一种新型航空蓄电池剩余容量预测方法研究  

A new method to predict residual capacity of aviation battery

作　　者：汤巍[1] 严东超[1] 刘勇智[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《电源技术》2012年第4期491-494,共4页Chinese Journal of Power Sources

摘　　要：在深入研究模糊神经网络模型的基础上,引入k-means聚类分析的方法对模型输入数据进行预处理,改进模型推理层,并对系统输出进行简化,提出了一种新的蓄电池剩余容量的预测模型。实验仿真表明,与基本模糊神经网络相比改进模糊神经网络的预测精度更高,所需时间更短,实现了对航空蓄电池剩余容量的实时、准确预测。Based on the in-depth study of fuzzy and neural network, a new prediction model for storage battery＇s residual capacity was presented. The clustering method of k-means was introduced to preprocess the inputs of the model, in order to improve the reasoning layer, and simplify the system output as well. Compared With the basic fuzzy and neural network, the improved fuzzy and neural network had high accuracy and needed less time, thus the method realized the real-time and precise forecast of aviation battery＇s residual capacity.

关 键 词：航空蓄电池 剩余容量 模糊神经网络 聚类分析 

分 类 号：TM912[电气工程—电力电子与电力传动]