0059] S1. Collect the historical health data and real-time monitoring data of the switch machine respectively;

[0060] S2. Construct a health assessment function based on the historical health data and using the normal distribution;

[0061] S3. Divide different stages according to different preset times, and perform stage scoring according to the real-time monitoring data and the health assessment function.

[0062] Both the historical health data and the real-time monitoring data include the current data, voltage data and gap data of the switch machine.

[0063] In this embodiment, the health assessment function based on the monitoring data: The core idea is to construct a health assessment function based on the normal distribution through statistical analysis of the historical health monitoring data of the switch. Compared with the existing health evaluation functions, the present invention constructs the function from the perspective of data-driven, avoiding the need for in-depth analysis of complex physical mechanisms and significantly reducing the implementation difficulty. At the same time, the present invention is data-based, has strong versatility, is applicable to the health evaluation of switches in different scenarios, and is easy to deploy and implement.

[0064] The health assessment function calculates a unique score based on the newly generated monitoring index data. In the evaluation process, each type of characteristic data will correspond to an independent scoring function. The characteristic data includes current data, voltage data and gap data, which are used to quantify the impact of this characteristic on the health state of the equipment. Specifically, the single-characteristic data scoring curve function is a statistical model constructed based on the historical data of specific characteristic monitoring. By calculating the score of the newly collected data, the contribution of this characteristic to the equipment health is quantified, and finally the health score is generated.

[0065] The health assessment function is:

[0066]

[0067] Among them, x is the current data or voltage data or gap data, and μ and σ are the mean and standard deviation of the current data or voltage data or gap data respectively.

[0068] The stages include a short-term stage, a medium-term stage and a long-term stage.

[0069] The preset times include a first preset time, a second preset time and a third preset. The preset times are in ascending order as the first preset time, the second preset time and the third preset. The short-term stage is the first preset time, the medium-term stage is the second preset structure, and the long-term stage is the third preset stage.

[0070] The μ and σ will be obtained by maximum likelihood estimation. When following the normal distribution:

[0071] X ~ N(μ, σ 2 ), σ ≥ 0. X is current data, voltage data, or notch data.

[0072] According to the maximum likelihood function L(μ, σ 2 ), we can obtain:

[0073]

[0074] Where:

[0075]

[0076] Therefore:

[0077]

[0078] We can obtain:

[0079]

[0080] Taking the log-likelihood function and continuing to simplify, we can obtain:

[0081]

[0082] By setting the partial derivative to zero, we get the first equation:

[0083]

[0084] By setting the partial derivative to zero, we get the second equation:

[0085]

[0086] From the first equation, the solution for μ is:

[0087]

[0088] Substituting μ into the second equation, we can obtain:

[0089]

[0090] n is the number of samples of the same data type, μ is the mean of the corresponding number of samples, x i is the current data, voltage data, or notch data of the corresponding number of samples, and σ 2 is the variance of the corresponding number of samples.

[0091] In this embodiment, through the above derivation, when a sufficient number of turnout health monitoring data is collected, the parameter σ can be calculated according to the formula 2And μ, and then construct a health assessment function. Combine the knowledge of normal distribution to create a health scoring function. We assume that under normal working conditions, the distributions of measured values such as voltage and current are close to normal distribution, where the mean represents the voltage in the healthy state, and the standard deviation represents the normal fluctuation range of the voltage. Therefore, we can use the distance between the voltage value and the mean to evaluate the health state. Specifically, through the constructed health assessment function, analyze the current monitoring data, and combine big data statistical methods to quantify and calculate the health score of the device.

[0092] According to the sample numbers of current data, voltage data, and notch data in the short-term stage, medium-term stage, and long-term stage respectively, calculate the μ and σ corresponding to the current data in the short-term stage, medium-term stage, and long-term stage 2 , calculate the μ and σ corresponding to the voltage data in the short-term stage, medium-term stage, and long-term stage 2 , calculate the μ and σ corresponding to the notch data in the short-term stage, medium-term stage, and long-term stage 2 .

