Inspection data storage method of track maintenance vehicle-mounted inspection system
By dividing the storage area and dimensionality-reducing the inspection data in the on-board inspection system for rail maintenance, the problem of insufficient data storage for special events was solved, and more efficient data storage and historical data query were achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 江苏中车机电科技有限公司
- Filing Date
- 2025-12-22
- Publication Date
- 2026-05-19
AI Technical Summary
In existing technologies, the storage capacity of special event inspection data in the on-board inspection system for track maintenance is insufficient, resulting in poor historical query function and inability to effectively support track inspection and maintenance.
The storage area is physically divided into a circular storage area for general event data and a circular storage area for special event data. The inspection data is reduced to one-dimensional data. Through reasonable storage ratio and encoding reorganization, the storage capacity of special event inspection data is increased and the data storage logic is simplified.
It increases the storage capacity of special event inspection data, simplifies data storage logic, facilitates the extraction and analysis of historical data, and reduces storage requirements.
Smart Images
Figure CN122064682A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data storage technology, and specifically to a data storage method for an onboard inspection system for rail maintenance. Background Technology
[0002] The onboard inspection system for rail maintenance is an advanced device integrating multiple intelligent detection technologies, primarily used to improve the efficiency and accuracy of inspections of urban rail transit lines. Due to the limited storage capacity of the onboard inspection system, inspection data is typically stored in a circular, cyclical structure. That is, when the storage capacity is full, new data overwrites the oldest data, and this cycle continues for data storage. The onboard inspection system itself intelligently judges the inspection data, marking abnormal data (such as track breakage, bending, or obstructions) as special event data, while normal data is marked as general event data. Special event data is the data that requires focused attention during rail maintenance.
[0003] While existing technologies for storing inspection data categorize it into special event inspection data and general event inspection data, the two types of data are still stored in a circular data storage structure in chronological order. Since the amount of special event inspection data is much smaller than that of general event inspection data, the amount of special event inspection data that can be retrieved during historical inspection data queries is very limited, resulting in the historical query function not being able to adequately support track inspection and maintenance. Summary of the Invention
[0004] The present invention aims to provide a data storage method for inspection vehicles in a rail maintenance vehicle, which can increase the amount of inspection data for special events in the storage area and enable the query of historical data to obtain more effective data.
[0005] To address the aforementioned technical problems, the present invention provides the following technical solution: This invention provides a method for storing inspection data in a track maintenance vehicle-mounted inspection system, comprising the following steps: S1: Construct a dual-loop inspection data storage structure that includes a general event data loop storage area, a special event data loop storage area, and a basic table storage area. The basic table storage area stores one-dimensional historical data. S2: Real-time collection of inspection data; S3: Temporarily store the inspection data in flash memory, analyze and determine whether the inspection data is normal; if the inspection data is normal, mark the inspection data as general event inspection data; if the inspection data is abnormal, mark the inspection data as special event inspection data. S4: Reduce the dimensionality of general event inspection data and special event inspection data to one-dimensional data, and store them in the general event data circular storage area and the special event data circular storage area respectively; S5: Compare the dimensionality-reduced inspection data with historical data. If they are the same, delete the inspection data; if they are different, store the difference between the inspection data and the historical data.
[0006] By adopting the above technical solution, the storage area is physically divided into two regions: one for storing general event inspection data and the other for storing special event inspection data. This reasonable storage ratio significantly increases the volume of special event inspection data, providing effective data support for subsequent historical data review and analysis. Simultaneously, by reducing the multidimensional data structure to one-dimensional data for storage, the data storage logic becomes clearer and simpler, facilitating the retrieval of historical data. Furthermore, by establishing basic data tables, unchanged data can be directly deleted, greatly reducing the amount of data that needs to be stored. Then, through encoding and reorganization, the amount of stored data is further reduced, thus increasing the amount of stored data.
[0007] Optionally, the inspection data in S2 includes at least one of track geometry data, track surface and structural status data, environmental and ancillary facility data, and spatiotemporal synchronization data.
[0008] Optionally, the track geometry data includes flatness, track geometry parameters, and dynamic response indicators.
[0009] Optionally, the track surface and structural condition data include track surface defect data, fastener and sleeper condition data, and track internal damage data.
[0010] Optionally, the environmental and ancillary facility data includes foreign object intrusion, clearance, and tunnel status.
[0011] Optionally, the spatiotemporal synchronization data includes positioning and mileage information, speed and operating parameters.
