RBF人工神经网络在坐姿评价中的应用

Application of RBF artificial neural network to sitting-posture evaluation

对传统坐姿评价算法计算过程冗杂、繁琐、准确率不高的问题,提出了建立基于径向基函数(radial basis function,RBF)神经网络的坐姿状态描述应用模型,通过搭建坐姿压力采集实验平台,利用MATLAB进行数据的归一化处理、计算、模型仿真,训练出一套具有良好学习能力、容错能力的坐姿状态描述神经网络。与常规的坐姿评价计算方法相比,该应用方法不但能够降低数据采集的难度,而且提高了坐姿评价5%的运算效率,为设计坐姿监测系统提供理论依据。

In view of the disadvantages of the traditional sitting-posture evaluation including jumbled calculation process,great complexity and low accurancy,a new method to calculate and simulate under the environment of MATLAb is put forward through establishing the application model of sitting posture description based on RBF neural network. By setting up the experimental platform of the sitting position,carrying out the data processing through MATLAB,calculating and model simulating,a set of sitting posture description neural network with good learning ability,and fault-tolerance capability has been completed. Compared with conventional method,this method can not only reduce the difficulty of data acquisition,but also improve the operation efficiency of 5% and provide a theoretical basis for the design of sitting-posture monitoring system.