Optimizing the design of physical structures/objects

Patent metadata

Field	Value
Publication number	US-10853528-B2
Application number	US-201414162159-A
Country	US
Kind code	B2
Filing date	Jan 23, 2014
Priority date	Jan 30, 2013
Publication date	Dec 1, 2020
Grant date	Dec 1, 2020

Abstract

In one aspect, a computer-assisted method for the optimization of the design of physical bodies, such as land, air and sea vehicles and robots and/or parts thereof, is provided comprising the steps of: representing the design to be optimized as a mesh, generating update signals to optimize the mesh representation, applying an optimization algorithm until a stop criterion has been reached, and outputting a signal representing the optimized design.

First claim

The invention claimed is:

1. A computer-assisted method for the optimization of the topology of a physical object resulting in an optimized distribution of material with respect to at least one of an aerodynamic, thermodynamic, crashworthiness, and mechanical compliance quality function and constraints relating to the physical properties of the physical object, wherein the physical object is at least one of a land vehicle, an air vehicle, a sea vehicle, a robot, or at least one of a part or design thereof, the method comprising the steps of:
obtaining design parameters of the physical object or a design of the physical object;
representing the design parameters in a design space;
performing a topology optimization of the design parameters until a stop criterion is met,
wherein the topology optimization comprises an iterative local redistribution of material of the physical object,
wherein the iterative local redistribution of material is performed using update strategies for the local redistribution of the local material based on local information at a location of the redistribution of a structure of the physical object,
wherein the update strategies define a functional mapping from the local information to an update signal, and the local information refers to properties of finite elements within a predetermined distance of the location in the design space depending on the design variable, and
wherein in response to the structure of the physical object at a time when the stop criterion is met not meeting a convergence criterion, updating the update strategies at the location of the local redistribution based on the local information including at least one of thermodynamic properties, energy absorption and position in an overall physical structure as an input for each finite element of the topology optimization separately, wherein the updating is performed without using gradient based sensitivity information, and repeating the performing of the topology optimization of the design parameters on the basis of the updated update strategies of another location of the local redistribution location; and
outputting a signal representing the optimized design parameters to a processor when the convergence criterion is met, wherein the processor is used to manufacture the physical object including the optimized distribution of materials based on the optimized design parameters.

2. The method of claim 1 ,
wherein the further optimization comprises a dual process comprising a first optimization process, wherein the first optimization process comprises at least one of machine learning or stochastic optimization, and providing update strategies to a second optimization process,
wherein the second optimization process comprises a topology optimization, which is initialized by the first optimization process.

3. The method of claim 2 ,
wherein the first optimization process starts with an initial update strategy, an initial set of the update strategies, or a set of the update strategies which can be chosen randomly.

4. The method of claim 3 ,
wherein the initial update strategies are supplied to and used in the second optimization process for the iterative redistribution of material.

5. The method of claim 2 , wherein the second optimization process is started on at least one of an initial physical object or an initial design of a physical object.

6. The method of claim 2 ,
wherein the topology optimization is started on at least one of the physical object or a design of the physical object utilizing the update strategies generated by the first optimization process.

7. The method of claim 2 ,
wherein the second optimization process is started on at least one of the physical object or a design of the physical object obtained from a preceding run of the second optimization process.

8. The method of claim 1 , further comprising:
performing a structural analysis using at least a finite element analysis on the optimized physical object.

9. The method of claim 2 ,
wherein the second optimization process stops when the stop criterion is met.

10. The method of claim 2 ,
wherein a quality determined according to at least one of a quality function or physical properties that are needed to evaluate a value of the quality of a resulting optimized physical object resulting from the second optimization process is returned to the first optimization process.

11. The method of claim 2 ,
wherein a performance of the evaluated physical object is evaluated based on at least one of a quality function or predetermined constraints.

12. The method of claim 2 ,
wherein if the optimized physical object satisfies boundary conditions, wherein the boundary conditions comprise an optimization objective, the second optimization process is stopped.

13. The method of claim 2 ,
wherein the results of an evaluation of the optimized physical object are used as feedback to evaluate a quality of a set of the update strategies provided to evaluate a quality of a current update strategies using improvement of the design during the second optimization process.

14. The method of claim 2 ,
wherein an improved set of update strategies is generated or a previous set of the update strategies is updated using at least one of the further optimization or a learning scheme.

15. The method of claim 1 ,
wherein the update strategies are generated by an automatic process that is utilized for adaptation of the design parameters representing physical properties of a structure of the physical object.

16. The method of claim 1 ,
wherein the update strategies are generated by an optimization or learning procedure that is utilized for adaptation of the design parameters representing physical properties of the structure.

17. The method of claim 1 ,
wherein the design parameters represent an amount of material within the structure.

18. The method of claim 1 ,
wherein final resulting update strategies are utilized for same or similar objective functions and constraints as has been used in the further optimization of the update strategies, in particular with different boundary conditions of the design space.

19. The method of claim 1 ,
wherein a desired mass of the physical object and/or a mass computed from the design of a physical object is fixed or not fixed initially.

20. The method of claim 1 ,
wherein the physical object is optimized in view of at least one of thermodynamic, aerodynamic, or hydrodynamic parameters and at least one of weight characteristic or mechanical characteristic.

21. A apparatus for optimization of a design of physical object resulting in an optimized distribution of material optimized with respect to at least one of an aerodynamic, thermodynamic, crashworthiness, and mechanical compliance quality function and constraints relating to the physical properties of the physical object, wherein the physical object is at least one of a land vehicle, an air vehicle, a sea vehicle, a robot, or at least one of a part or design thereof, the apparatus comprising a processor, wherein the processor is configured to at least:
obtain design parameters of the physical object or a design of the physical object,
represent the design parameters in a design space,
perform a topology optimization of the design parameters until a stop criterion is met, wherein the topology optimization comprises an iterative local redistribution of material of the physical object,
wherein the iteratively redistributing material of the ph

Assignees

• Honda Res Inst Europe Gmbh

Inventors

• Olhofer Markus
• Aulig Nikola

Classifications

• G06F30/23 Primary

  using finite element methods [FEM] or finite difference methods [FDM] · CPC title

• G06F30/00 Primary

  Computer-aided design [CAD] · CPC title

• G06F2111/08

  Probabilistic or stochastic CAD · CPC title

• G06F2111/06

  Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA] · CPC title

• G06F30/15

  Vehicle, aircraft or watercraft design · CPC title