Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Prediction of Anthropometric Dimensions Using Multiple Linear Regression and Artificial Neural Network Models
Ergonomic design of any product requires anthropometric study of prospective users of the product. Commonly more or less 20 anthropometric dimensions of representative population of users are to be collected for designing or manufacturing most of the products ergonomically. This work attempts to find out if there exists any relationship or group of relationships in anthropometric dimensions which will help to ensure that the anthropometry-based applications like design of fashion mannequins, humanoids, and human dummies are not having disproportionate dimensions. Moreover, this establishment of inter-relationship also helps in reducing the number of anthropometric dimensions to be measured. Anthropometric data measured for 18 dimensions of workers working in the various industries is utilized for this purpose. Exhaustive set of Multiple Linear Regression (MLR) and Artificial Neural Networks (ANN) models are developed for all combinations of 1 through 9 anthropometric dimensions as independent variables to relate with each anthropometric dimension not present in the set of independent variables. The set of MLR and ANN models having minimum Normalized Root Mean Square Error (NRMSE) between computed and actual anthropometric dimensions are drawn out from all the models obtained.
Prediction of Anthropometric Dimensions Using Multiple Linear Regression and Artificial Neural Network Models
Ergonomic design of any product requires anthropometric study of prospective users of the product. Commonly more or less 20 anthropometric dimensions of representative population of users are to be collected for designing or manufacturing most of the products ergonomically. This work attempts to find out if there exists any relationship or group of relationships in anthropometric dimensions which will help to ensure that the anthropometry-based applications like design of fashion mannequins, humanoids, and human dummies are not having disproportionate dimensions. Moreover, this establishment of inter-relationship also helps in reducing the number of anthropometric dimensions to be measured. Anthropometric data measured for 18 dimensions of workers working in the various industries is utilized for this purpose. Exhaustive set of Multiple Linear Regression (MLR) and Artificial Neural Networks (ANN) models are developed for all combinations of 1 through 9 anthropometric dimensions as independent variables to relate with each anthropometric dimension not present in the set of independent variables. The set of MLR and ANN models having minimum Normalized Root Mean Square Error (NRMSE) between computed and actual anthropometric dimensions are drawn out from all the models obtained.
Prediction of Anthropometric Dimensions Using Multiple Linear Regression and Artificial Neural Network Models
J. Inst. Eng. India Ser. C
Zanwar, Dinesh R. (Autor:in) / Zanwar, Hitesh D. (Autor:in) / Shukla, Himanshu M. (Autor:in) / Deshpande, Ambarish A. (Autor:in)
Journal of The Institution of Engineers (India): Series C ; 104 ; 307-314
01.04.2023
8 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
| Emerald Group Publishing | 2023