基于小波神经网络的航空蓄电池容量预测

Prediction of capacity of aeronautic battery based on wavelet neutral network

为有效预测航空蓄电池的剩余容量,引入小波神经网络,建立了蓄电池内阻和SOC的小波网络模型,通过实验数据对小波网络模型进行训练,得到了用于内阻和SOC预测的小波网络,最后将小波网络的预测结果和BP网络的预测结果进行对比,结果表明小波网络比BP的预测精度要高,更适合用于航空蓄电池容量的预测。

In order to predict the residual capacity of aeronautic battery effectively, the WNN was used to build the wavelet neutral network models of internal resistance and SOC of battery. After training the wavelet neutral network model by experiment data, the wavelet neutral network for predicting internal resistance and SOC was obtained. Comparing the predicting results of wavelet neutral network with BP, the results show that the result of comparing the predicting results of the WNN with the BP shows that the wavelet neutral network with a higher precision is more suitable for the prediction of SOC than BP NN.