Abstract: The nonlinear error of load cell is not same in weight range.The character of load cell’s nonlinear error is formulated and a method for adaptive compensation is proposed.The nonlinear error compensation network based on Radial Basis Function Neural Network（RBFNN） is used in upper limit of load cell’ weighing range，the digital filter is applied in the low limit range，and the load cell is not compensated in the middle range.The adaptive selective network is use to choose the subnet for error compensation.The experimental results show that the maximum relative error of load cell with this method respectively drops from 0.2% in its lower interval scale，0.4% in its middle interval scale，and 1.37% in its upper interval scale without compensation to 0.16%，0.04%，and 0.07% after compensation，and its weighing result is more accurate.

Key words: load cell, nonlinear error, adaptive compensation, radial basis function neural network

摘要： 额定量程内称重传感器的非线性误差不同，为此阐述了称重传感器的非线性误差特性，提出了一种非线性误差自适应分段补偿方法：在额定量程的上限区，采用基于径向基函数神经网络（RBFNN）的补偿网络完成传感器非线性误差补偿;在下限区，采用数字滤波器完成非线性误差补偿;在中间区，传感器不补偿。同时利用自适应选择网络，完成了分段补偿的选择。实验表明，采用这种方法补偿后的称重传感器下限区、中间区与上限区的最大相对误差分别由补偿前的0.2%、0.4%、1.37%下降到0.16%、0.04%、0.07%，补偿效果明显。

关键词: 称重传感器, 非线性误差, 自适应补偿, 径向基函数神经网络