论文部分内容阅读
为实现风电机组齿轮箱故障模式的有效识别,提出一种基于混沌量子粒子群优化BP神经网络(CQPSOBP)的故障诊断方法。在该算法中,利用混沌序列来初始化粒子的初始角位置,可提高种群的遍历性;通过引入变异操作,避免算法陷入早熟收敛,并依此来对BP神经网络的初始权值和阈值进行优化。实例表明,同粒子群优化BP神经网络(PSO-BP)与BP网络的诊断结果相比,CQPSO-BP算法具有收敛速度快、识别精度高的优点,可有效用于风电机组齿轮箱的故障诊断系统中。
In order to effectively identify the gearbox fault mode of wind turbines, a fault diagnosis method based on chaotic quantum particle swarm optimization BP neural network (CQPSOBP) is proposed. In this algorithm, chaos sequence is used to initialize the initial angular position of the particle to improve the ergodicity of the population. By introducing the mutation operation, the algorithm avoids the premature convergence of the algorithm and optimizes the initial weights and thresholds of the BP neural network . The example shows that CQPSO-BP algorithm has the advantages of fast convergence speed and high recognition accuracy compared with the results of BPO (Particle Swarm Optimization) BP neural network (PSO-BP) and BP network, which can be effectively used in the fault diagnosis of wind turbine gearbox System.