CN112380622A - Vehicle body lightweight parameter optimization method based on cloud computing technology - Google Patents

Abstract

The invention relates to a vehicle body lightweight parameter optimization method based on a cloud computing technology, and belongs to the technical field of automobile research and development and computer simulation. The method comprises the following steps: step 1: selecting variables and analytic variables, and selecting the thickness of the metal plate as the variable by the lightweight target; step 2: extracting samples in a set variable range by using an optimized Latin hypercube method; and step 3: submitting a cloud computing platform solving sample; and 4, step 4: extracting response data and establishing an approximate model; and 5: and defining optimization conditions to carry out optimization calculation. According to the invention, the computing capability and efficiency are greatly improved by configuring the cloud computing platform of the master control service.

Description

Vehicle body lightweight parameter optimization method based on cloud computing technology

Technical Field

The invention relates to a vehicle body lightweight parameter optimization method based on a cloud computing technology, and belongs to the technical field of automobile research and development and computer simulation.

Background

Parameter optimization is a method for achieving a design objective, by parameterizing design variables and the design objective, and adopting an optimization method, the design variables are continuously adjusted, so that the design result is continuously close to a parameterized target value. The general process of parameter optimization is divided into: parameterizing variables, generating a DOE (design of experiment) sample, solving the sample generated in the DOE process, establishing an approximate model and solving an optimization model.

The parameter optimization mainly comprises three processes: DOE, approximate model and optimization, wherein the DOE samples are numerous, the requirements on computing capacity and efficiency are huge, and in the conventional computing platform, after the task is submitted, resource allocation cannot be performed among a plurality of servers, so that the resource utilization rate is low.

In the whole process of parameter optimization, the most critical step is how to establish an approximate model; establishing an accurate approximate model is crucial to iterating to obtain an effective optimization scheme; the more the variable quantity is, the more the requirement on the quantity of the samples is, for a large model with a long single sample calculation period, the calculation force is tested by solving a plurality of DOE samples, the common workstation is difficult to meet the calculation requirement, and the cloud calculation technology can effectively solve the problem of insufficient calculation force, so that the effects of improving the optimization precision and shortening the calculation period are achieved.

Disclosure of Invention

The invention provides a vehicle body lightweight parameter optimization method based on a cloud computing technology, and computing capacity and efficiency are greatly improved by configuring a cloud computing platform of a master control service.

The invention adopts the following technical scheme for solving the technical problems:

a vehicle body lightweight parameter optimization method based on a cloud computing technology comprises the following steps:

step 1: selecting variables and analytic variables, and selecting the thickness of the metal plate as the variable by the lightweight target;

step 2: extracting samples in a set variable range by using an optimized Latin hypercube method;

and step 3: submitting a cloud computing platform solving sample;

and 4, step 4: extracting response data and establishing an approximate model;

and 5: and defining optimization conditions to carry out optimization calculation.

And the analytical variable in the step 1 is a sheet metal thickness parameter extracted by software.

The specific process of step 2 is as follows:

step 2.1, determining the variation range of design variables, wherein the upper limit of the thickness variation of all the metal plates is 2.0 times of the basic thickness, and the lower limit of the thickness variation of all the metal plates is 0.5 times of the basic thickness;

step 2.2, determining the number of samples, selecting the thicknesses of 70 metal plates as variables, and setting the number of generated samples to be 350;

and 2.3, after the variable range and the number of samples are determined, generating and outputting the samples.

The invention has the following beneficial effects:

1. according to the invention, the computing capability and efficiency are greatly improved by configuring the cloud computing platform of the master control service.

2. After the method disclosed by the invention is adopted to greatly improve the calculation efficiency, the number of samples can be greatly improved, and the problem that the high-precision establishment of the approximate model cannot be realized by improving the number of samples due to the limitation of the calculation capacity can be solved.

Drawings

FIG. 1 is a graph of optimized Latin hypercube sampling effect.

Fig. 2 is an architectural logic diagram of a cloud computing platform.

FIG. 3 is a flow chart of a general process of building an approximation model.

