ELM在航空铅酸蓄电池容量检测中的应用

Aviation lead-acid battery capacity detection using extreme learning machine

针对传统BP神经网络训练速度慢、参数选择难、易陷入局部极值等缺点,提出基于极限学习机(ELM)的航空铅酸蓄电池容量检测模型。极限学习机是一种新的单隐层前馈神经网络(SLFNs)学习算法,不但可以简化参数选择过程,而且可以提高网络的训练速度。在确定最优参数的基础上,建立ELM的航空铅酸蓄电池容量检测模型。实验结果表明:LM获得较高的分类准确率和较快的训练速度,从而验证ELM用于航空铅酸蓄电池容量检测模型的可行性和有效性。

Aiming at the traditional BP neural network models are inefficient and prone to fall into local extreme values, the extreme learning machine （ELM） was proposed as an alternative in the detection of aviation lead-acid battery capacities. This new learning algorithm for the studies of single-hidden layer feed forward neural networks（SLFNs） can both simplify the parameter selection process and improve the network training speed. The optimal parameters obtained by this algorithm were used to design a model for detecting aviation lead-acid battery capacities. According to the experimental results, the ELM has made classification more accurate and has quickened network training. Thus, it can be used to test the capacity of aviation lead-acid batteries.