基于SSA-PNN的矿井提升机主轴装置故障诊断

Fault diagnosis of mine hoist spindle device based on SSA-PNN

对于矿井提升机来说,主轴装置是其核心,它能否正常运行关系着整个矿井提升机的工作进程,应用概率神经网络可以有效地进行矿井提升机的故障诊断,但是传统概率神经网络存在平滑因子具有主观性的缺点。针对这个不足,文章引入樽海鞘群算法对其进行优化,提出了樽海鞘群算法优化的概率神经网络(SSA-PNN)的主轴装置故障诊断。实验表明,SSA-PNN与遗传算法、BP神经网络相比,能够更加准确、快速地进行分类,实现对矿井提升机主轴装置案例已知故障的有效诊断。

For the mine hoist, the spindle device is its core. Its normal operation is related to the working process of the whole mine hoist. The application of probabilistic neural network can effectively diagnose the fault of the mine hoist, but there exists the shortcoming that smoothing factor in traditional probabilistic neural network is subjective. Aiming at this deficiency, in this paper the Schonhage.Strassen algorithm was introduced to optimize it. The fault diagnosis of the probabilistic neural network(SSA-PNN)of the the Schonhage.Strassen algorithm was proposed. The experiment shows that SSA-PNN can be classified more accurately and quickly compared with genetic algorithm, BP neural network, and realized the effective diagnosis of the known failure of the case of the mine hoist spindle device.