应变数据融合测力传感器的仿真与实验研究

Simulation & Experiment Study On Strain Data Fusion Load Sensor

为了满足测量精度和线性度要求,并根据其设计准则,大量程应变测力传感器的安装空间往往较大。而在一些大型机械和工程结构中,常常需要测量大力值,但能提供的安装空间又比较小,只有超薄型大力值测力传感器才能适用于这类工程应用。本文采用仿真与实验的方法,探讨了不同高度的柱式应变测力传感器在各种偏心和斜载条件下,电桥直接输出值的百分比误差和利用神经网络进行应变数据融合后输出的百分比误差。仿真与实验结果均表明:在柱式应变传感器高度直径(h/d)比为1/4的情况下,神经网络应变数据融合输出仍然可以保持较高的测量精度,且基本不受传感器高度变化的影响,这为大力值小空间测力传感器的制造提供了一种有效途径。

For the sake of the linearity and measuring precision, according to designing rule, when the range of load is increased, the height of the sensor grows, so there must have larger workroom for sensor. But in huge machine and engineering structure, only load sensor with huge load and low height can be used. Aiming at it, in all kind of off-center load and inclined load conditions, with different height of cylinder strain load sensor, the output percent error of electric bridge and that of strain data fusion based on neural network are discussed by simulation and experiment. The simulation and experiments indicates that for cylinder strain load sensor the measuring precision of the strain data fusion load sensor, with the ratio of height and with huge load and low workroom diameter about the 1/4(H/D=1/4), still keeps high. Thus the method presented in the paper provides an effective way to build the load sensor