首页>
外国专利>
WAVELET NOISE REDUCTION AND RELEVANCE VECTOR MACHINE-BASED METHOD FOR PREDICTING REMAINING LIFE OF LITHIUM BATTERY
WAVELET NOISE REDUCTION AND RELEVANCE VECTOR MACHINE-BASED METHOD FOR PREDICTING REMAINING LIFE OF LITHIUM BATTERY
展开▼
机译:基于小波降噪和相关矢量机的锂电池剩余寿命预测方法
展开▼
页面导航
摘要
著录项
相似文献
摘要
A wavelet noise reduction and relevance vector machine-based method for predicting the remaining life of a lithium battery. The present invention relates to a method for estimating the health of a lithium battery and predicting the remaining life thereof, and comprises the following steps: measuring health data of a lithium battery with charge and discharge cycles (1); performing wavelet secondary noise reduction on measured lithium battery capacity data (2); calculating a capacity threshold for lithium battery failure (3); on the basis of a lithium battery capacity data sequence and a charge and discharge cycle data sequence, applying a differential evolution algorithm to a width factor of a relevance vector machine algorithm to perform optimisation selection (4); applying the relevance vector machine algorithm optimised by the differential evolution algorithm to predict the remaining life of the lithium battery (5). The method is simple and effective, and accurately predicts the remaining life of a lithium battery.
展开▼