文献类型：期刊文献

中文题名：基于INTLBO-SVR的低压断路器热脱扣时间预测

英文题名：INTLBO-SVR based prediction of thermal tripping time of low-voltage circuit breaker

作者：张凌波[1];操春名[1];徐震浩[1]

机构：[1]华东理工大学信息科学与工程学院

年份：2019

卷号：42

期号：16

起止页码：30

中文期刊名：现代电子技术

外文期刊名：Modern Electronics Technique

收录：CSTPCD;;北大核心:【北大核心2017】；

基金：国家自然科学基金资助项目(61573144);国家自然科学基金(61773165);国家自然科学基金(61673175)~~

语种：中文

中文关键词：低压断路器;热脱扣时间预测;支持向量回归;隔离小生境教学算法;参数优化;预测模型

外文关键词：low-voltage circuit breaker;thermal tripping time prediction;support vector regression;INTLBO;parameter optimization;prediction model

摘要：热脱扣时间是低压断路器的关键指标,利用断路器生产过程中可检测数据可以实现热脱扣时间的预测。针对支持向量回归(SVR)进行热脱扣时间预测,参数的选择对预测的精度和泛化性能影响较大问题,提出一种基于隔离小生境教学算法(Isolated Niche Teaching-Learning-Based Optimization Algorithm,INTLBO)优化支持向量回归的热脱扣时间预测方法。该方法针对教学算法易陷入局部最优的缺点,采用隔离机制的小生境技术对其进行改进,然后利用INTLBO寻优找到最优的SVR参数。根据低压断路器生产历史数据,建立基于INTLBO-SVR的热脱扣时间预测模型。仿真结果表明,与TLBO-SVR和常规SVR等方法相比,INTLBO-SVR模型具有较好的预测性能。
Thermal tripping time is the key index of low-voltage circuit breaker.The prediction of thermal tripping time can be realized by using the detectable data in the production process of circuit breaker.In allusion to the problem that the pa-rameter selection has great influence on the prediction accuracy and generalization ability when support vector regression(SVR)is used to predict thermal tripping time,a thermal tripping time prediction method based on INTLBO-SVR is proposed.In order to overcome the shortcoming that the teaching-learning-based optimization(TLBO)algorithm is easy to fall into local optimiza-tion,an isolated niche technique is used to improve the original TLBO so that the optimal SVR parameters can be found.Ac-cording to the historical production data of low-voltage circuit breaker,a thermal tripping time prediction model based on INTL-BO-SVR is established.Simulation results show that the INTLBO-SVR model has better prediction performance in comparison with TLBO-SVR and classical SVR.