ADJUSTING WHEEL DIAMETER

• US 20120136637A1
• Filed: 11/30/2011
• Published: 05/31/2012
• Est. Priority Date: 11/30/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Duan, Ning, Jiang, Zhong Bo, Sun, Wei, Mo, Wen Ting
• Granted Patent: US 8,855,975 B2
• Time in Patent Office: Days
• US Class Current: 703/2
• CPC Class Codes: B61K 9/12 Measuring or surveying wheel-rims measuring techniques G01
• CPC Class Codes: G01B 21/10 for measuring diameters

Abstract

A method and apparatus for adjusting wheel diameter by acquiring a current wheel diameter distribution vector; calculating similarity scores between the current wheel diameter distribution vector and known wheel diameter distribution vector clusters to determine a potential wear rate of the current wheel diameter distribution vector; calculating similarity scores between candidate target wheel diameter distribution vectors and the known wheel diameter distribution vector clusters to determine a potential wear rates of the candidate target wheel diameter distribution vectors; and selecting a target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors, such that the reduction of potential wear rate achieved by switching the current wheel diameter distribution vector to the target wheel diameter distribution vector meets a predetermined condition. Future potential wear rates and the amount of grinding required during train maintenance can be optimized through changing wheel diameter distribution vectors.

Claims

1. A method for adjusting wheel diameter, comprising:

acquiring a current wheel diameter distribution vector;

calculating similarity scores between the current wheel diameter distribution vector and known wheel diameter distribution vector clusters to determine a potential wear rate of the current wheel diameter distribution vector;

calculating similarity scores between candidate target wheel diameter distribution vectors and the known wheel diameter distribution vector clusters to determine potential wear rates of the candidate target wheel diameter distribution vectors; and

selecting a target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors, such that the reduction of the potential wear rate achieved by switching the current wheel diameter distribution vector to the target wheel diameter distribution vector meets a predetermined condition.

2. The method according to claim 1, further comprising:

generating the known wheel diameter distribution vector clusters from known wheel diameter distribution vectors in history records.

3. The method according to claim 2, wherein generating the known wheel diameter distribution vector clusters from the known wheel diameter distribution vectors in the history records comprises:

acquiring a plurality of known wheel diameter distribution vectors and corresponding actual wear rates from the history records;

clustering the plurality of known wheel diameter distribution vectors and corresponding actual wear rates to generate the known wheel diameter distribution vector clusters; and

calculating an average wear rate of each known wheel diameter distribution vector cluster.

4. The method according to claim 1, wherein calculating the similarity scores between the current wheel diameter distribution vector and the known wheel diameter distribution vector clusters comprises:

calculating a weighted-average wheel diameter distribution vector of each known wheel diameter distribution vector cluster;

calculating a distance between the current wheel diameter distribution vector and the average wheel diameter distribution vector, and taking the distance as the similarity score.

5. The method according to claim 1, wherein calculating the similarity scores between the current wheel diameter distribution vector and the known wheel diameter distribution vector clusters comprises:

for a certain known wheel diameter distribution vector cluster, calculating a distance between each known wheel diameter distribution vector included therein and the current wheel diameter distribution vector;

determining the similarity score between the known wheel diameter distribution vector cluster and the current wheel diameter distribution vector according to the number of known wheel diameter distribution vectors within the known wheel diameter distribution vector cluster whose distance to the current wheel diameter distribution vector meets a distance condition.

6. The method according to claim 1, wherein calculating the similarity scores between candidate target wheel diameter distribution vectors and the known wheel diameter distribution vector clusters comprises:

enumerating all candidate target wheel diameter distribution vectors that can be obtained through adjusting an axle sequence of the train;

for each enumerated candidate target wheel diameter distribution vector, calculating the similarity score between it and the known wheel diameter distribution vector clusters.

7. The method according to claim 1, wherein selecting the target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors comprises:

taking a reduction of potential wear rate achieved by switching the current wheel diameter distribution vector to a candidate target wheel diameter distribution vector as a switch benefit of the candidate target wheel diameter distribution vector, and taking a cost required for switching the current wheel diameter distribution vector to the candidate target wheel diameter distribution vector as a switch cost of the candidate target wheel diameter distribution vector, selecting the candidate target wheel diameter distribution vector for which the switch benefit is larger than the switch cost to a greatest extent as the target wheel diameter distribution vector.

