基于改进BP神经网络的设备管理与维修人员培训效果评估方法

Improved BP Neural Network-Based Training Effect Assessment Method for Equipment Management and Maintenance Personnel

摘要: 设备管理与维修培训领域，目前缺乏对受训者培训效果的评估方法，针对这一问题，建立了基于BP神经网络的设备管理与维修人员培训效果评估模型，利用麻雀搜索算法(SSA)对BP神经网络进行全局优化。实验结果表明，麻雀优化算法将误差值收敛至0.5以下。经过与其他模型对比测试，改进BP神经网络模型在R²、精度和召回率上表现优异。模型具有可重复性，在训练和测试集上的实验结果稳定，可以为设备管理与维修培训效果评估提供支持。

Abstract: In the field of equipment management and maintenance training, there is a lack of methods to assess the training effect of trainees, to address this problem. Thus, a BP neural network-based model for assessing the training effect of equipment management and maintenance personnel was established, and a global optimisation of the BP neural network was carried out by using the Sparrow Search Algorithm (SSA). The experimental results show that the sparrow optimisation algorithm converges the error value to below 0.5. After comparison testing with other models, the improved BP neural network model performs well in terms of R², precision and recall. The model is reproducible and the experimental results are stable on both training and test sets, which can support the evaluation of equipment management and maintenance training effectiveness.