基于1DCNN-LSTM的锂离子电池SOH预测
|
王英楷, 张红, 王星辉
|
Hybrid 1DCNN-LSTM model for predicting lithium ion battery state of health
|
Yingkai WANG, Hong ZHANG, Xinghui WANG
|
|
表1 算法指标
|
Table 1 Algorithm indicators
|
|
数据 | 50个预测点 | 80个预测点 |
---|
MAE | MAPE/% | | | RUL | MAE | MAPE/% | | | RUL |
---|
B5 | LSTM | 0.032 | 2.3 | 125 | 117 | -8 | 0.038 | 2.68 | 125 | 112 | -13 | 1DCNN-LSTM | 0.0092 | 0.62 | 125 | 125 | 0 | 0.013 | 0.91 | 125 | 125 | 0 | B6 | LSTM | 0.024 | 1.9 | 110 | 117 | 7 | 0.036 | 2.12 | 110 | 120 | 10 | 1DCNN-LSTM | 0.011 | 0.71 | 110 | 111 | 1 | 0.015 | 1.02 | 110 | 112 | 2 | B7 | LSTM | 0.035 | 2.6 | 97 | 103 | 6 | 0.040 | 2.8 | 97 | 111 | 14 | 1DCNN-LSTM | 0.013 | 0.83 | 97 | 99 | 2 | 0.02 | 1.13 | 97 | 103 | 6 |
|
|
|