基于属性偏好学习的配电网综合评价方法

为了摆脱在传统地区配电网评价方法中对参评人员个人评价偏好的过度依赖,实现合理、精准的属性权重确定,提出了一种基于属性偏好学习的配电网多指标智能综合评价方法。依据属性测度理论,在置信度准则与评分准则下完成对配电网综合评价模型的构造;进而提出数值绝对偏移率指标以实现对中间值指标的数据预处理;最后,应用随机权神经学习方法,通过对配电网历史训练样本进行有监督学习,计算得到指标属性偏好权重,并依据配电网综合评价模型以及计算所得属性偏好权重完成对配电网待测样本的智能综合评价。与传统的AHP、PSO-SVM以及RWN算法的对比仿真实验验证了该方法的精确性与稳定性。该方法实现了合理、客观的配电网综合评价,对地...

Full description

Saved in:
Bibliographic Details
Published in计算机应用研究 Vol. 34; no. 3; pp. 785 - 790
Main Author 谈元鹏 李买林 许刚
Format Journal Article
LanguageChinese
Published 华北电力大学电气与电子工程学院,北京,102206 2017
Subjects
多属性决策
属性测度
神经网络
评价偏好
配电网评价
evaluation preference
multiple attribute decision making
distribution network evaluation
属性测度
neural networks
评价偏好
神经网络
多属性决策
attribute measure
配电网评价
Online AccessGet full text
ISSN1001-3695
DOI10.3969/j.issn.1001-3695.2017.03.034

Cover

More Information
Summary:为了摆脱在传统地区配电网评价方法中对参评人员个人评价偏好的过度依赖,实现合理、精准的属性权重确定,提出了一种基于属性偏好学习的配电网多指标智能综合评价方法。依据属性测度理论,在置信度准则与评分准则下完成对配电网综合评价模型的构造;进而提出数值绝对偏移率指标以实现对中间值指标的数据预处理;最后,应用随机权神经学习方法,通过对配电网历史训练样本进行有监督学习,计算得到指标属性偏好权重,并依据配电网综合评价模型以及计算所得属性偏好权重完成对配电网待测样本的智能综合评价。与传统的AHP、PSO-SVM以及RWN算法的对比仿真实验验证了该方法的精确性与稳定性。该方法实现了合理、客观的配电网综合评价,对地区配电网评价具有一定的实际应用价值。
Bibliography:51-1196/TP
Tan Yuanpeng, Li Mailin, Xu Gang （School of Electrical ＆ Electronic Engineering, North China Electric Power University, Beijing 102206, China）
To avoid the over reliance of personal preferences on traditional regional distribution network evaluation and achieve reasonable,accurate attribute weights,this paper proposed a multiple index attribute preference learning based intelligent comprehensive evaluating method for distribution network. The proposed method established the distribution network comprehensive evaluation model under confidence criterion and score criterion,according to the attribute measure theory. Then,it pre-processed the intermediate value indexes by introducing numerical absolute deviation rate. Finally,based on the historical training samples of distribution network,the proposed method calculated the preference weights of indexes by employing supervised random weighted neural network learning model. The paper performed intelligent evaluation of test samples by using the distributi
ISSN:1001-3695
DOI:10.3969/j.issn.1001-3695.2017.03.034