8. The method according to claim 1, wherein selecting the target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors comprises:

determining the target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors by solving an optimal problem.

9. The method according to claim 1, wherein acquiring the current wheel diameter distribution vector comprises:

acquiring a current wheel diameter distribution vector of left wheels and a current wheel diameter distribution vector of right wheels; and

combining the current wheel diameter distribution vector of the left wheels and the current wheel diameter distribution vector of the right wheels as the current wheel diameter distribution vector.

10. A computer apparatus for adjusting wheel diameter, comprising:

at least one processing device for implementing components;

acquisition component, configured to acquire a current wheel diameter distribution vector;

first potential wear rate determining component, configured to calculate similarity scores between the current wheel diameter distribution vector and known wheel diameter distribution vector clusters to determine a potential wear rate of the current wheel diameter distribution vector;

second potential wear rate determining component, configured to calculate similarity scores between candidate target wheel diameter distribution vectors and the known wheel diameter distribution vector clusters to determine potential wear rates of the candidate target wheel diameter distribution vectors; and

selection component, configured to select a target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors, such that the reduction of potential wear rate achieved by switching the current wheel diameter distribution vector to the target wheel diameter distribution vector meets a predetermined condition.

11. The apparatus according to claim 10, further comprising:

generation component, configured to generate the known wheel diameter distribution vector clusters from known wheel diameter distribution vectors in history records.

12. The apparatus according to claim 11, wherein the generation component comprises:

an acquiring component configured to acquire a plurality of known wheel diameter distribution vectors and their corresponding actual wear rates from the history records;

a clustering component configured to cluster the plurality of known wheel diameter distribution vectors and corresponding actual wear rates to generate the known wheel diameter distribution vector clusters; and

a calculating component configured to calculate an average wear rate of each known wheel diameter distribution vector cluster.

13. The apparatus according to claim 10, wherein the first potential wear rate determining component comprises:

a weighted average calculation component configured to calculate a weighted-average wheel diameter distribution vector of each known wheel diameter distribution vector cluster;

a distance calculation component configured to calculate the distance between the current wheel diameter distribution vector and the average wheel diameter distribution vector and take the distance as the similarity score.

14. The apparatus according to claim 10, wherein the first potential wear rate determining component comprises:

a distance component configured to, for a certain known wheel diameter distribution vector cluster, calculate the distance between each known wheel diameter distribution vector included therein and the current wheel diameter distribution vector;

a similarity component configured to determine the similarity score between the known wheel diameter distribution vector cluster and the current wheel diameter distribution vector according to the number of known wheel diameter distribution vectors within the known wheel diameter distribution vector cluster, whose distance to the current wheel diameter distribution vector meets a distance condition.

15. The apparatus according to claim 10, wherein the second potential wear rate determining component comprises:

an enumeration component configured to enumerate all candidate target wheel diameter distribution vectors that can be obtained through adjusting axle sequence of the train; and

a similarity score component configured to, for each enumerated candidate target wheel diameter distribution vector, calculate the similarity score between it and the known wheel diameter distribution vector clusters.

16. The apparatus according to claim 10, wherein the selection component comprises:

a candidate selection component configured to, with the reduction of potential wear rate achieved through switching the current wheel diameter distribution vector to a candidate target wheel diameter distribution vector being taken as a switch benefit of the candidate target wheel diameter distribution vector, and with the cost required for switching the current wheel diameter distribution vector to the candidate target wheel diameter distribution vector as a switch cost of the candidate target wheel diameter distribution vector, select the candidate target wheel diameter distribution vector for which the switch benefit is larger than the switch cost to a largest extent as the target wheel diameter distribution vector.

17. The apparatus according to claim 10, wherein the selection component comprises:

a vector determination component configured to determine the target wheel diameter distribution vector from the candidate target wheel diameter distribution vectors through solving an optimal problem.

18. The apparatus according to claim 10, wherein the acquisition component comprises:

a acquiring component configured to acquire a current wheel diameter distribution vector of left wheels and a current wheel diameter distribution vector of right wheels;

a combining component configured to combine the current wheel diameter distribution vector of the left wheels and the current wheel diameter distribution vector of the right wheels as the current wheel diameter distribution vector.