>> 自然科学版期刊 >> 2018年02期 >> 正文
基于模糊系统的接触电阻预测算法研究
供稿: 王刚;姚永琪;张博;蒋晓旭;苏计东;王旭;王文博 时间: 2018-03-29 次数:

作者:王刚姚永琪张博蒋晓旭苏计东王旭王文博

第一作者单位:‍平高集团有限公司

摘要:运用模糊系统对接触电阻进行预测,通过试验得到试验数据,将试验数据分成训练数据与测试数据,训练数据用于训练模糊系统,测试数据用于测试模糊系统。将梯度下降算法与禁忌搜索算法混合得到改进禁忌搜索算法,训练模糊系统,所用算法依次为梯度下降算法、禁忌搜索算法、所得改进算法,并建立相应的接触电阻回归模型。对所得模型进行测试,结果表明,由改进算法训练模糊系统所得接触电阻模型的测试结果优于其他模型,并且该方法预测接触电阻精确可靠。

Abstract:The contact resistance is predicted by fuzzy system. The enough data are obtained through experiment, and all data are divided into two parts, training data and testing data. The training data are used to train fuzzy system, and the testing data are used to test fuzzy system. New algorithm is obtained by combining gradient descent algorithm and tabu search algorithm to train the fuzzy system. The training algorithms include gradient descent algorithm, tabu search algorithm and the new algorithm. The regression models of contact resistance are founded. All models are tested, and the prediction effect of the model obtained from new algorithm is best. The results show that the fuzzy system trained by new algorithm is reliable and precise to predict the contact resistance.

关键词:接触电阻;模糊系统;梯度下降算法;禁忌搜索算法;回归分析;

DOI:10.16186/j.cnki.1673-9787.2018.02.014

分类号:TM743

最近更新