针对压阻式压力传感器的温度补偿问题,提出一种将样条插值与最小二乘拟合相结合的补偿算法,并结合传感器标定实验数据进行仿真试验。结果表明该方法相比曲面拟合方法、BP神经网络和RBF神经网络,补偿最大相对误差和平均补偿时间分别为0.103%和0.135 4 s,不仅能够满足高精度测试要求,而且可减少标定工作量达到提升生产效率的目的。
To address this temperature compensation problem of Piezo-resistive pressure sensor,a combined method of spline interpolation and least square fitting was presented.The simulation results based on a calibration experiment demonstrate the maximum relative error of the proposed compensation method is 0.103% as well as the mean compensation time is about 0.135 4 s.Compared with the results come from surface fitting,BP neural networks and RBF neural networks,in addition to short the calibration process the proposed compensation method can also obtain a more satisfactory compensation precision. The compensation results also indicate that the presented temperature compensation method is able to reach a balance between the time cost and the compensation effectiveness,which lays a foundation to the further reach.
Instrument Technique and Sensor
piezo-resistive pressure sensor
least square fitting