论文部分内容阅读
针对BP神经网络鲁棒性、容错性不强的问题,提出双向BP神经网络,更直接地建立与先前状态的映射关系;利用量子粒子群算法(QPSO)优化双向BP神经网络的权值和阈值,克服其学习算法复杂、收敛速度慢的缺点,来得到精度更高的网络。将改进的双向BP神经网络应用于逆变电路的故障诊断,测试结果表明该算法比双向BP神经网络具有更强的收敛性和精确率,为逆变电路的故障诊断提出一个新的思路。
Aiming at the problem of poor robustness and fault tolerance of BP neural network, bidirectional BP neural network is proposed to more directly establish the mapping relationship with previous state. QPSO is used to optimize the weights and thresholds of bidirectional BP neural network , To overcome its learning algorithm is complex, slow convergence shortcomings, to get a more accurate network. The improved bi-directional BP neural network is applied to the fault diagnosis of the inverter. The test results show that the proposed algorithm has more convergence and accuracy than the bi-directional BP neural network, and provides a new idea for the fault diagnosis of the inverter.