Search

The posture prediction is important in ergonomics evaluation system and virtual human animation. The posture choosing phase by human is a non-logical evolutionary approach, so we adapt orthogonal interactive genetic algorithm (OIGA) to calculate the weights of different factors which affect posture choosing. The case analysis of driving posture prediction and the application of the Ergonomics Driving Simulation System proves that OIGA can use a less interactive algebra to reach a higher value. So it is a feasibility and practicability algorithm.

The emulation of the ergonomics design on automobile requires us to combine human models with prediction of body comfort. The comfort prediction model of posture needs to express the real driving instances. Most of the ergonomics systems used posture databases and reverse-movement kinematics in order to adjust a posture. However, it is difficult to describe the body comfort involving human body joints. This paper introduces a research on comfort prediction model, which is based on our experiments from which we have obtained the relations among driving posture, and the body parameters. Using support vector machine (SVM) regression analysis methods, we have derived the functional expression between these parameters, which formed the basis of a prediction model on driving posture. In this approach, a sport-chain model was used to confirm the driving posture, which was captured using motion-capture equipment, with the marking points pasted on the surface of a human body.