Comparison examination between fuzzy logic model and multiple regression model.

概要

論文の詳細を見る
Research on seating comfort has been often conducted in the field of ergonomics. Consideration of differences in the attributes of a person who wishes to sit when he or she chooses a cushion is important. Therefore, focus has been placed on an attribute of body type and a model to predict an appropriate cushion for an individual has been proposed. The relationship between body size and seating comfort has already been in preceding research, and an item for evaluation that is extensively related to overall cushion evaluation is an item related to hardness. In the past, multiple regression was used in a model to indicate the relationship between items for evaluation as expressed in cushion characteristics and overall evaluation as expressed by seating comfort. In the current research, fuzzy logic using non-linearity has been proposed. An experiment for verification was conducted; results indicated a high correlation between predicted values and measured values, and the validity of the system itself was noted. Further progress in research indicated that the model would be highly superior with addition of dummy variables to multiple regression, which is inappropriate, in a conventional subjective evaluation model. Evaluation and examination of both systems were also conducted.
一般社団法人日本人間工学会の論文