人工神经网络用于人体振动响应的分析

Analysis of the human vibration response by artificial neural networks

要对振动环境中的人体进行保护，必须研究人体振动特性。人体全身振动共振频率是一个很重要的动力学参数。对不同身高ｈ、体重ｗ及不同振级ａ下坐姿人体全身振动的第１阶共振频率ｆ进行了测量，获得了３９组数据。利用三层人工神经网络对这些数据进行了分析，排除了个体差异，寻找出ｈ，ｗ及ａ与ｆ的普遍映射关系，从而可以根据ｈ，ｗ及ａ预测呈非线性关系的ｆ值，对人机工程学设计及相关研究有重要意义。

Dynamic characters of human body must be researched for human vibration protection. The resonant frequency of whole human body is a very important dynamic parameter. The data of 39 groups were measured from human iliocristale during whole body vibration in sitting position. Using three-level arificial neural network, the relation between human vibration resonant frequency and height, weight of human body, vibration level was analysed. The common reflective relation of the resonant frequency to the height, weight of the body and exciting acceleration was obtained by eliminating the individual differences.Based on the height, weight of body arid vibration level the resonant frequency could be predicted. This study is important to the ergonomical design and relevant study.