基于Elman神经网络的高压电缆导体温度动态计算方法

导体温度作为运行电缆的关键状态参数,是影响电缆载流能力、绝缘性能的重要因素。针对传统方法难以克服电缆本身物性参数和外部环境变化影响的局限,构建了以电缆运行电流和电缆实时外表面温度为输入,以导体温度为输出的Elman神经网络模型,并引入粒子群算法优化网络的初始权值和阈值。最后,提出了具体的电缆导体温度动态计算方法,通过设计不同工况下电缆温升实验,进而验证了所提出方法的有效性。实验结果表明,所提方法的计算准确度可以不受电缆物性参数、负荷变化方式及外部环境变化的影响,有助于实现电缆导体温度的实时监测。

Dynamic Calculation Method of High-voltage Cable Conductor Temperature Based on Elman Neural Network

Conductor temperature is an important parameter affecting the ampacity and insulation performance of power cable. Traditional conductor temperature calculation methods are difficult to overcome the limitation that their calculation results are affected by variations of the cable' s physical parameters and the external environment. In this study, an Elman neural network model is established with operating current and actual surface temperature of power cable as the input, and conductor temperature as the output. The particle swarm algorithm is introduced to optimize the initial weights and the thresholds of the network. Consequently, a specific dynamic calculation method of cable conductor temperature is proposed, and its validity is verified by cable temperature rise tests under different working conditions. Experimental results show that the calculation accuracy of the proposed method is independent of physical parameters of the cable, load variation mode and external environment. This method may contribute to implementation of real-time temperature monitoring of power cable conductor.