A curve health assessment method based on vehicle body vibration and noise response

Abstract

The application provides a curve health assessment method based on vehicle body vibration and noise response, characterized by comprising: collecting vehicle body vibration response and noise signal data on a plurality of different curve devices, including vehicle body horizontal acceleration data, vertical acceleration data and noise data in the vehicle compartment; using the collected data to calculate and construct an index system, including horizontal acceleration effective value E H , vertical acceleration effective value E V and noise effective value E P ; based on the horizontal acceleration effective value E H , the vertical acceleration effective value E V and the noise effective value E P , using a clustering algorithm to construct a respective curve health index CHI of the plurality of different curve devices, wherein CHI=f(E V +E H ,E P )=f(E A ,E P ), E A =E V +E H , indicating the acceleration effective value, f indicating the mapping relationship between E A , E P and CHI, determined by the clustering algorithm; based on the value of CHI, the health of the curve device is assessed. Corresponding devices, computer readable storage media and systems are also provided.

Description

Technical Field

[0001] This invention relates to the field of rail transit technology, and in particular to a curve-based health assessment method based on vehicle vibration and noise response. Background Technology

[0002] A curve is a railway line that changes direction from one direction to another on a horizontal plane, connecting various points along the way. Its primary function is to change the direction of train travel on a railway line. Curves constitute a significant portion of urban rail transit lines, and their radii are relatively small. The health status assessment results of curves are crucial for managers to formulate curve maintenance strategies; unscientific health assessment methods will lead to unreasonable maintenance activities. Therefore, how to scientifically and comprehensively assess the health status of curves is of great importance for optimizing the development of maintenance plans.

[0003] According to the "Methods for Monitoring and Evaluation of Urban Rail Transit Facilities Operation - Part 4: Track and Subgrade" (GB / T39559.4-2020), in actual operation and maintenance management, curve quality status assessment evaluates the quality of curves from four dimensions: static irregularities (or static geometric dimensional exceedances), dynamic irregularities (or dynamic geometric dimensional exceedances), curve roundness (or curve versine), and curve wear. Data sources for curve status assessment include track inspection vehicle data, track inspection instrument data, rail corrugation detector data, manual track inspection data, and curve versine inspection data.

[0004] However, the methods described above have some obvious shortcomings:

[0005] 1. The curve detection cycle specified in "Methods for Monitoring and Evaluation of Urban Rail Transit Facilities Operation - Part 4: Track and Subgrade" (GB / T39559.4-2020) is relatively long, and the detection cycles vary for different detection methods, as detailed in Table 1 below. This makes it difficult to obtain the data for the above four dimensions of status assessment indicators at the same time point.

[0006] Table 1 shows the detection cycle of the curve device under different conditions.

[0007]

[0008] 2. The classification of defects in different dimensions of curved rail equipment varies, making a comprehensive assessment of curve health based on the aforementioned four dimensions of condition evaluation indicators difficult. Dynamic irregularities are classified into four levels: Level I, Level II, Level III, and Level IV; static irregularities are classified into three levels: operational acceptance, planned maintenance, and temporary repair. Curve smoothness (or curve versine) defects are classified into two levels: operational acceptance and daily maintenance. Curve wear mainly includes different forms such as upper rail side wear, lower rail head crushing, and wave wear. Curve wear is a type of rail damage defect, and rail damage is classified into three levels: minor damage, severe damage, and breakage.

[0009] Therefore, new technologies and methods are needed to at least partially address the shortcomings of existing technologies and enable a scientific assessment of the health status of curved devices. Summary of the Invention

[0010] Unlike traditional methods that assess curve health based on dynamic and static smoothness, roundness, and wear, this invention proposes a method for comprehensively evaluating the health status of curve equipment based on the vibration response of the bus body and the noise inside the carriage during the curve passage process, which at least partially solves the problems existing in the prior art.

[0011] To achieve the above objectives, this application provides a curve-based health assessment method for vehicle body vibration and noise response, comprising:

[0012] a. Collect vehicle vibration response and noise signal data from multiple different curve devices, including vehicle vibration horizontal acceleration data, vertical acceleration data, and noise data inside the vehicle compartment;

[0013] b. Utilize the collected data to calculate and construct an indicator system, including using horizontal acceleration data and vertical acceleration data to construct the effective value of horizontal acceleration E respectively. H and the effective value of vertical acceleration E V Using noise data to construct the effective noise value E P ;

[0014] c. Based on the effective value of horizontal acceleration E H Effective value of vertical acceleration E V and the effective value of noise E P To construct a discrete quantitative index for measuring the health status of curved equipment, namely the Curve Health Index (CHI), where CHI ∈ [1, 2, 3, 4], where...

