论文部分内容阅读
针对测量参数存在的非线性、参数间的耦合性以及噪声干扰,将量子粒子群算法引入到流形学习的参数选择中,结合径向基神经网络,提出了一种故障诊断方法。邻域个数和约简维数是流形学习中的关键问题。结果表明:该方法首先利用量子粒子群算法优选邻域个数、约简维数和径向基函数的参数,再利用等距特征映射(ISOMAP)对原始参数进行非线性降维,提取其低维流形特征,从而进行故障分类。结果表明:该方法能够有效地对发动机各种复合故障进行分类,精度达到97.33%,量子粒子群优于基本粒子群优化的分类结果;其分类精度明显优于主元分析(PCA)、核主元分析(KPCA)方法,且有很强的抗噪能力。
Aiming at the nonlinearity of measured parameters, the coupling between parameters and the noise interference, the quantum particle swarm optimization algorithm is introduced into the parameter selection of manifold learning. Combined with RBF neural network, a fault diagnosis method is proposed. The number of neighborhoods and the number of reductions are the key issues in manifold learning. The results show that this method firstly uses the particle swarm optimization algorithm to optimize the parameters of neighborhood number, reduction dimension and radial basis function, and then uses ISOMAP to nonlinearly reduce the original parameters and extract the low Dimensional manifold features, thus fault classification. The results show that the proposed method can effectively classify various complex faults of the engine with an accuracy of 97.33%. The quantum particle swarm optimization is superior to the classification results of the basic particle swarm optimization. The classification accuracy is better than that of the principal component analysis (PCA) Element Analysis (KPCA) method, and has a strong anti-noise ability.