【摘要】 在瞬动校验电流控制方法中,提出用补偿系数提高瞬动电流的精度。在分析瞬动电流影响因素并建立等效模型的基础上,采用神经网络控制方法进行控制。以LabVIEW为语言环境,通过Matlab Script节点调用训练好的神经网络,计算出不同等级电流所需的补偿系数,最终由电流产生机构产生瞬动调试电流。为了提高系统的适应性,提出通过变更训练数据样本,实现神经网络的在线调整,进一步提高了瞬动调试电流的控制精度。更多还原

【Abstract】 This paper improves the accuracy of current by referring compensation coefficient in the way of instantaneous calibration current control. The system is controlled by neural network after analyzing the influencing factors of instantaneous calibration current and establishing equivalent model.In LabVIEW,calling neural network which has been trained by Matlab Script node to compute the compensation coefficient.This method can compute the compensation coefficient of different current.Finally, the current producing institution creates instantaneous current.In order to enhance system’s adaptability,the neural network online adjustment is achieved by changing the training samples to improve the accuracy of instantaneous current furthely.更多还原