[0015] CHI=f(E V +E H E P )=f(E A E P ),

[0016] E A =E V +E H , represents the effective value of acceleration, f represents E A and E P The mapping relationship with CHI was determined using clustering methods;

[0017] d. Assess the health of the curve device based on the CHI value.

[0018] According to an embodiment of the present invention, the clustering method includes a hierarchical clustering algorithm and a K-means clustering algorithm.

[0019] According to an embodiment of the present invention, in step b, the effective value of vertical acceleration E is calculated using the formula and respectively. V and the effective value of horizontal acceleration E H :

[0020]

[0021]

[0022] Among them, y V Represents the vertical acceleration signal value, y H represents the horizontal acceleration signal value, x represents the signal sampling interval, and L represents the total length of the curve;

[0023] Using formulas and calculating the effective value of noise E P :

[0024]

[0025]

[0026] Among them, L p The sound pressure level is represented by p, which represents the difference in atmospheric pressure when there is sound and when there is no sound. p0 is the reference sound pressure.

[0027] According to an embodiment of the present invention, in step c, the clustering method includes:

[0028] Step 1: In the case of E A E P In the constructed two-dimensional space, it is assumed that the curve health index (CHI) of the multiple curve devices are all different, and each of the multiple curve devices is a cluster;

[0029] Step 2: Calculate the curve device cluster D using formula (7). v With D e The distance D between each pair ve :

[0030]

[0031] Where t v Represents curve device cluster D v The number of curved devices in the t e Represents curve device cluster D e The number of curved devices in the diagram; d ij G represents the curve device in the cluster. i and G j The distance between any two pairs of objects is calculated using formula (6):

[0032]

[0033]

[0034] Among them, E i It is curve device G i State attribute variable, E i =(E i,A E i,P ) T E j It is curve device G j State attribute variable, E j =(E j,A E j,P ) T ;

[0035] Step 3: Place D ve The two smallest clusters are merged into a new cluster;

[0036] Step 4: Repeat steps 2 and 3 until the multiple curved devices are divided into 4 clusters to obtain the results of the health feature clustering of the curved devices.

[0037] According to an embodiment of the present invention, the clustering method further includes:

[0038] Based on the clustering results of health characteristics of curved devices, the possible set of CHI values ​​for curved devices and the cluster of curved devices corresponding to each CHI value are determined; and

[0039] The CHI values ​​of the plurality of curve devices are determined based on their hierarchical relationship with the cluster.

[0040] According to an embodiment of the present invention, the curve health assessment method based on vehicle vibration and noise response further includes determining the set of curve devices that ultimately need repair based on the CHI values ​​of the plurality of curve devices and the results of cluster analysis.

[0041] According to another aspect of the present invention, a curve-based health assessment device based on vehicle body vibration and noise response is provided, comprising:

[0042] Vibration response and noise data acquisition module, used to acquire vibration response data and noise data;

[0043] The module for calculating and constructing the indicator system is used to calculate and construct the indicator system.

[0044] A comprehensive evaluation algorithm module for curve health is used to construct discrete quantitative indicators for measuring the health status of curve-based equipment; and

[0045] The health assessment module for curve equipment assesses the health of the curve equipment based on the CHI value.

[0046] According to another aspect of the invention, the invention also includes a computer-readable storage medium storing executable instructions that, when executed by a processor, cause the aforementioned curve-based health assessment method based on vehicle vibration and noise response to be executed. The readable storage medium can be a non-volatile memory, such as a hard disk or magnetic disk, and can be used in various terminals, such as computers and servers.

[0047] According to another aspect of the present invention, the present invention also includes a curve health assessment system based on vehicle body vibration and noise response. The system includes a processor and a storage device, the storage device being used to store executable instructions. When the executable instructions are executed by the processor, the aforementioned curve health assessment method based on vehicle body vibration and noise response can be implemented.

[0048] The above and other objects, advantages and features of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description

[0049] The following description will detail some specific embodiments of the invention by way of example and not limitation, with reference to the accompanying drawings. The objectives and features of the invention will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:

[0050] Appendix Figure 1 This is a flowchart of a curve-based health assessment method for vehicle vibration and noise response according to an embodiment of the present invention.

[0051] Appendix Figure 2 This is a schematic diagram of the health assessment device based on vehicle vibration and noise response curves according to an embodiment of the present invention.

