CN106096086B - Method for converting multipoint optimization into multi-objective optimization - Google Patents
Method for converting multipoint optimization into multi-objective optimization Download PDFInfo
- Publication number
- CN106096086B CN106096086B CN201610373357.3A CN201610373357A CN106096086B CN 106096086 B CN106096086 B CN 106096086B CN 201610373357 A CN201610373357 A CN 201610373357A CN 106096086 B CN106096086 B CN 106096086B
- Authority
- CN
- China
- Prior art keywords
- optimization
- converting
- working conditions
- objective
- solver
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
Abstract
The invention discloses a method for converting multi-point optimization into multi-objective optimization aiming at a leaf fan, which comprises the following steps: (1) generating a set of design variables x based on the selected optimization strategy; (2) calling a solver to obtain a group of solution values according to a plurality of set working conditions in sequence; (3) converting multiple points into multiple targets, processing the solving values of the weighted mean target and the weighted standard deviation target, and converting the solving values into final objective function values fx; (4) judging whether a termination condition is met, and if the termination condition is met, returning a preferred solution set; if not, the steps are sequentially executed from the step 1, multi-point optimization is converted into a multi-objective optimization method, multi-objective optimization of a weighted average value objective and a weighted standard deviation objective under a group of working conditions is considered, more reliable design optimization can be performed on a performance curve or under the condition of multiple working conditions, the obtained scheme or design is more reliable, the product performance is effectively improved, and the loss is reduced.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method for converting multipoint optimization into multi-objective optimization.
Background
In life, many problems are composed of multiple targets which conflict with each other and influence each other, and people often encounter an optimization problem which is used in a given area and is as optimal as possible, namely a multi-target optimization problem, and when the optimization targets of the optimization problem exceed one and need to be processed simultaneously, the problem becomes the multi-target problem. In practice, the optimization problem is mostly a multi-objective optimization problem, and in general, there are contradictions between the sub-objectives of the multi-objective optimization problem, and it is likely that an improvement of one sub-objective will cause a performance reduction of one or more other sub-objectives, that is, it is unlikely that multiple sub-objectives will reach the optimal value together, and only coordination and such processing among them will be performed to optimize each sub-objective as much as possible.
The multi-point optimization technology, namely the optimization of a certain performance curve, means that a certain group of design variables are optimized, so that a satisfactory effect can be achieved under different working conditions. For example, the performance optimization problem of a hydraulic torque converter needs a set of design variables, such as the blade wrap angle of a pump and a turbine, a metal angle and the like, so that under the condition of different rotation speed ratios, the volume coefficient and the torque ratio achieve relatively good effects, and the problem with practical requirements is the multi-point optimization problem.
At present, a method for converting multiple targets into a single target by adopting a weighting coefficient is generally adopted, such as the optimization of a certain characteristic curve:
where wi is the weighting coefficient. Knowing a certain design variable
Solving for a1,a2,…,anN Performance parameters under different operating conditions
(taking the example of minimizing the loss factor here). Due to a cluster of design variables
Loss coefficients under multiple working conditions need to be minimized at the same time, and the processing method inevitably sacrifices margin information of a performance curve, so that the variable working condition margin cannot be widened while the multi-point performance parameters are maintained.
At present, a mechanism for optimizing by means of program automation processing, which converts multipoint optimization into multi-objective optimization, is still blank, and solving the problem with practical requirements becomes more important.
Disclosure of Invention
The invention mainly aims to overcome the difficulties and provide a method for converting multi-point optimization into multi-objective optimization, namely, an optimization method for meeting requirements is obtained by optimizing the problem of performance curve optimization by combining an optimization strategy of automatic optimization and optimization parameter self-adaption, a more reliable scheme or design is obtained, the product performance is effectively improved, and the loss is reduced.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows: a method for converting multipoint optimization into multi-objective optimization comprises the following steps: (1) generating a set of design variables x based on the selected optimization strategy; (2) calling a solver to obtain a group of solution values in sequence according to a plurality of set working conditions; (3) converting the multiple points into multiple targets, processing the solving values of the weighted mean target and the weighted standard deviation target, and converting into a final objective function value fx,
the weighted mean target definition formula is:
the weighted standard deviation target definition formula is as follows:
wherein, wiIs a weighting coefficient under different working conditions, f (x, a)i) The objective function values under different working conditions in the optimization problem, and n is the number of the working conditions;
(4) judging whether a termination condition is met, and if the termination condition is met, returning a preferred solution set; if not, the steps (1) are executed in sequence.
In the above scheme, preferably, the solver is CAD software for geometric modeling, grid generation software for generating a grid, and CAE software is combined or is a display function expression.
In the above-described aspect, the CAE software is preferably CFD software for fluid analysis or other analysis software such as thermal, force, and electromagnetic software.
