基于RBF神经网络的电子秤非线性误差补偿

Compensation of flectronic scale's nonlinear error based on RBF neural netwrok

称重传感器、秤盘机械结构的非线性环节影响了电子秤称重结果的准确度。本文分析了电子秤的非线性误差来源与误差机理,在此基础上利用径向基函数神经网络(RBFNN)构建了一种电子秤非线性误差补偿网络,完成了电子秤的非线性校正。经现场检测表明,采用这种方法补偿后的电子秤称重误差小于国家标准《JJG 555-1996非自动秤通用检定规程》规定的中级秤允许误差,提高了称重准确度。

The nonlinear errors derived from load cell and scale pan are resulted in the weighing errors of electronic scale. In this paper, the source and the mechanics of nonlinear errors in electronic scale are expounded. A compensation method for electronic scale＇ s nonlinear error based on radial basis function neural netwok（RBFNN） is designed, and then the nonlinear error is corrected. The experimental results are show that the weighing error of the electronic scale with this proposed method is less than the permissible error defined by the ＂ JJG555-1996 General Verification Regulation for Nonautomatic Weighing Instrument＂, and its weighing result is more accurate.