无损检测和人工神经网络模型联用评估核电电缆剩余使用寿命的研究

The remaining life assessment of the cable in nuclear power plants and nuclear ship has been one of the most important topics in the area of electrical aging management. Currents testing methods can only detect defects and damage of cable insulation, but can’t predict the remaining life. Therefore, It has great researching potential and application value that the development of novel nondestructive evaluation techniques and models for the remaining life assessment of the cable. The main object of this study is EPR insulated cables, which are one of the primary cable types used in nuclear power plants (NPPs) and ship. Through the accelerated thermal aging test, detect and analyze a variety of chemical, physical and electrical indicators closely related to aging of cable. Further study the relationship between the frequency domain spectroscopy and ultrasonic measurement curve with physical properties, chemical properties and electrical indicators. Find the sensitive measuring range of frequency domain spectroscopy and ultrasonic measurement to aging of insulation material, and design the in-situ monitoring systems. Application of the mathematical methods to derive single factors and multiple factors aging equations. Then, construct the artificial neural network models of aging factors, time, key indicators and remaining useful life. Finally, Application of the non-destructive in-situ testing technology (frequency domain spectroscopy and ultrasonic measurement) and artificial neural network method to evaluate the remaining life of the cable in nuclear power plants and nuclear ship.

核电站和核动力船舶中的电缆剩余使用寿命评估是电气老化管理的重要课题之一。目前检测手段仅可发现电缆绝缘层的缺陷和损伤情况，而不能预计剩余使用寿命。因此开发针对电缆在核环境的剩余使用寿命无损评估技术及相关模型具有极大的研究潜力和应用价值。本研究以在核电和船舶使用频率较高的EPR绝缘电缆为主要研究对象，以电缆绝缘层老化性能失效作为电缆使用寿命终止的判据。首先通过实验室加速热老化试验，检测与老化形成密切相关的化学、物理和电气关键参数；进一步分析介电频谱和超声波测量曲线与物理性能、化学性能及其他电气参数之间的关系；发现绝缘材料对老化较为敏感的介电频谱和超声波测量区间，设计原位监测系统；随后用数学分析方法推导出单因素和多因素老化状态方程，并用人工神经网络建立老化因子、时间、检测指标和剩余使用寿命之间的关系模型。最终实现超声波和介电频谱无损在线检测与人工神经网络模型联用评估电缆的剩余使用寿命。

核电站和核动力船舶中的电缆剩余使用寿命评估是电气老化管理的重要课题之一。目前检测手段仅可发现电缆绝缘层的缺陷和损伤情况，而不能预计剩余使用寿命。因此开发针对电缆在核环境的剩余使用寿命无损评估技术及相关模型具有极大的研究潜力和应用价值。本研究以在核电和船舶使用频率较高的EPR绝缘电缆为主要研究对象，以电缆绝缘层老化性能失效作为电缆使用寿命终止的判据。首先通过实验室加速热老化试验，检测与老化形成密切相关的化学、物理和电气关键参数；进一步分析介电频谱和超声波测量曲线与物理性能、化学性能及其他电气参数之间的关系；发现绝缘材料对老化较为敏感的介电频谱和超声波测量区间，设计原位监测系统；确定了EAB和硬度计压头模量数值与电缆老化高度相关，发现随着老化时间的延长，检测到声速在加速老化1080h后快速上升，随后趋于平稳，在1200h的声速为为1.61 mm /µsec。在7.5GHz频率时检测不同老化时间的最大ε'值，而老化时间为1152 hr和1224 hr的样品的ε'增大幅度尤其明显。并用人工神经网络建立老化因子、时间、检测指标和剩余使用寿命之间的关系模型。最终实现超声波和介电频谱无损在线检测与人工神经网络模型联用评估电缆的剩余使用寿命。