镍氢合金及电极的偏最小二乘法-人工神经网络优化设计研究

摘要

A Partial Least Square-Back Propagation Network (PLS-BPN) method is applied to materials design of hydrogen storage alloys and property prediction of metal hydride (MH) electrodes. Two samples of hydrogen st orage alloys designed from an optimal region of the PLS sub-space have been pre pared and the predicted target values have been verified by experiment. Some pro perties of MH elecgtrodes:initial capacity, capacity ratio and discharge curve p redicted by the method are in agreement with experimental results.