Predicting aircraft seat comfort using an artificial neural network

Aircraft seat pitch is one of the most important factors affecting passengers ability to sit comfortably for a longer time on the plane. The purpose of this study was to illustrate the association between different seat pitches and overall comfort index, and seat-interface pressure variables as well as try to predict aircraft seat comfort. Through an experimental study, 11 subjects (age, 26.3 ± 1.6; height, 169.8 ± 8.5; body mass, 65.1 ± 15.6) rated five different seat pitch settings (55 examples, 11 participants and five scenarios). Descriptive statistics, together with one-way analysis of variance were used to determine which of the metrics could be used to distinguish the overall comfort index. The results show that overall comfort index was statistically significant (p < 0.05) between different seat pitches, but there was no statistically significant difference (p > 0.05) in interface pressure variables. In addition, a multilayer feed forward neural network with one hidden layer was proposed with the help of Matlab Neural Network Toolbox. The model explained 99% of the variance in overall comfort index with the root mean square error (RMSE) of 0.12551. The remainder of the total data was used for validation purposes; the correlation is r = 0.775, p < 0.01, and the RMSE is 1.21031. That suggests this model has a significant relationship between the actual and predicted overall comfort index.