In the optimization strategy, for a specific optimization method, a corresponding strategy obtains a group of design variables x in the design variable space, which are determined according to the specific optimization method; setting working conditions according to actual problems, and under different working conditions, corresponding cae (solver) execution scripts, wherein the specific solution is only to introduce design variables into specific cae (solver) scripts to realize calculation; and processing the weighted mean target and the weighted mean target solution value to convert the weighted mean target and the weighted mean target solution value into an actual objective function value fx, judging whether the termination condition is that whether the calculation result is Frontier and whether Frontier meets the customer requirements are judged, if yes, reaching the termination condition, and returning an optimal solution set which is the previously calculated design variable x and the corresponding fx. Compared with the prior art, the invention has the following beneficial effects: by converting multi-point optimization into multi-target optimization and observing the multi-target optimization of a weighted average value target and a weighted standard deviation target under a group of working conditions, the design optimization of a performance curve or under the condition of multiple working conditions can be more reliable, the obtained scheme or design is more reliable, the performance of a product is effectively improved, and the loss is reduced.
Detailed Description
The following description is presented to disclose the invention so as to enable any person skilled in the art to practice the invention. The preferred embodiments in the following description are given by way of example only, and other obvious variations will occur to those skilled in the art.
The fan blade optimization is taken as an example, the design parameters of the fan blade are optimized, and the design variables are the values of parameters such as the twist degree of the blade, the radius of an air channel, the hub ratio, the sweepback degree of the blade, the number of the blade and the circumferential camber of the blade, so that the fan blade can achieve lower noise and higher efficiency under the condition of simultaneously meeting two working conditions of working condition 1 and working condition 2.
A method for converting multipoint optimization into multi-objective optimization comprises the following steps: (1) generating a group of design variables x based on a selected optimization strategy, wherein the x represents the value taking situation of variables such as the twist degree of the blades, the radius of an air duct, the hub ratio, the sweep degree of the blades, the number of the blades, the circumferential camber of the blades and the like under a certain working condition; (2) calling a solver to obtain a group of solution values according to a plurality of set working conditions in sequence, setting two working conditions, and under the condition of the working condition 1, based on the design variable x generated in the step 11Invoking solver and returning the calculation result
Under the condition of working condition 2, based on the design variable x generated in the step 12Invoking solver and returning the calculation result
(3) Converting the multiple points into multiple targets, processing the solving values of the weighted mean target and the weighted standard deviation target, and converting the solving values into actual valuesThe objective function value fx, where,
the weighted mean target definition formula is:
the weighted mean standard deviation target definition formula is as follows:
wherein, wiIs a weighting coefficient under different working conditions, f (x, a)i) Is the objective function value under different working conditions in the optimization problem, n is the number of working conditions, and is obtained by the design variable x returned by the summary solver
And
then applying the above formula F1And F2Calculating to obtain a target value fx;
(4) judging whether a termination condition is met, and if the termination condition is met, returning a preferred solution set; if the parameters do not meet the requirements, the steps (1) are sequentially executed, namely the termination conditions are blade twist, air channel radius, hub ratio, blade sweep, the number of blades, the circumferential camber of the blades and other parameters can achieve lower noise and higher efficiency under two working conditions;
the solver is CAD software for geometric modeling, grid generation software for generating grids, and CAE software combined or displaying a function expression.
The CAE software may be CFD software for fluid analysis or other analysis-like software such as thermal, force, electromagnetic, etc.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (3)
1. A method for converting multipoint optimization into multiobjective optimization, the method comprising the steps of:
(1) generating a group of design variables x based on a selected optimization strategy, wherein the x represents the twist degree of the blades, the radius of an air duct, the hub ratio, the sweep degree of the blades, the number of the blades and the value taking situation of the circumferential camber variable of the blades under a certain working condition;
(2) calling a solver to obtain a group of solution values according to a plurality of set working conditions in sequence, setting two working conditions, and under the working condition 1, based on the design variable x generated in the step 11Invoking solver and returning the calculation result
Under the condition of working condition 2, based on the design variable x generated in the step 12Invoking solver and returning the calculation result
(3) Converting multiple points into multiple targets, processing the solving values of the weighted mean target and the weighted standard deviation target, and converting into actual objective function values fx,
the weighted mean target definition formula is:
the weighted standard deviation target definition formula is as follows:
wherein, wiIs a weighting coefficient under different working conditions,
is the objective function value under different working conditions in the optimization problem, n is the number of working conditions, and is obtained by the design variable x returned by the summary solver
And
then applying the above formula F1And F2Calculating to obtain a target value fx;
(4) judging whether a termination condition is met, and if the termination condition is met, returning a preferred solution set; if the parameters do not meet the requirements, the steps (1) are sequentially executed, namely the termination conditions are blade twist, air channel radius, hub ratio, blade sweep, the number of blades and the circumferential camber parameters of the blades can achieve lower noise and higher efficiency under two working conditions.
2. The method for converting multipoint optimization into multi-objective optimization as claimed in claim 1, wherein the solver is CAD software for geometric modeling, grid generation software for generating grids, CAE software combined or displaying function expressions.
