https://www.xiaoxiangtao.cn/zh/tags/generative-design/Recent content in Generative-Design on Xiangtao's logHugozh-Hans15989119660@163.com (Xiangtao XIAO)15989119660@163.com (Xiangtao XIAO)Fri, 15 May 2020 00:00:00 +0000https://www.xiaoxiangtao.cn/zh/posts/what-is-generative-design/Fri, 15 May 2020 00:00:00 +000015989119660@163.com (Xiangtao XIAO)https://www.xiaoxiangtao.cn/zh/posts/what-is-generative-design/<p>Generative design is an <a href="%5C%22//en.wikipedia.org/wiki/Iteration%5C%22">iterative</a> <a href="%5C%22//en.wikipedia.org/wiki/Design%5C%22">design</a> <a href="%5C%22//en.wikipedia.org/wiki/Process_%28engineering%29%5C%22">process</a> that involves a <a href="%5C%22//en.wikipedia.org/wiki/Computer_program%5C%22">program</a> that will <a href="%5C%22//en.wikipedia.org/wiki/Generator_%28mathematics%29%5C%22">generate</a> a certain number of <a href="%5C%22//en.wikipedia.org/wiki/Output_%28computing%29%5C%22">outputs</a> that meet certain <a href="%5C%22//en.wikipedia.org/wiki/Constraint_%28mathematics%29%5C%22">constraints</a>, and a <a href="%5C%22//en.wikipedia.org/wiki/Designer%5C%22">designer</a> that will fine tune the <a href="%5C%22//en.wikipedia.org/wiki/Feasible_region%5C%22">feasible region</a> by changing <a href="%5C%22//en.wikipedia.org/wiki/Maxima_and_minima%5C%22">minimal and maximal</a> values of an <a href="%5C%22//en.wikipedia.org/wiki/Interval_%28mathematics%29%5C%22">interval</a> in which a <a href="%5C%22//en.wikipedia.org/wiki/Variable_%28mathematics%29%5C%22">variable</a> of the program meets the <a href="%5C%22//en.wikipedia.org/wiki/Set_%28mathematics%29%5C%22">set</a> of constraints, in order to reduce or augment the number of outputs to <a href="%5C%22//en.wikipedia.org/wiki/Choice%5C%22">choose</a> from. The program doesn&rsquo;t need to be run on a <a href="%5C%22//en.wikipedia.org/wiki/Machine%5C%22">machine</a> like a <a href="%5C%22//en.wikipedia.org/wiki/Digital_computer%5C%22">digital computer</a>, it can be run by a human for example with <a href="%5C%22//en.wikipedia.org/wiki/Pen_and_paper%5C%22">pen and paper</a>. <a href="%5C%22//en.wikipedia.org/wiki/Generative_Design#cite_note-1%5C%22">[1]</a> The designer doesn&rsquo;t need to be a <a href="%5C%22//en.wikipedia.org/wiki/Human%5C%22">human</a>, it can be a <a href="%5C%22//en.wikipedia.org/wiki/Test_%28assessment%29%5C%22">test</a> program in a testing environment or an <a href="%5C%22//en.wikipedia.org/wiki/Artificial_intelligence%5C%22">artificial intelligence</a> (see for example <a href="%5C%22//en.wikipedia.org/wiki/Generative_adversarial_network%5C%22">Generative adversarial networks</a>). The designer <a href="%5C%22//en.wikipedia.org/wiki/Learning%5C%22">learns</a> to <a href="%5C%22//en.wikipedia.org/wiki/Refinement_%28computing%29%5C%22">refine</a> the program (usually involving <a href="%5C%22//en.wikipedia.org/wiki/Algorithm%5C%22">algorithms</a>) with each iteration as his design <a href="%5C%22//en.wikipedia.org/wiki/Goal%5C%22">goals</a> become better <a href="%5C%22//en.wikipedia.org/wiki/Definition%5C%22">defined</a> over time.</p>