热老化加速下高压套管聚醚酰亚胺复合材料抗拉强度预测  

Tensile Strength Prediction of High-voltage Bushing Polyetherimide Composites Under Accelerated Thermal Aging

作　　者：罗汉武 曹阳 王坤涵 粱佳宇 刘文斌 LUO Hanwu;CAO Yang;WANG Kunhan;LIANG Jiayu;LIU Wenbin(State Grid East Inner Mongolia Electric Power Company,Hohhot,Inner Mongolia 010010,China;State Grid East Inner Mongolia Electric Power Research Institute,Hohhot,Inner Mongolia 010010,China)

机构地区：[1]国网内蒙古东部电力有限公司,内蒙古呼和浩特010010 [2]国网内蒙古东部电力有限公司电力科学研究院,内蒙古呼和浩特010010

出　　处：《广东电力》2024年第8期117-124,共8页Guangdong Electric Power

基　　金：国网内蒙古东部电力有限公司科技项目(526604210003)。

摘　　要：热老化易导致玻璃纤维/聚醚酰亚胺(glass fiber polyetherimide,GF/PEI)电缆套管失效,因此研究GF/PEI复合材料热老化加速作用下的力学性能变化规律对提高电缆套管可靠性至关重要。首先,对GF/PEI复合板进行85~145℃范围内不同温度的热时效处理;然后,对不同热老化温度下的样品进行拉伸试验,测试其抗拉强度,分析不同热老化作用下材料色差变化规律;最后,基于抗拉强度和色差的回归分析结果,采用支持向量回归模型建立色差和抗拉强度之间映射模型,实现抗拉强度预测。测试结果表明,随着热老化作用温度逐渐上升,GF/PEI复合材料的抗拉强度逐渐下降,而色差总体上保持增长趋势,且两者呈现显著相关性。建模结果表明,对比传统的线性回归模型和BP神经网络方法,支持向量回归模型能够更加有效地表征色差和抗拉强度之间的映射关系,且预测GF/PEI复合材料抗拉强度的误差更小。Thermal aging can easily lead to the failure of glass fiber/polyetherimide(GF/PEI)cable sleeves.Therefore,it is very important to study the change law of mechanical properties of GF/PEI composites under the effect of thermal aging to improve the reliability of cable sleeves.Firstly,the GF/PEI composite board was subjected to thermal aging treatment at different temperatures in the range of 85~145℃,and then tensile experiments were carried out on the samples at different thermal aging temperatures to test the tensile strength.At the same time,the law of color difference of the materials under different thermal aging effects was investigated.Finally,a support vector regression model was used to establish the relationship between color difference and tensile strength to predict the tensile strength.The test results show that the tensile strength of GF/PEI composites decreases gradually with the increase of the thermal aging temperature,while the color difference generally maintains an increasing trend,and the both show significant correlation.The modeling results show that compared with traditional linear regression model and BP neural network,l the proposed support vector regression model can effectively characterize the nonlinear mapping relationship between color difference and tensile strength,and can successfully predict the tensile strength of materials based on color difference with smaller errors.

关 键 词：玻璃纤维/聚醚酰亚胺 热老化 支持向量回归 强度预测 

分 类 号：TM216[一般工业技术—材料科学与工程] TQ322.3[电气工程—电工理论与新技术]