Development of Decision-Support System for Aesthetic Design of Bridges Using Genetic Algorithm

近年来，桥梁的美学设计已成为人们关注的焦点，需要考虑桥梁的观赏性正成为桥梁设计的重要条件之一。在过去，透视和蒙太奇技术已经被用来在现场拍摄桥梁的完工。然而，这些技术需要大量的努力和时间，即使是一个图像。相反，由于计算机技术和数据处理的显著发展，图形工作站上的桥梁图像在设计过程中有效地准备了几个备选方案。本研究的目的是开发一个桥梁美学设计决策支持系统。该系统利用遗传算法处理离散设计变量和不确定目标函数的卓越能力，将设计者（用户）的偏好和主观性引入决策过程。本研究开发的原型系统是基于计算机图形学和遗传算法的。虽然遗传算法适用于产生许多设计方案，但计算机图形学对于准备其视图模拟是有用的。通过在斜拉桥美学设计中的应用，说明了该系统的适用性。

Recently, the aesthetic design of bridges has been focused since the need to consider the view of the bridges is becoming one of the important design conditions. In the past, perspective and photomontage techniques have been utilized to image the completion of bridges at site. However, these techniques have required a lot of efforts and time even for an image. On the contrary, due to the remarkable development of computer technology and data processing, the image of bridges on a graphic workstation is effective to prepare several alternatives in the design process. The objective of this study is to develop a decision-supporting system for the aesthetic design of bridges. Using such remarkable ability of genetic algorithm that it can deal with discrete design variables and indefinite objective functions, the present system can introduce the preference and subjectivity of designers (users) into the process of decision-making. The prototype system developed in this study is based on computer graphics as well as genetic algorithm. While genetic algorithm is applicable to produce mane design alternatives, computer graphics is useful to prepare their view simulations. The applicability of the system developed here is demostrated through applications to the aesthetic design of cable-stayd bridges.