基于粒子群混合算法的风力发电机齿轮箱故障诊断

针对传统方法在风力发电机齿轮箱故障诊断中存在精度不高的问题,引入了一种改进粒子群算法优化神经网络的方法。该算法的惯性权重可进行自适应调整,以平衡全局和局部搜索能力。同时,收缩因子可加快算法的收敛速度,以更快收敛到全局最优。仿真结果表明,该方法能较好地识别故障模式,具有一定的实用性。

Fault diagnosis of gearbox for wind turbine based on PSO hybrid algorithm

For the reasons of low accuracy of traditional diagnosis methods on gearbox of wind turbine,a method with the neural networks evolved by modified particle swarm optimization is introduced.In this algorithm,the inertia weight is adjusted to balance the searching capability between whole and partial range.At the same time,the convergence can be accelerated by setting the shrinkage factor,which to find the global optimal solution quickly.The simulation results show that the method can identify the failure modes and has a certain practicality.