基于RBFNN的称重传感器温度误差补偿

摘要

Load cell has nonlinear error because of different ambient temperature in work, and it is necessary to compensate. The principle of temperature error of load cell is illustrated, and a method for compensating this error based on radial basis function neural network (RBFNN) is proposed. The training algorithm of RBFNN is described. Using this proposed temperature error compensation method and the load cell which the weighing range is 100 kg,the experiments are implemented under 0 ～60℃. The results show that temperature errors of load cell are decreased, and its weighing accuracy is increased.%称重传感器存在因环境温度不同导致的非线性误差,需要进行补偿.阐述了称重传感器的温度误差机理,提出了一种基于径向基函数神经网络(RBFNN)的称重传感器温度误差补偿方法,并给出了训练算法.采用该方法,利用量程为100kg的称重传感器,在0～60℃范围内进行温度误差补偿实验.实验表明:采用这种方法补偿后,称重传感器温度误差大大减少,提高了称重准确度.