Abstract: Rolling bearing is one of the most commonly used components in rotating machinery. It’s easy to be damaged. But its working condition usually is very complex, which makes it difficult to diagnose rolling bearing faults accurately. In order to improve the effectiveness of rolling bearing fault?diagnosis, the improved quantum bee colony algorithm is introduced into BP neural network and the BP neural network diagnosis model based on Improved Quantum Artificial Bee Colony algorithm（IQABC-BP） is proposed. Firstly, an improved quantum bee colony algorithm is proposed to solve the problem of quantum bee colony algorithm. Then the improved quantum bee colony algorithm is applied to optimize the initial weight, threshold and the number of hidden layer of BP neural network. Finally the model of BP neural network based on the improved quantum bee colony algorithm with super parallel and ultra-high speed is proposed and is applied to?fault diagnosis of rolling bearing. The experimental results show that IQABC-BP with convergence speed faster and better fault diagnosis has the very good application value.

Key words: BP neural network, quantum bee colony algorithm, fault diagnosis

摘要： 滚动轴承是旋转机械中最常用的部件之一。滚动轴承很容易损坏，而它的工作条件通常比较复杂，很难对其故障进行准确判断。为了提高滚动轴承故障诊断的有效性，构建了一种新的基于改进量子蜂群算法和BP神经网络的滚动轴承故障诊断模型（IQABC-BP）。首先针对量子蜂群算法在种群初始化和进化过程中存在的问题，提出了一种改进量子蜂群算法，然后利用改进量子蜂群算法对BP神经网络的初始权值、阈值和隐含层单元数进行优化，建立了一种具有超并行超高速的基于改进量子蜂群算法的BP神经网络模型，并应用于滚动轴承的故障诊断中。实验结果表明，IQABC-BP模型收敛速度更快，故障诊断效果更好，具有很好的应用价值。

关键词: BP神经网络, 量子蜂群算法, 故障诊断