EXACT PARAMETER SPACE REDUCTION

• US 20150142399A1
• Filed: 12/15/2014
• Published: 05/21/2015
• Est. Priority Date: 11/17/2010
• Status: Abandoned Application

• Current Assignee: Toyota Motor Engineering & Manufacturing North America Incorporated (Toyota Motor Corporation)
• Original Assignee: Toyota Motor Engineering & Manufacturing North America Incorporated (Toyota Motor Corporation)
• Inventors: Postma, Erik Jelle, Shmoylova, Elena, Roche, Austin Duncan, Gerhard, Jrgen
• Application Number:
  US14/570,465
• Publication Number:
  US 20150142399A1
• Time in Patent Office: Days
• US Class Current: 703/2
• CPC Class Codes: G06F 17/13 Differential equations using digital differential analysers G06F7/64
• CPC Class Codes: G06F 30/20 Design optimisation, verification or simulation optimisation, verification or simulation of circuit designs G06F30/30

Abstract

In a computational environment including at least one processor, an example method of reducing the number of symbolic parameters in a model of a physical system includes receiving an initial model such as a system of differential algebraic equations (DAEs), eliminating isolated symbolic parameters (if any) from the initial model, extracting parameter sub-expressions from the DAEs, establishing minimal disconnected clusters of parameter sub-expressions, and for each cluster, attempting to generate a reduced cluster having a reduced number of symbolic parameters using one or more algorithms. If more than one approach is successful, that which is most successful in reducing the number of symbolic parameters is selected. A revised model is created having fewer symbolic parameters than the initial model and based on simulation results obtained from the revised model, a prototype of a physical system is manufactured and/or changes are made to an existing physical system.

Claims

1. A method for designing and manufacturing a prototype of a physical system or improving an existing physical system, the method comprising:

providing an initial model of the physical system, the initial model having a plurality of differential algebraic equations (DAEs), the plurality of DAEs having a plurality of variables and a plurality of symbolic parameters;

receiving the initial model of the physical system within a specifically configured computational environment having at least one processor, the specifically configured computational environment executing the following steps;

a. extracting symbolic parameter sub-expressions from the plurality of DAEs;

b. establishing at least one initial cluster of symbolic parameter sub-expressions;

c. executing a parameter reduction algorithm on the symbolic parameter sub-expressions of the at least one initial cluster;

d. generating at least one reduced cluster having a reduced number of symbolic parameter sub-expressions compared with the at least one initial cluster using the parameter reduction algorithm, the reduced number of symbolic parameter sub-expressions being a combination of the symbolic parameter sub-expressions of the at least one initial cluster;

e. creating a revised model using the at least one reduced cluster, the revised model having fewer symbolic parameter sub-expressions than the initial model;

f. modeling the physical system with the revised model having fewer parameter sub-expressions, the modeling of the physical system a having a reduction in computation time compared to modeling of the physical system using the initial model of the physical system;

and

manufacturing a prototype of the physical system or an improved version of an existing physical model as a function of results obtained from the modeling of the physical system with the revised model.

2. The method of claim 1, wherein the at least one initial cluster is a plurality of initial clusters with each initial cluster having a plurality of initial symbolic parameters sub-expressions.

3. The method of claim 2, wherein the exact parameter reduction algorithm is executed on each of the plurality of initial clusters.

4. The method of claim 3, wherein the parameter reduction algorithm is selected from the group consisting of a linear decomposition algorithm, a uni-multivariate polynomial decomposition algorithm and a heuristical decomposition algorithm.

5. The method of claim 4, wherein the parameter reduction algorithm is the linear decomposition algorithm.

6. The method of claim 4, wherein the parameter reduction algorithm is the uni-multivariate decomposition algorithm.

7. The method of claim 4, wherein the parameter reduction algorithm is the heuristical decomposition algorithm.

9. The method of claim 1, further including removing all isolated symbolic parameters from the model before establishing initial clusters of parameter sub-expressions.

8. The method of 3, wherein the parameter reduction algorithm includes a linear decomposition algorithm, a uni-multivariate polynomial decomposition algorithm and a heuristical decomposition algorithm;

executing the linear decomposition algorithm, the uni-multivariate polynomial decomposition algorithm and the heuristical decomposition algorithm on each of the plurality of initial clusters, each of the executed algorithms generating a separate cluster of symbolic parameter sub-expressions; and

selecting at least one reduced cluster from the separate clusters of symbolic parameter sub-expressions, the selected at least one reduced cluster having a minimum number of symbolic parameter sub-expressions.