关键词: 姿势预测, 人机工程, 计算机辅助人机工程

Abstract: Traditional research methods for postural control were analyzed. Influencing factors on posture selection through experiments were studied, and importance of these factors were compared. Then, cost functions optimization was put forward to simulate the man's behavior in reality. Cost functions for posture prediction were defined and applied to algorithm. Genetic algorithm was adopted to deal with posture prediction. Comfort, potential energy and joint ranges of motion were set as constraints, as a result the posture prediction was more realistic. Experiment proved that the model could solve the reachable problem of space point in computer aided ergonomics, and predict reasonable posture.

Key words: posture prediction, ergonomics, computer aided ergonomics