[0012] Optionally, the difference between the inspection data and historical data stored in S5 at different times includes: S5-1: Sample the historical data of the inspection data to obtain a set of sample data; S5-2: Count the number of identical values Si in the sampled data, Pi, to obtain a set of values S={S1, S2, ..., Si, ... Sn}, and the corresponding number P={P1, P2, ..., Pi, ... n}; S5-3: Treat each value in S as a binary tree Ti with only one node, and put all the binary trees Ti into a priority queue, and arrange them in ascending order according to the size of the number Pi corresponding to Ti; S5-4: Take the two binary trees with the smallest number Pi from the priority queue, namely T1 and T2, and their corresponding numbers are P1 and P2; S5-5: Create a new binary tree Tnew, with T1 and T2 as its left and right subtrees. The number of binary trees Tnew is the sum of the number of the two subtrees: Pnew = P1 + P2. S5-6: Put the binary tree Tnew back into the priority queue and return to S5-3; S5-7: Repeat S5-3 to S5-6 until the root node of the binary tree is finally obtained by merging. That is, the priority queue has only two binary trees left. After merging these two binary trees, the root node Troot is obtained. A bit value is assigned to each branch of the binary tree. S5-8: The bit sequence along the path from the root node to each leaf node is the new storage code corresponding to Si.
[0013] In summary, the present invention has at least the following beneficial technical effects: 1. This invention physically divides the storage area into two regions, which are used to store general event inspection data and special event inspection data respectively. By dividing the storage area in a reasonable proportion, the amount of special event inspection data is greatly increased, thereby providing effective data support for subsequent review and analysis of historical data.
[0014] 2. This invention reduces complex multidimensional data structures into one-dimensional data for storage, making the data storage logic clearer and simpler, and facilitating the retrieval of historical data. By establishing a basic data table, unchanged data can be directly deleted, greatly reducing the amount of data that needs to be stored. Then, through encoding and recombination, the amount of stored data is reduced again, thereby increasing the amount of stored data. Attached Figure Description
[0015] Figure 1 This is a flowchart of a data storage method for an onboard inspection system for rail maintenance. Detailed Implementation
[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0017] This invention provides a method for storing inspection data in a track maintenance vehicle-mounted inspection system.
[0018] refer to Figure 1 A method for storing inspection data in a track maintenance vehicle-mounted inspection system includes the following steps: S1: Construct a dual-loop inspection data storage structure that includes a general event data loop storage area, a special event data loop storage area, and a basic table storage area. The basic table storage area stores one-dimensional historical data.
[0019] The ratio of the general event data circular storage area and the special event data circular storage area to the inspection data storage area is K:1, and K>1.
[0020] The general event data circular storage area is used to store general event inspection data, the special event data circular storage area is used to store special event inspection data, and the one-dimensional historical data is the historical data obtained from previous inspections after dimensionality reduction, which is stored through the basic data table in the basic table storage area.
[0021] S2: Real-time collection of inspection data.
[0022] The inspection data includes track geometry data, track surface and structural status data, environmental and ancillary facility data, and spatiotemporal synchronization data.
[0023] Track geometry data includes: flatness, obtained by collecting horizontal and vertical vibration signals of the track through devices such as accelerometers and lidar; track geometry parameters, obtained by real-time monitoring of track spacing, superelevation, track alignment, and elevation; and dynamic response indicators, obtained by detecting lateral and vertical vibrations of the car body, frame, and axle boxes, and calculating the track short-wave response and vehicle balance indicators.
[0024] The track surface and structural condition data include: track surface defect data, which is obtained by using high-definition cameras, line scan cameras, etc. to collect track surface images and identify surface defects such as missing pieces, scratches, and abnormal welds; fastener and sleeper condition data, which is obtained by using image analysis technology to detect structural problems such as missing elastic clips, loose bolts, and cracks in the track bed; and track internal damage data, which is obtained by using an ultrasonic flaw detection system to detect hidden damage inside the rail.
[0025] Environmental and ancillary facility data include: foreign object intrusion, obtained by monitoring foreign objects on the track, such as gravel and obstacles; clearance and tunnel status, obtained by simultaneously detecting railway clearance, tunnel walls and surrounding environment.
[0026] Spatiotemporal synchronization data includes: positioning and mileage information, which is obtained by tagging time and mileage data collected by devices such as GPS and odometers to ensure spatiotemporal synchronization; and speed and operating parameters, which are dynamically adjusted based on the vehicle's operating speed to improve data accuracy.
[0027] S3: Temporarily store the inspection data in flash memory, analyze and determine whether the inspection data is normal; if the inspection data is normal, mark the inspection data as general event inspection data; if the inspection data is abnormal, mark the inspection data as special event inspection data.
