MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK
In order to accurate reflect the highly non-linear relationship between structural parameter uncertainty of frame and its responses, and promote reliability design level of frame. A multi-objective reliability optimization design method based on GA weighted network is proposed. Firstly, establish th...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Journal of Mechanical Strength
2022-01-01
|
| Series: | Jixie qiangdu |
| Subjects: | |
| Online Access: | http://www.jxqd.net.cn/thesisDetails#10.16579/j.issn.1001.9669.2022.04.015 |
| _version_ | 1797768036558569472 |
|---|---|
| author | WU GaoFeng WU YaWen ZHI PengPeng LI YunQuan |
| author_facet | WU GaoFeng WU YaWen ZHI PengPeng LI YunQuan |
| author_sort | WU GaoFeng |
| collection | DOAJ |
| description | In order to accurate reflect the highly non-linear relationship between structural parameter uncertainty of frame and its responses, and promote reliability design level of frame. A multi-objective reliability optimization design method based on GA weighted network is proposed. Firstly, establish the parametric analysis model of the frame and carry out the experimental design. After that, the sample data of frame responses and uncertain parameters are obtained. Secondly, the initial BP neural network between responses and uncertain parameters is established. Furthermore, using mathematical model to express BP neural network. On this basis, four kinds of weight coefficients are introduced. Then the GA method is used to find optimal combination mode of weight coefficients, so as to improve the accuracy of the initial BP neural network model. Moreover, make the internal relationship between frame responses and uncertain parameters more accurate. Finally, taking the multi index failure probability of the frame as the objective function, the optimization model is constructed. And the multi-objective reliability optimization design of frame is carried out. The results show that there is a better uniformity of prediction results between improved model and test sample. The absolute error of maximum stress response is less than 1 MPa of them under standard working condition. The new frame parameter combination which is obtained by this method not only improve the frame comprehensive reliability, but also provides a reference for the adjustment and optimization of its design parameters. |
| first_indexed | 2024-03-12T20:48:41Z |
| format | Article |
| id | doaj.art-f58c95f150314b4b852cf6f0c1e43cb1 |
| institution | Directory Open Access Journal |
| issn | 1001-9669 |
| language | zho |
| last_indexed | 2024-03-12T20:48:41Z |
| publishDate | 2022-01-01 |
| publisher | Editorial Office of Journal of Mechanical Strength |
| record_format | Article |
| series | Jixie qiangdu |
| spelling | doaj.art-f58c95f150314b4b852cf6f0c1e43cb12023-08-01T07:39:12ZzhoEditorial Office of Journal of Mechanical StrengthJixie qiangdu1001-96692022-01-014485986629913954MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORKWU GaoFengWU YaWenZHI PengPengLI YunQuanIn order to accurate reflect the highly non-linear relationship between structural parameter uncertainty of frame and its responses, and promote reliability design level of frame. A multi-objective reliability optimization design method based on GA weighted network is proposed. Firstly, establish the parametric analysis model of the frame and carry out the experimental design. After that, the sample data of frame responses and uncertain parameters are obtained. Secondly, the initial BP neural network between responses and uncertain parameters is established. Furthermore, using mathematical model to express BP neural network. On this basis, four kinds of weight coefficients are introduced. Then the GA method is used to find optimal combination mode of weight coefficients, so as to improve the accuracy of the initial BP neural network model. Moreover, make the internal relationship between frame responses and uncertain parameters more accurate. Finally, taking the multi index failure probability of the frame as the objective function, the optimization model is constructed. And the multi-objective reliability optimization design of frame is carried out. The results show that there is a better uniformity of prediction results between improved model and test sample. The absolute error of maximum stress response is less than 1 MPa of them under standard working condition. The new frame parameter combination which is obtained by this method not only improve the frame comprehensive reliability, but also provides a reference for the adjustment and optimization of its design parameters.http://www.jxqd.net.cn/thesisDetails#10.16579/j.issn.1001.9669.2022.04.015Bogie frame;Multi-objective optimization;Reliability;Genetic algorithm;BP neural network |
| spellingShingle | WU GaoFeng WU YaWen ZHI PengPeng LI YunQuan MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK Jixie qiangdu Bogie frame;Multi-objective optimization;Reliability;Genetic algorithm;BP neural network |
| title | MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK |
| title_full | MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK |
| title_fullStr | MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK |
| title_full_unstemmed | MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK |
| title_short | MULTI-OBJECTIVE RELIABILITY OPTIMIZATION DESIGN OF BOGIE FRAME BASED ON GA WEIGHTED NETWORK |
| title_sort | multi objective reliability optimization design of bogie frame based on ga weighted network |
| topic | Bogie frame;Multi-objective optimization;Reliability;Genetic algorithm;BP neural network |
| url | http://www.jxqd.net.cn/thesisDetails#10.16579/j.issn.1001.9669.2022.04.015 |
| work_keys_str_mv | AT wugaofeng multiobjectivereliabilityoptimizationdesignofbogieframebasedongaweightednetwork AT wuyawen multiobjectivereliabilityoptimizationdesignofbogieframebasedongaweightednetwork AT zhipengpeng multiobjectivereliabilityoptimizationdesignofbogieframebasedongaweightednetwork AT liyunquan multiobjectivereliabilityoptimizationdesignofbogieframebasedongaweightednetwork |