Generative Design by Reinforcement Learning: Enhancing the Diversity of Topology Optimization Designs
نویسندگان
چکیده
Generative design refers to computational methods that can automatically conduct exploration under constraints defined by designers. Among many approaches, topology optimization-based generative designs aim explore diverse designs, which cannot be represented conventional parametric approaches. Recently, data-driven optimization research has started exploit artificial intelligence, such as deep learning or machine learning, improve the capability of exploration. This study proposes a reinforcement (RL) based process, with reward functions maximizing diversity designs. We formulate sequential problem finding optimal parameter combinations in accordance given reference design. Proximal Policy Optimization is used framework, demonstrated case an automotive wheel problem. To reduce heavy burden process required our RL formulation, we approximate neural networks. With efficient data preprocessing/augmentation and architecture, networks achieve generalized performance symmetricity-reserving characteristics. show RL-based produces large number within short inference time exploiting GPU fully automated manner. It different from previous approach using CPU takes much more processing involving human intervention.
منابع مشابه
the impact of generative learning strategy use on efl learners’ reading achievement and critical thinking
این پژوهش کوشیده است تأثیر به کارگیری راهبردهای زایا را بر پیشرفت و بسندگی در خواندن و درک مطلب و نیز بر تفکر انتقادی زبان آموزان انگلیسی را بررسی کند. در این پژوهش تعداد 46 دانش آموز دختر در مقطع متوسطه اول به صورت تصادفی انتخاب وبه دو گروه تجربی و شاهد با تعداد دانش آموزان مساوی در هر گروه تقسیم شدند. به هر دو گروه سه داستان کوتاه به عنوان موضوع خواندن برای کلاس داده شد. به گروه تجربی درابتدا...
simultaneous shape and reinforcement topology optimization of shell structures by using the method of moving asymptotes
this article is devoted to the simultaneous optimization of shape and topology of shell structures. the optimum shape is obtained together with the optimum layout for the reinforcement layers at the both sides of the shell surface. to solve this problem the finite element method is employed. it is assumed that each element is comprised of a porous media with microscopic rectangular voids where ...
متن کاملDeep learning for topology optimization design
Generative modeling techniques are being rapidly developed in the field of deep learning, and they have been applied to topology optimization. The variational autoencoder (VAE) is a generative modeling technology that extends the autoencoder to generate new images with a limited latent space. We modified the basic VAE structure to encode optimization conditions and decode latent variables for t...
متن کاملMap-based Multi-Policy Reinforcement Learning: Enhancing Adaptability of Robots by Deep Reinforcement Learning
In order for robots to perform mission-critical tasks, it is essential that they are able to quickly adapt to changes in their environment as well as to injuries and or other bodily changes. Deep reinforcement learning has been shown to be successful in training robot control policies for operation in complex environments. However, existing methods typically employ only a single policy. This ca...
متن کاملthe effect of consciousness-raising activities on learning grammatical structures by iranian guidance school efl learners
abstract the first purpose of this study was to investigate the effect of consciousness-raising (c-r) activities on learning grammatical structures (simple present tense in this case) by iranian guidance school efl learners. the second one was to investigate the effect of gender on learning the simple present tense through c-r activities and tasks. finally, this study aimed to investigate the ...
ذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Computer Aided Design
سال: 2022
ISSN: ['1879-2685', '0010-4485']
DOI: https://doi.org/10.1016/j.cad.2022.103225