Prediction of Plasterer''s Working Postures and Maximum Working Time Using Back-Propagation Network

• Main Authors: Huang-yen Chan, 詹皇彥
• Other Authors: Tao-ming Cheng
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/31248188678973109656

碩士 === 朝陽科技大學 === 營建工程系碩士班 === 95 === Two BPN (back-propagation network) models were established in this study. The first is to predict the risk level of plasterer’s working postures and the second is to forecast plasterer’s maximal working time (MWT) under specific posture. Input variables in the first model are work’s height, weight, gender, working experience presented by time, and age. Output variables of this model is the risk level acquired by guestionnaire. Input variables in the second model consist of worker’s height, weight, gender, working experience presented by time, and age, and posture’s risk level accessed by REBA (rapid entire body assessment) method. The value of MWT used in this model were obtained from interviewing plasterer’s. Case studies indicate that the developed models are capable of good precision.