基于灰色预测的电力电缆老化过程预警仿真研究

摘要

为了对电力电缆的老化状况进行监测和预警,对理想条件下电缆老化过程进行了模拟和仿真,得到了可以表征电缆运行状况的电流信号.将灰色理论引入到了老化预测中,并与基于时间序列的神经网络预测模型做了对比分析.结果表明,在本文条件下,基于新陈代谢的灰色预测模型的预测精度、过程可重复性和计算速度上要优于神经网络预测模型.因此,通过预测电流变化趋势,反映电缆的绝缘老化程度时,改进的灰色预测与神经网络预测相比具有一定的优势.同时,建立了一种电缆老化预警机制,为进一步诊断和维护老化严重的电力电缆提供一种判据.%In order to monitor the process of power cable aging and give a pre-alarm, the simulation of cable aging process is made in perfect condition, and then the current signal that can characterize the status of power cable is gotten. The grey prediction theory is introduced to cable aging prediction, and then a comparative analysis is made with a neural network prediction model based on time series. The results show that, in the paper's condition, prediction accuracy, repeatability of the process and computing speed of grey prediction model based on metabolism are superior to that of the neural network prediction model. Therefore, when reflecting the aging of cable insulation by predicting the current trend, the improved grey prediction compared with neural network has certain advantages. Meanwhile, a cable aging early warning mechanism is established, which provides a criterion for further diagnosis and maintenance of aging serious power cables.