[0093] Take the square root of the variance σ 2 to obtain the standard deviation σ.

[0094] Substitute the current data, voltage data, notch data, and the corresponding μ and σ into the health assessment function, calculate the corresponding scores of the current data in the short-term stage, medium-term stage, and long-term stage, calculate the corresponding scores of the voltage data in the short-term stage, medium-term stage, and long-term stage, and calculate the corresponding scores of the notch data in the short-term stage, medium-term stage, and long-term stage.

[0095] In this embodiment, for the assessment of the short-term stage, medium-term stage, and long-term stage, different time units are adopted for assessment according to the scale of the existing data volume. Specifically, the short-term assessment is carried out on a daily basis, focusing on capturing the immediate changes in the device performance; the medium-term assessment is carried out on a weekly basis, focusing on analyzing the performance fluctuations and trends within a week; while the long-term assessment is carried out on a monthly basis, aiming to gain insights into the monthly performance changes of the device and its long-term operating conditions. This hierarchical time frame makes the assessment process more refined and systematic, and can provide corresponding data support and insights for different operation and maintenance needs.

[0096] Furthermore, the evaluation of equipment health in the short - term stage: This part focuses on real - time monitoring and rapid response. The model analyzes the recently collected monitoring data to judge the immediate health status of the equipment. Currently, three health status evaluation methods are designed, which are based on current data, voltage data, and notch data respectively. Corresponding health status evaluation models are constructed for each type of data. Under each health status evaluation method, it can be further subdivided according to the short - term stage, medium - term stage, and long - term stage, realizing the evaluation of the short - term stage, medium - term stage, and long - term stage of current data, the evaluation of the short - term stage, medium - term stage, and long - term stage of voltage data, and the evaluation of the short - term stage, medium - term stage, and long - term stage of notch data.

[0097] Adopting a data - analysis method based on statistics, the model can promptly identify any signs of abnormality or deviation from the normal operating range, thus supporting timely fault diagnosis and maintenance decision - making. In addition, the model also supports on - site data verification at sampling points, improving the accuracy of diagnosis and response speed.

[0098] Taking the switch voltage data as an example, using the collected and sorted historical data of switch representation voltage, calculate through the health evaluation function. First, calculate the values of μ and σ:

[0099] μ = 61.5182

[0100] σ = 1.2169

[0101] Substitute μ and σ into the health evaluation function to calculate the equipment score of voltage data in the short - term stage.

[0102] The evaluation of equipment health in the medium - term stage: In this stage, the model focuses on the medium - term performance evaluation of the equipment. Through in - depth statistics and analysis of the monitoring data over a period of time, the model can not only detect current faults but also judge the health degree of each modal data within a certain duration (such as one week). This method helps to identify potential hidden faults in advance, so as to take preventive measures and reduce unexpected equipment shutdowns. The model also supports verification using the historical daily report data of non - model sample equipment, thus enhancing its adaptability and flexibility.

[0103] The evaluation of equipment health in the long - term stage: By analyzing the historical monitoring data and maintenance records of the equipment, it helps to evaluate the life of the equipment. Long - term evaluation is another important part of the model. Through comprehensive analysis of the historical monitoring data and maintenance records of the equipment, the model can evaluate the overall service life and performance degradation trend of the equipment. This is crucial for formulating effective maintenance strategies and budget planning. Long - term health evaluation can help urban rail operators optimize resource allocation, extend equipment life, and ensure safety and reliability at the same time.

[0104] When the voltage data is 60V, it is calculated through the constructed health assessment function, and the result is mapped to the range of 0 - 100. Finally, the health score of the switch device is 95.55 (rounded to two decimal places). This score indicates that the device is currently in a relatively healthy state, which can provide a clear quantitative index of the health status for the operation and maintenance personnel, assist in scientifically judging the maintenance priority of the device, and formulate a more accurate maintenance plan.

