In recent years, styling (shape) of industrial products has diversified, attention is paid to automatic generation technology by computer such as topology optimization method. However, the design evaluation method of the shape is not established, and the designer selects subjectively and inefficiently the shape generated by the computer. This research focuses on “complexity” which is one of the important macroscopic shape features in human shape recognition, and constructed an automatic shape generation system using quantification method of “complexity” of curved surface shape. Specifically, Gaussian curvature integration, which is an index of “complexity”, is incorporated in the process of calculating the fitness of the System for Deriving Diverse Solutions based on emergence. This study constructed a system that generates shapes with arbitrary “complexity”. As a validity verification of the system, a cognitive experiment on "complexity" were conducted using 15 chair shapes generated by the system, and the relationship between the sensory evaluation value and the index was analyzed. As a result, the coefficient of determination by logarithmic approximation is 0.76, the correlation between the index and sensory evaluation value of “complexity” is confirmed.
抄録全体を表示