Vehicle dynamics production system and method

• US 20040064235A1
• Filed: 10/28/2003
• Published: 04/01/2004
• Est. Priority Date: 12/20/2000
• Status: Active Grant

• Current Assignee: Central Queensland University, Queensland Rail
• Original Assignee: Central Queensland University, Queensland Rail
• Inventors: Cole, Colin Robert
• Granted Patent: US 6,853,889 B2
• Time in Patent Office: Days
• US Class Current: 701/70
• CPC Class Codes: B60T 8/172 Determining control parameters used in the regulation, e.g. by calculations involving measured or detected parameters B60T8/17551 takes precedence
• CPC Class Codes: B60T 8/174 characterised by using special control logic, e.g. fuzzy logic , neural computing
• CPC Class Codes: B60W 2556/50 of positioning data, e.g. GPS [Global Positioning System] data
• CPC Class Codes: B60W 50/0097 Predicting future conditions
• CPC Class Codes: B61L 15/0058 On-board optimisation of vehicle or vehicle train operation
• CPC Class Codes: B61L 15/0092 Memory means reproducing during the running of the vehicle or vehicle train, e.g. smart cards
• CPC Class Codes: B61L 2205/04 Satellite based navigation systems, e.g. global positioning system [GPS]
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 13/048 using a predictor
• CPC Class Codes: Y10S 706/905 Vehicle or aerospace

Abstract

A vehicle dynamics prediction system for providing prediction of vehicle velocity for a predetermined future time period. The predictions utilise future vehicle control setting anticipated for the time period. An artificial intelligence database coupled to a processor that utilises weighted values for neural network models representing dynamic performance of the vehicle forms part of the vehicle dynamics prediction system.

Claims

1. A vehicle dynamics prediction system for providing predictions, over a predetermined future time period, of vehicle velocity, which predictions utilize future vehicle control settings anticipated for that period, said prediction system including:

(a) a processing means coupled to a user interface for facilitating selection of future vehicle control settings and selection of a route for travel of the vehicle;

(b) a first input arrangement for inputting data about current vehicle position to the processing means;

(c) a second input arrangement for inputting data about current vehicle control settings and data about current vehicle operational parameters to the processing means;

(d) a first memory arrangement coupled to the first input arrangement for storing historical vehicle position data and for buffering current vehicle position data prior to storage;

(e) a second memory arrangement coupled to the second input arrangement for storing historical control setting data and historical operational parameter data and for buffering current control setting data and current operational parameter data prior to storage;

(f) an artificial intelligence database coupled to the processing means, containing a plurality of weighting values for neural network models representing dynamic performance of respective units comprising the vehicle;

(g) a route topographical database coupled to the processing means, containing position data about available routes of travel for the vehicle;

wherein the processing means calculates future conditions of the vehicle based on the current vehicle position data relative to the selected route utilizing position data obtained from the route topographical database, and predicts the vehicle velocity in the vehicle during said predetermined period by processing the vehicle control setting data and operational parameter data through the neural network models, which models employ the weighting values from the artificial intelligence database, for calculated future vehicle positions and associated future control settings.

2. The vehicle dynamics prediction system of claim 1 wherein the system also provides predictions of longitudinal forces present in the vehicle, wherein the processing means uses the vehicle velocity predictions combined with future control data to further predict longitudinal forces for at least one selected location in the length of the vehicle during said predetermined period.

3. The vehicle dynamics predication system of claim 1 wherein the vehicle control settings include throttle and brake settings.

4. The vehicle dynamics prediction system of claim 1 wherein the vehicle operational parameters include the vehicle velocity and a force between at least two inter-coupled units of the vehicle.

5. The vehicle dynamic prediction system of claim 4 wherein the vehicle operational parameters further include a force between a motive power unit and a plurality of trailed units.

6. The vehicle dynamics prediction system of claim 1 used on vehicles in the form a train consisting of at least locomotive unit and a plurality of wagon units and/or carriage units.

7. The vehicle dynamics prediction system of claim 1 used on vehicles in the form a multiple trailer road vehicle comprising a prime mover or tractor and at least three trailer units.

8. A method of predicting, over a predetermined future time period, vehicle velocity which predictions utilize future vehicle control settings anticipated for that period, said prediction method including the steps of:

(a) selecting future control settings and a desired route for travel of the vehicle;

(b) inputting data about current vehicle position to a processing means;

(c) inputting data about current vehicle control settings and data about current vehicle operational parameters to the processing means;

(d) storing in a first memory arrangement coupled to the first input arrangement, historical vehicle position data and buffering current vehicle position data prior to storage;

(e) storing in a second memory arrangement coupled to the second input arrangement, historical control setting data and historical operational parameter data and buffering current control setting data and current operational parameter data prior to storage;

(f) providing an artificial intelligence database coupled to the processing means, containing a plurality of weighting values for neural network models representing dynamic performance of respective units comprising the vehicle;

(g) providing a route topographical database coupled to the processing means, containing position data about available routes of travel for the vehicle;

(h) calculating future conditions of the vehicle based on the current vehicle position data relative to the desired route and position data obtained from the route topographical database;

(i) processing the vehicle control setting data and operational parameter data through at least one neural network model, which models employ the weighting values from the artificial intelligence database, for calculated future positions and associated future control settings to provide predictions of the vehicle velocity during said predetermined period.

9. The method of claim 8 that includes the step of providing predictions of longitudinal forces present in the vehicle, wherein the processing means uses the vehicle velocity predictions combined with future control data to further predict longitudinal forces for at least one selected location in the length of the vehicle during said predetermined period.

10. The method of claim 9 that includes the step of providing predictions of the forces between a motive power unit, and trailed units in order to predict both force input into the trailed units by the motive power unit and force input into the motive power unit by the trailed units.

11. The method of claim 10 that includes the step of providing a prediction of energy transfer occurring in the vehicle during the predetermined period may be predicted from the following formula, using predictions of forces between the motive power unit and trailed units and predictions of vehicle velocity:

12. The method of claim 9 wherein the predictions of longitudinal forces are provided by a separate neural network model.

13. The method of claim 9 wherein the predictions of longitudinal forces are provided by a single integrated neural network model.