Fig. 4 is a flow chart of the method of the present invention.

Detailed Description

The invention is described in further detail below with reference to the accompanying drawings.

A vehicle body lightweight parameter optimization method based on a cloud computing technology is shown in FIG. 4 and comprises the following steps: step 1: selecting variables and analytic variables, wherein the thickness of a metal plate is generally selected as the variable for the lightweight target; step 2: extracting samples in a set variable range by using an optimized Latin hypercube method; and step 3: submitting a cloud computing platform solving sample; and 4, step 4: extracting response data and establishing an approximate model; and 5: and defining optimization conditions to carry out optimization calculation.

Step 1: the specific process of selecting and analyzing the variables is as follows:

step 1.1, the automobile body is formed by splicing hundreds of sheet metal parts with different sizes, and the thickness, the material and the like of the sheet metal parts can be used as variables to be optimized; the principle of selecting the sheet metal as the design variable is to select parts on the main body structure of the vehicle body, and some accessory supports with smaller structures have low cost performance and can avoid selection.

Step 1.2, making the selected thickness attribute of the good metal plate into an independent include file and outputting the file; after software is imported, selecting the thickness parameter of each metal plate as an analysis object to complete analysis; the thickness parameter is extracted by the software as a design variable for subsequent sampling.

Step 2: the optimal latin hypercube sampling method is selected as shown in fig. 1, and the specific process of generating the sample is as follows:

and 2.1, determining the variation range of the design variables, wherein the upper limit of the thickness variation of all the metal plates is 2.0 times of the basic thickness, and the lower limit of the thickness variation of all the metal plates is 0.5 times of the basic thickness.

Step 2.2, determining the number of samples, wherein the number of the samples is determined according to the modeling precision requirement of the radial basis function neural network, generally, the number of variables is more than 5 times, 70 sheet metal thicknesses are selected as variables, and the number of generated samples is at least set to 350.

And 2.3, after the variable range and the number of the samples are determined, generating and outputting the samples for next solving.

And step 3: the concrete process of solving the sample is as follows

The number of samples is large, and a cloud computing platform with a master control service is adopted for computing, so that the computing efficiency is improved; the basic architecture of this computing service is shown in FIG. 2.

The computing task is input to a master control service, the master control service is responsible for searching resources, scheduling and distributing tasks, and the proxy service is responsible for receiving the tasks, starting software computing and feeding back task states;

and 4, step 4: establishing an approximate model;

after extracting the variables and the response data, establishing an approximate model, and the general process of establishing the approximate model is as shown in fig. 3, if the approximate model has enough credibility, the approximate model can be used to develop the optimization design instead of a simulation program.

And 5: defining an optimization problem; design variables are determined, and then constraint conditions and optimization targets need to be determined; the constraint condition is not less than the value of the existing vehicle body rigidity, and the target is defined as the minimum mass.

Claims (3)

1. A vehicle body lightweight parameter optimization method based on a cloud computing technology is characterized by comprising the following steps:

step 1: selecting variables and analytic variables, and selecting the thickness of the metal plate as the variable by the lightweight target;

step 2: extracting samples in a set variable range by using an optimized Latin hypercube method;

and step 3: submitting a cloud computing platform solving sample;

and 4, step 4: extracting response data and establishing an approximate model;

and 5: and defining optimization conditions to carry out optimization calculation.

2. The method for optimizing the vehicle body weight reduction parameters based on the cloud computing technology as claimed in claim 1, wherein the analytical variables in step 1 are sheet metal thickness parameters extracted by software.

3. The vehicle body lightweight parameter optimization method based on the cloud computing technology is characterized in that the specific process of the step 2 is as follows:

step 2.1, determining the variation range of design variables, wherein the upper limit of the thickness variation of all the metal plates is 2.0 times of the basic thickness, and the lower limit of the thickness variation of all the metal plates is 0.5 times of the basic thickness;

step 2.2, determining the number of samples, selecting the thicknesses of 70 metal plates as variables, and setting the number of generated samples to be 350;

and 2.3, after the variable range and the number of samples are determined, generating and outputting the samples.

Images