摘要
构建了一种改进对传神经网络,提出一种量子蛙跳算法,采用概率幅作为个体编码,利用量子旋转门来实现个体更新,量子非门实现个体变异,用小波包提取各频段的能量组成特征向量作为网络输入,通过神经网络输出结果判别油井的故障类型。仿真试验结果表明,该方法可以得到更好的分类结果,有一定的应用价值。
In this paper,an improved counter propagation network is built,and a quantum frog leaping learning algorithm is proposed.The individual encodes for probability amplitudes of quantum bits,the updates of individual are performed by quantum rotation gates,the mutations of individual are performed by quantum non-gate.Extracting the feature vectors composed of energy of different bands constitute network input.The types of well fault are judged by output results of neural network.The results of simulation experiments show that the method can get better classification results,it has a certain application value.
出处
《长江大学学报（自科版）（上旬）》
CAS
2013年第3期70-73,6,共4页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI ＆ ENG
关键词
量子蛙跳
对传神经网络
算法
故障诊断
Quantum Frog Leaping
counter propagation network
algorithm
fault diagnosis