3. The method for converting multi-point optimization into multi-objective optimization of claim 2, wherein the CAE software is CFD software for fluid analysis or other analysis software such as thermal, force, electromagnetic.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610373357.3A CN106096086B (en) | 2016-05-30 | 2016-05-30 | Method for converting multipoint optimization into multi-objective optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610373357.3A CN106096086B (en) | 2016-05-30 | 2016-05-30 | Method for converting multipoint optimization into multi-objective optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106096086A CN106096086A (en) | 2016-11-09 |
| CN106096086B true CN106096086B (en) | 2020-08-14 |
Family
ID=57229431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610373357.3A Active CN106096086B (en) | 2016-05-30 | 2016-05-30 | Method for converting multipoint optimization into multi-objective optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106096086B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063372A (en) * | 2018-08-27 | 2018-12-21 | 山东理工大学 | A kind of wheel hub driving electric car suspension system and damping element parameter design method for optimization of matching |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102968568A (en) * | 2012-11-30 | 2013-03-13 | 湖南大学 | Reverse determination method of material parameters of high-strength steel in multi-working-condition mode |
| CN103136428A (en) * | 2013-03-12 | 2013-06-05 | 上海交通大学 | Vehicle body structure steady design method based two uncertain saloon cars |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8560472B2 (en) * | 2010-09-30 | 2013-10-15 | The Aerospace Corporation | Systems and methods for supporting restricted search in high-dimensional spaces |
-
2016
- 2016-05-30 CN CN201610373357.3A patent/CN106096086B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102968568A (en) * | 2012-11-30 | 2013-03-13 | 湖南大学 | Reverse determination method of material parameters of high-strength steel in multi-working-condition mode |
| CN103136428A (en) * | 2013-03-12 | 2013-06-05 | 上海交通大学 | Vehicle body structure steady design method based two uncertain saloon cars |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106096086A (en) | 2016-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106650125B (en) | Optimization method and system for centrifugal compressor impeller | |
| CN113673129B (en) | Multi-objective robust optimization design method for ultra-efficient induction motor | |
| CN110378016B (en) | Multi-objective optimization design method for pump impeller adjustable hydraulic torque converter | |
| CN103514308B (en) | The method and apparatus of design blade of wind-driven generator | |
| CN106096086B (en) | Method for converting multipoint optimization into multi-objective optimization | |
| KR102109493B1 (en) | Method for building fatigue load models for wind turbine | |
| CN114091377B (en) | Method for calculating wake flow wind speed of dynamic double-Gaussian wind turbine based on spatial variation | |
| CN105465037A (en) | Hydraulic power optimizing method and device for double-suction centrifugal pump impeller | |
| CN112861362B (en) | Power assembly performance parameter optimization method and device based on vehicle oil consumption | |
| CN102682348A (en) | Complex equipment component maintenance level optimization system and establishing method of thereof | |
| CN109409019A (en) | A kind of pre-bending optimization method of pneumatic equipment bladess aerodynamic configuration | |
| CN110397553B (en) | Model-free wind power plant wake flow management method and system | |
| Li et al. | Overall design optimization of dedicated outboard airfoils for horizontal axis wind turbine blades | |
| CN111461443A (en) | Optimal design method and system for wind power plant, electronic device and storage medium | |
| CN109540459B (en) | Pneumatic characteristic numerical calculation result correction method | |
| TW201943951A (en) | A method and a system for designing a foundation for a wind turbine | |
| DK201870706A1 (en) | A dynamic optimisation strategy for improving the operation of a wind farm | |
| CN109472389B (en) | Multi-working-condition intelligent optimization method for impeller with medium and low specific speed | |
| Safari et al. | A novel combination of adaptive tools for turbomachinery airfoil shape optimization using a real-coded genetic algorithm | |
| CN112366681A (en) | Power system random dynamic model linearization threshold condition and stability analysis method | |
| Zhang et al. | Investigation on multi-objective performance optimization algorithm application of fan based on response surface method and entropy method | |
| CN114528749A (en) | Model determination method, model determination device, electronic equipment and storage medium | |
| CN113294297B (en) | Variable weight adjusting method for wind turbine generator nonlinear model prediction torque control | |
| CN106934124B (en) | Adaptive variable window method based on measurement change detection | |
| CN115577453A (en) | Variable pitch propeller pneumatic design method, device and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211206 Address after: 211106 floor 11, building B1, No. 19, Suyuan Avenue, Jiangning District, Nanjing City, Jiangsu Province Patentee after: NANJING TIANFU SOFTWARE Co.,Ltd. Address before: 315000 building 006 (19-1), No. 375, Jiangdong North Road, Ningbo, Zhejiang Patentee before: NINGBO HIPOINT INDUSTRIAL DESIGN CO.,LTD. |
|
| TR01 | Transfer of patent right |