[0052] Appendix Figure 3 This is a structural diagram of a curve-based health assessment system for vehicle vibration and noise response, consisting of a processor 201 and a memory 202, according to an embodiment of the present invention.

[0053] Figure 4This is a clustering result diagram of the curve equipment health index (CHI) obtained by analyzing the curve health assessment method based on vehicle vibration and noise response according to an embodiment of the present invention. Detailed Implementation

[0054] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0055] like Figure 1 As shown, the curve-based health assessment method based on vehicle vibration and noise response according to this embodiment includes:

[0056] Step 1: Vibration Response and Noise Data Acquisition

[0057] To acquire data on the horizontal and vertical acceleration of train body vibration and noise levels inside the carriages, this embodiment expands the hardware structure of the track quality inspection instrument by adding a noise measurement module to the existing acceleration measurement module. The improved track quality inspection instrument can be carried by personnel or installed on the electric train to acquire the latest data on the train's horizontal and vertical acceleration of body vibration and noise levels inside the carriages.

[0058] Step Two: Construction and Calculation of the Indicator System

[0059] (1) Effective value of horizontal acceleration

[0060] Studies have found that poor track smoothness, roundness, and wear of curved rails are the main causes of aggravated contact vibration, leading to excessive impact acceleration, intensifying the wheel-rail dynamic interaction, causing abnormal vehicle vibration, and potentially resulting in breakage of rail clips, detachment of fasteners, and impact damage to critical components at the bottom of the vehicle. Therefore, the effective value of vibration acceleration can be used to measure the vibration response of the vehicle body; the calculation formula is shown in Formula [Formula Name].

[0061]

[0062]

[0063] Among them, E V E represents the effective value of vertical acceleration. H This represents the effective value of horizontal acceleration. y V Represents the vertical acceleration signal value, y HThe value represents the horizontal acceleration signal, x represents the signal sampling interval (e.g., 0.25m), and L represents the total length of the curve.

[0064] (2) Effective value of noise

[0065] There is a positive correlation between noise levels inside the train carriage and the severity of rail wear; severe rail wear significantly increases noise levels inside the carriage. A poor acoustic environment inside the carriage reduces service quality. Unlike the definition formula for the effective value of acceleration, the average sound pressure level E can be used. P This describes the overall noise level of the carriage when passing through a curve, as shown in the formula.

[0066]

[0067]

[0068] In the formula, E P Indicates the effective value of noise; L p Sound pressure level (p) describes the intensity of noise inside the carriage, measured in dB (decibels). Sound pressure p represents the difference in atmospheric pressure between when sound is present and when there is no sound. p0 is the reference sound pressure, which can be, for example, 2 × 10⁻⁶. -5 Pa.

[0069] Step 3: Construction of a comprehensive evaluation algorithm for curve health

[0070] To comprehensively evaluate the health status of the curve, this embodiment proposes the Curve Health Index (CHI) from the perspective of the vibration and noise response of the train during its journey along the curve. CHI is a discrete quantitative index that measures the health status of the equipment along the curve; CHI ∈ [1,2,3,4] is a measure of the health status of the curve's equipment. V E H and E P Based on comprehensive analysis, different CHI values ​​represent different health characteristics of the curve device, and the calculation formula is as follows:

[0071] CHI=f(E V +E H E P )=f(E A E P )\*MERGEFORMAT (5)

[0072] Among them, E A E represents the effective value of acceleration. A =E V +E H f represents E V E H and E PThe mapping relationship with CHI can be determined, for example, using a hierarchical clustering algorithm. Optionally, the hierarchical clustering algorithm in this embodiment can be replaced with other clustering algorithms, such as the K-means clustering algorithm.

[0073] Cluster analysis is performed on multiple curve device samples, dividing them into 4 clusters (or classes). Curve devices belonging to the same cluster have a high degree of similarity in health status, while curve devices belonging to different clusters have a large difference in health status. The algorithm flow is as follows.

[0074] Step 1: In the case of E A E P In the two-dimensional space, assume that the health characteristics of the U curved devices are different, and each of the U curved device samples is a cluster.

[0075] Step 2: Calculate the pairwise distances between each cluster of curved devices to determine the similarity of their health characteristics. Curved device sample G i With G j The distance d between ij The calculation, using the Euclidean distance definition, is shown in the formula. The distance d between curved device samples... ij Used to measure the similarity of health characteristics among samples from different equipment. Among them, E i It is curve device G i State attribute variable, E i =(E i,A E i,P ) T E j It is curve device G j State attribute variable, E j =(E j,A E j,P ) T .