[0028] The specific standards for determining whether inspection data is normal are based on existing technology and will not be elaborated upon here. They mainly involve setting safety limits, such as whether changes in track geometry parameters exceed specified safety thresholds. Image processing, such as scratches and weld abnormalities, is based on the detection results of the relevant equipment.
[0029] S4: Reduce the dimensionality of general event inspection data and special event inspection data to one-dimensional data, and store them in the general event data circular storage area and the special event data circular storage area respectively.
[0030] Taking track geometry parameter data as an example, it contains geometric parameter data in four dimensions: track gauge, superelevation, track alignment, and elevation. The track maintenance vehicle-mounted intelligent inspection system parses the track geometry parameter data and then transforms the four-dimensional track geometry parameter data into a one-dimensional structure using a data structure of start symbol, geometric parameter, track gauge + value, superelevation + value, track alignment + value, elevation + value, and end symbol.
[0031] Finally, the geometric parameter data and the flatness data are concatenated to form flatness and geometric parameter data. Then, the flatness and geometric parameter data are concatenated with the dynamic response index data to form track geometric state data. Finally, the track geometric state data is concatenated with track surface and structural state data to form complete one-dimensional data, and event labels are added to the one-dimensional data.
[0032] Then, the one-dimensional general event inspection data is stored in the general event data circular storage area in chronological order, and the one-dimensional special event inspection data is stored in the special event data circular storage area in chronological order.
[0033] S5: Compare the dimensionality-reduced inspection data with historical data. If they are the same, delete the inspection data; if they are different, store the difference between the inspection data and the historical data.
[0034] When storing the dimensionality-reduced inspection data (one-dimensional general event inspection data and special event inspection data) in the on-board inspection system for track maintenance, the system reads the historical data in the basic table storage area and then compares the inspection data with the historical data in sequence.
[0035] For any inspection data item, if the inspection data is the same as the historical data of the corresponding item, then the inspection data is deleted. If they are different, then the difference between the inspection data and the historical data of the corresponding item is calculated, and the difference is stored as the new value of the inspection data in the corresponding position in the basic table storage area.
[0036] Since most inspection data remains unchanged under normal circumstances—for example, the presence of cracks, welds, and the track's geometric parameters remain constant within a certain range—the above processing method can remove items identical to those in the base data table. During data retrieval, it's only necessary to sequentially search for missing items in the inspection data and fill them in with historical data from the base data table. The inspection data values are then summed to restore the true values. This reduces the actual data storage volume of the inspection data.
[0037] Based on this, the data items that were not deleted are mainly the differences between the data in the base table and the data in the base table. These differences are usually within a certain fixed range. Under normal circumstances, the differences in the inspection data are close to the base values, meaning that the amount of data with smaller differences is much larger than the amount of data with larger differences.
[0038] The data storage size is further reduced by the above re-encoding method, as shown in the following steps: S5-1: Sample the historical data (difference) of the inspection data to obtain a set of sampled data; S5-2: Count the number of identical values Si in the sampled data, Pi, to obtain a set of values S={S1, S2, ..., Si, ... Sn}, and the corresponding number P={P1, P2, ..., Pi, ... n}; S5-3: Treat each value in S as a binary tree Ti with only one node, and put all the binary trees Ti into a priority queue, and arrange them in ascending order according to the size of the number Pi corresponding to Ti; S5-4: Take the two binary trees with the smallest number Pi from the priority queue, namely T1 and T2, and their corresponding numbers are P1 and P2; S5-5: Create a new binary tree Tnew, with T1 and T2 as its left and right subtrees. The number of binary trees Tnew is the sum of the number of the two subtrees: Pnew = P1 + P2. S5-6: Put the binary tree Tnew back into the priority queue and return to S5-3; S5-7: Repeat S5-3 to S5-6 until the final merging results in the root node of the binary tree. That is, the priority queue only has two binary trees left. After merging these two binary trees, the root node Troot is obtained. Assign a bit value to each branch of the binary tree (e.g., assign 0 to the left branch and 1 to the right branch). S5-8: The bit sequence along the path from the root node to each leaf node (i.e., Si) is the new storage code corresponding to Si.