[0105] Embodiment 2

[0106] A health assessment system for a switch machine, the system includes a current sensor, a voltage sensor, a notch sensor and a control module. The current sensor, the voltage sensor and the notch sensor are respectively communicatively connected to the control module. The current sensor, the voltage sensor and the notch sensor respectively collect current data, voltage data and notch data. The control module implements the method described in any one of the above according to the current data, voltage data and notch data.

[0107] As described above, only the specific embodiments of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A health evaluation method for a turnout machine, characterized in that: The method specifically comprises: S1, collect historical health data and real-time monitoring data of turnout machines respectively; S2, constructing a health evaluation function based on historical health data and using normal distribution; S3. Divide different stages according to different preset times, and perform stage scoring based on real-time monitoring data and health assessment functions.

2. A health evaluation method for a turnout machine according to claim 1, characterized in that: The historical health data and real-time monitoring data both include current data, voltage data and gap data of the turnout machine.

3. A health evaluation method for a turnout machine according to claim 1, characterized in that: The health evaluation function is: Wherein, x is the current data or voltage data or gap data, μ and σ are the mean and standard deviation of the current data or voltage data or gap data, respectively.

4. A health evaluation method for a turnout machine according to claim 3, characterized in that: The stages include a short-term stage, a medium-term stage and a long-term stage.

5. A health evaluation method for a turnout machine according to claim 4, characterized in that: The preset time includes a first preset time, a second preset time and a third preset time. The preset times are the first preset time, the second preset time and the third preset time from short to long. The short-term stage is the first preset time, the medium-term stage is the second preset structure, and the long-term stage is the third preset stage.

6. A health evaluation method for a turnout machine according to claim 3, characterized in that: The μ and σ are obtained by maximum likelihood estimation, and when they obey the normal distribution: X~N(μ,σ 2 ),σ≥0 X is current data or voltage data or gap data. According to the maximum likelihood function L(μ,σ 2 ) can be obtained: in: so: We can get: Taking the log-likelihood function and continuing to simplify, we get: By setting the partial derivative to zero, we get the first formula: By setting the partial derivative to zero, we get the second formula: From the first formula, the solution of μ is: Substituting μ into the second formula yields: n is the number of samples of the same data type, μ is the mean of the corresponding number of samples, and x i is the current data or voltage data or gap data corresponding to the number of samples, σ 2 is the variance corresponding to the sample size.

7. A health evaluation method for a turnout machine according to claim 6, characterized in that: According to the number of samples of current data, voltage data and gap data in the short-term stage, medium-term stage and long-term stage respectively, μ and σ corresponding to the current data in the short-term stage, medium-term stage and long-term stage are calculated. 2 , calculate the μ and σ corresponding to the voltage data in the short-term, medium-term and long-term stages 2 , calculate the μ and σ corresponding to the gap data in the short-term, medium-term and long-term stages 2 .

8. A health evaluation method for a turnout machine according to claim 7, characterized in that: The variance σ 2 Taking the square root gives the standard deviation σ.

9. A health evaluation method for a turnout machine according to claim 8, characterized in that: The current data, voltage data, gap data and corresponding μ and σ are brought into the health assessment function to calculate the corresponding scores of the current data in the short-term, medium-term and long-term stages, calculate the corresponding scores of the voltage data in the short-term, medium-term and long-term stages, and calculate the corresponding scores of the gap data in the short-term, medium-term and long-term stages.

10. A health evaluation system for a turnout machine, characterized in that: The system includes a current sensor, a voltage sensor, a notch sensor and a control module, wherein the current sensor, voltage sensor and notch sensor are respectively connected to the control module for communication, and the current sensor, voltage sensor and notch sensor collect current data, voltage data and notch data respectively. The control module implements the method described in any one of claims 1 to 9 according to the current data, voltage data and notch data.

Images