[0076]

[0077]

[0078] Curve device cluster D v With D e Distance D ve The Average Linkage strategy is used, as shown in the formula, where the distance between curve device clusters is the average distance between all pairs of curve device samples within the cluster. The similarity of health features among curve device clusters is represented by D. ve To measure, D ve The smaller the value, the higher the similarity of health characteristics. Among them, t... v Represents curve device cluster D v The number of samples in t.e Indicates cluster D e The number of samples in the sample.

[0079]

[0080] Step 3: Place D ve The two smallest clusters are merged into a new cluster, that is, the two curve device clusters with the most similar health characteristics are merged into a new cluster.

[0081] Step 4: Repeat Step 2 and Step 3 until the U curved devices are divided into 4 clusters to obtain the results of the health feature clustering of the curved devices.

[0082] Step 5: Based on the clustering results of the health features of the curved devices, determine the possible set of CHI values ​​for the curved devices and the cluster of curved devices corresponding to each CHI value.

[0083] Step 6: Determine the CHI values ​​of U curve device samples based on the affiliation relationship between the curve device samples and the cluster, and then end.

[0084] Table 2 shows the curve health characteristics and recommended maintenance strategies for different CHI values.

[0085] Table 2. Different health characteristics of the curved equipment and recommended maintenance strategies.

[0086]

[0087]

[0088] Step 4: CHI-based Curve Maintenance Decisions

[0089] Based on the different health characteristics of the curved equipment described in Table 2, it is recommended to prioritize the maintenance activities of the curved equipment with CHI=4. At the same time, the maintenance capabilities of the management department should be taken into account to determine the final set of curved equipment that needs maintenance.

[0090] like Figure 2 As shown, this embodiment also provides a curve-based health assessment device for vehicle body vibration and noise response, including:

[0091] Vibration response and noise data acquisition module 10 is used to acquire vibration response data and noise data;

[0092] The indicator system calculation and construction module 20 is used to calculate and construct the indicator system;

[0093] The comprehensive health assessment algorithm construction module 30 is used to construct discrete quantitative indicators for measuring the health status of curve equipment; and

[0094] The health assessment module 40 for the curve equipment assesses the health of the curve equipment based on the CHI value.

[0095] According to another aspect of the invention, the invention also includes a computer-readable storage medium storing executable instructions that, when executed by a processor, cause the aforementioned health assessment method based on vehicle vibration and noise response curves to be executed. The readable storage medium can be a non-volatile memory, such as a hard disk or magnetic disk, and can be used in various terminals, such as computers and servers. The readable storage medium can include high-speed random access memory, and can also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device, and can be used in various terminals, such as computers and servers.

[0096] Storage media also include, but are not limited to, any type of disk (including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic cards, or optical cards. In other words, storage media includes any medium by which a device (e.g., a computer) stores or transmits information in a readable form. Storage media can also be read-only memory, magnetic disks, or optical disks, etc.

[0097] Embodiments of the present invention also provide a computer program product that, when run on a computer, causes the computer to perform the aforementioned related steps to implement the methods described in the above embodiments.

[0098] The apparatus, computer storage medium, computer program product or chip provided by the present invention are all used to execute the corresponding methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0099] In addition, such as Figure 3 As shown, a curve health assessment system based on vehicle vibration and noise response is also provided. The system includes a processor 201 and a storage device 202. The storage device 202 is used to store computer execution instructions. When the device is running, the processor 201 can execute the computer execution instructions stored in the storage device 202 to make the chip execute the above-mentioned curve health assessment method based on vehicle vibration and noise response. Specific Implementation

[0101] The method of this invention was used to inspect a subway line 1 in a northern city. The line has 61 curved sections. Using a track quality detector, the effective values ​​of vertical acceleration, horizontal acceleration, and noise were measured for each curved section. A comprehensive health assessment algorithm for the curved sections was used to obtain the corresponding health index (CHI) value for each section. The results are as follows: Figure 4 As shown. Analysis Figure 2 It can be seen that (1) the set of health characteristic CHI values ​​of the 61 curved devices on Line 1 is {1,3,4}. There are no curved devices with CHI=2 on Line 1. The corresponding health characteristics are small effective noise value and large effective acceleration value; (2) there are 8 curved devices with CHI=4, and maintenance should be arranged as soon as possible.