[0039] For example Given data S = {A, B, C, D}, and quantities P = {40, 30, 20, 10}; There are four trees in the queue: (A:40), (B:30), (C:20), (D:10); First merge: Take out the two smallest ones (C: 20) and (D: 10) and merge them into a new tree T1. The number is 20 + 10 = 30. The queue becomes (A: 40), (B: 30), (T1: 30). Second merge: Take out the two smallest ones (B:30) and (T1:30) and merge them into a new tree T2. The number is 30+30=60. The queue becomes (A:40) and (T2:60). Second merge: Take out the two smallest ones (A:40) and (T2:60) and merge them into the final tree Troot, with a count of 40 + 60 = 100; Let the first dequeue member be the left subtree and be assigned 0, and the last dequeue member be the right subtree and be assigned 1. Starting from the root node Troot(100): From left to (A:40): If the path is 0, then the data encoding of A is 0; Go right to (T2:60), go left to B (30): the path is 1+0, then the data code of B is 10; To the right to T1 (30): To the left to C (20): The path is 1+1+0, so the data encoding of C is 110; To the right to D(10): the path is 1+1+1, then the data encoding of D is 111.
[0040] By re-encoding the inspection data as described above, the storage code for the largest amount of data is the shortest, thereby reducing the final amount of data stored.
[0041] The above description of the embodiments is only used to provide a detailed introduction to the technical solution of the present invention. However, the description of the above embodiments is only for the purpose of helping to understand the method and core idea of the present invention, and should not be construed as a limitation of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for storing inspection data in a track maintenance vehicle-mounted inspection system, characterized in that, Includes the following steps: S1: Construct a dual-loop inspection data storage structure that includes a general event data loop storage area, a special event data loop storage area, and a basic table storage area. The basic table storage area stores one-dimensional historical data. S2: Real-time collection of inspection data; S3: Temporarily store the inspection data in flash memory, analyze and determine whether the inspection data is normal; If the inspection data is normal, it will be marked as general event inspection data; if the inspection data is abnormal, it will be marked as special event inspection data. S4: Reduce the dimensionality of general event inspection data and special event inspection data to one-dimensional data, and store them in the general event data circular storage area and the special event data circular storage area respectively; S5: Compare the dimensionality-reduced inspection data with historical data. If they are the same, delete the inspection data; if they are different, store the difference between the inspection data and the historical data.
2. The inspection data storage method of a track maintenance vehicle-mounted inspection system as described in claim 1, characterized in that, The inspection data in S2 includes at least one of the following: track geometry data, track surface and structural status data, environmental and ancillary facility data, and spatiotemporal synchronization data.
3. The inspection data storage method of a track maintenance vehicle-mounted inspection system as described in claim 2, characterized in that, The track geometry data includes flatness, track geometry parameters, and dynamic response indicators.
4. The inspection data storage method of a track maintenance vehicle-mounted inspection system as described in claim 2, characterized in that, The track surface and structural condition data include track surface defect data, fastener and sleeper condition data, and track internal damage data.
5. The inspection data storage method of a track maintenance vehicle-mounted inspection system as described in claim 2, characterized in that, The environmental and ancillary facility data includes foreign object intrusion, clearance, and tunnel status.
6. The inspection data storage method of a track maintenance vehicle-mounted inspection system as described in claim 2, characterized in that, The spatiotemporal synchronization data includes positioning and mileage information, speed and operating parameters.
7. A method for storing inspection data in a track maintenance vehicle-mounted inspection system as described in any one of claims 1-6, characterized in that, The differences between the non-simultaneously stored inspection data and historical data in S5 include: S5-1: Sample the historical data of the inspection data to obtain a set of sample data; S5-2: Count the number of identical values Si in the sampled data, Pi, to obtain a set of values S={S1, S2, ..., Si, ... Sn}, and the corresponding number P={P1, P2, ..., Pi, ... n}; S5-3: Treat each value in S as a binary tree Ti with only one node, and put all the binary trees Ti into a priority queue, and arrange them in ascending order according to the size of the number Pi corresponding to Ti; S5-4: Take the two binary trees with the smallest number Pi from the priority queue, namely T1 and T2, and their corresponding numbers are P1 and P2; S5-5: Create a new binary tree Tnew, with T1 and T2 as its left and right subtrees. The number of binary trees Tnew is the sum of the number of the two subtrees: Pnew = P1 + P2. S5-6: Put the binary tree Tnew back into the priority queue and return to S5-3; S5-7: Repeat S5-3 to S5-6 until the root node of the binary tree is finally obtained by merging. That is, the priority queue has only two binary trees left. After merging these two binary trees, the root node Troot is obtained. A bit value is assigned to each branch of the binary tree. S5-8: The bit sequence along the path from the root node to each leaf node is the new storage code corresponding to Si.