基于灰色BP神经网络模型的文胸压力预测研究

Grey BP Neural Network Model Based Bra Pressure Prediction

为了更准确地预测文胸压力,以40名在校健康女大学生为研究对象,在其穿着合适尺码的文胸后,对文胸10个重要部位进行压力测试。建立BP神经网络和灰色BP神经网络预测模型,利用人体5个测量参数对文胸10个部位的压力值进行预测。结果表明,通过BP神经网络模型预测的文胸压力值只能达到真实压力值的73.55%,而通过灰色BP神经网络模型预测的文胸压力值可以达到真实压力值的84.70%。灰色BP神经网络模型可以结合BP神经网络模型和灰色模型的优点,发挥两种模型的优势,对文胸压力值做出较为准确的预测。

In order to predict bra pressure more accurately,40 healthy female college students were selected as subjects to wear suitable size bras,and the pressures on 10 important parts of the bra were tested.The BP neural network model and gray BP neural network prediction model were established,and the pressure of the 10 parts were predicted by using five measured parameters of the human body.The results show that the bra pressure values predicted by BP neural network model can only reach 73.55%of the real pressure values,while that predicted by gray BP neural network model can reach 84.70%of the real pressure values.Gray BP neural network model can combine the advantages of BP neural network model and gray model in order to take full use of the advantages of both,which can make a more accurate prediction of the bra pressure value.