[0102] Compared with existing technologies, the method of this invention can achieve beneficial technical effects:

[0103] 1) Starting from the basic function of curve equipment, this invention innovatively utilizes vehicle response indicators under wheel-rail contact to assess the health status of curve equipment in real time, which is an effective supplement to the current general condition assessment methods for curve quality. The method of this invention can be measured when the vehicle is working normally, which is convenient and fast.

[0104] (2) Unlike traditional evaluation methods, the four dimensions of the curve state evaluation index have a long detection cycle and different detection cycles, which makes it difficult to obtain the above indexes at the same time cutoff point. The vehicle response indexes used in this invention are all obtained from the same time cutoff point.

[0105] (3) The detection device (improved line quality detector) proposed in this invention can be carried by personnel and placed or installed on subway trains to realize the real-time and accurate assessment of the health status of the entire line curve, which helps managers to grasp the deterioration process of the curve status and understand its deterioration law.

[0106] Although this application has been described through embodiments, those skilled in the art will know that this application has many modifications and variations without departing from the spirit of this application, and it is intended that the appended claims cover such modifications and variations without departing from the spirit of this application.

Claims

1. A curve-based health assessment method for vehicle body vibration and noise response, characterized in that, include: a. Collect vehicle vibration response and noise signal data from multiple different curve devices, including vehicle vibration horizontal acceleration data, vertical acceleration data, and noise data inside the vehicle compartment; b. Utilize the collected data to calculate and construct an indicator system, including using horizontal acceleration data and vertical acceleration data to construct effective values ​​of horizontal acceleration respectively. and the effective value of vertical acceleration ; Constructing RMS noise data ; c. Based on the effective value of horizontal acceleration Effective value of vertical acceleration and the effective value of noise To construct a discrete quantitative index for measuring the health status of curved equipment, namely the Curve Health Index (CHI). ,in ， , represents the effective value of acceleration, f represents and The mapping relationship with CHI was determined using clustering methods; d. Assess the health of the curved equipment based on the CHI value; In step c, the clustering method includes: Step 1: In the process of... In the two-dimensional space formed, it is assumed that the curve health characteristics of the multiple curve devices are different, and each of the multiple curve devices is a cluster. Step 2: Calculate the curve device cluster using formula (7) and Distance between each pair : （7） in Representing curve device clusters The number of curve devices in the diagram; Representing curve device clusters The number of curve devices in the diagram; Curve device in cluster and The distance between any two pairs of objects is calculated using formula (6): （6） in, It is a curve device State attribute variables, , It is a curve device State attribute variables, ; Step 3: Put The two smallest clusters are merged into a new cluster; Step 4: Repeat steps 2 and 3 until the multiple curved devices are divided into 4 clusters to obtain the results of the health feature clustering of the curved devices.

2. The method for health assessment based on vehicle body vibration and noise response curves according to claim 1, characterized in that, The clustering methods include hierarchical clustering algorithms and K-means clustering algorithms.

3. The method for health assessment based on vehicle body vibration and noise response curves according to claim 1, characterized in that, In step b, the effective value of vertical acceleration is calculated using formulas (1) and (2) respectively. and the effective value of horizontal acceleration : （1） （2） in, This represents the vertical acceleration signal value. represents the horizontal acceleration signal value, x represents the signal sampling interval, and L represents the total length of the curve; The effective value of noise is calculated using formulas (3) and (4). : （3） （4） in, The sound pressure level (p) represents the difference in atmospheric pressure when sound is present and when there is no sound. The reference sound pressure level is used.

4. The method for curve-based health assessment of vehicle body vibration and noise response according to claim 1, characterized in that, The clustering method also includes: Based on the clustering results of health characteristics of curved devices, the possible set of CHI values ​​for curved devices and the cluster of curved devices corresponding to each CHI value are determined; and The CHI values ​​of the plurality of curve devices are determined based on their hierarchical relationship with the cluster.

5. The curve-based health assessment method for vehicle body vibration and noise response according to claim 4, characterized in that: It also includes determining the final set of curve devices that need repair based on the CHI values ​​of the multiple curve devices and the results of cluster analysis.

6. A computer-readable storage medium storing executable instructions, which, when executed by a processor, cause the execution of the curve health assessment method based on vehicle vibration and noise response as described in any one of claims 1-5.

7. A curve health assessment system based on vehicle vibration and noise response, the system comprising a processor and a storage device, the storage device being used to store executable instructions, which, when executed by the processor, enable the execution of the curve health assessment method based on vehicle vibration and noise response as described in any one of claims 1-5.

Images