Power cable insulation aging life advanced prediction method and system

The invention provides a power cable insulation aging life advanced prediction method and system, and the method comprises the steps: selecting index historical data, carrying out the index predictionthrough employing a recurrent neural network method, enabling a recurrent neural network to generate a memory function for the state at a previous moment, and calculating an output vector at a current moment. The method has unique superiority in the aspect of predicting the time sequence data, can accurately evaluate the operation reliability of the cable and calculate the residual insulation aging life of the cable, and has higher prediction precision compared with a traditional BP neural network method under the same data scale and calculation environment. 本发明提供了一种电力电缆绝缘老化寿命超前预测方法及系统，本发明通过选取指标历史数据，采用循环神经网络方法进行指标预测，循环神经网络会对上一个时刻的状态产生记忆功能，并用作当前时刻输出向量的计算，在预测时间序列数据方面具有独有的优越性，能够准确评价电缆运行可靠性，计算电缆剩余绝缘老化寿命，在相同数据规模和计算环境下，预测精度相比于传统的BP神经网络方法更高。