Electric vehicle power battery remaining service life prediction method

Abstract

The invention discloses a method for predicting the remaining service life of a power battery of an electric vehicle. The method is used for predicting the RUL of the battery under a driving conditionon the basis of a light gradient elevator (GBM) method. The LightGBM uses a histogram optimization strategy to reduce the number of traversal times of the data sample set and improve robustness of the method, the risk of overfitting is reduced through a depth-first splitting (leaf-wise) strategy; a gradient single-side sampling (GOSS) strategy is adopted, so the data dimension can be reduced; a feature dimension is reduced by using an Exclusive Feature Bundling (EFB); however, the LightGBM method has difficulty in parameter setting, so the Hyperopt based on distributed asynchronous algorithmconfiguration / hyper-parameter optimization is adopted to optimize the complicated hyper-parameters of the LightGBM method. Then, the method is applied to battery RUL prediction under the simulated driving working condition, and results show that the method can ensure rapidity, accuracy and robustness of RUL prediction under the condition of low memory usage.

Description

technical field [0001] The invention relates to the field of electric vehicle batteries, in particular to a method for predicting the remaining service life of a power battery of an electric vehicle. Background technique [0002] In today's battery market, power batteries are mainly lead-acid batteries, nickel-metal hydride batteries and lithium-ion batteries, and lithium-ion batteries rely on their energy density, cycle life, self-discharge rate, charge and discharge performance, operating temperature range, etc. The excellent performance in the indicators occupies a dominant position in applications such as electric vehicles (Electric vehicle, EV) or hybrid electric vehicles (Hybrid electric vehicle, HEV), portable electronic products, etc. However, as the number of charge and discharge cycles increases, the performance of lithium-ion batteries will decline, specifically manifested as capacity fading. Among them, the remaining useful life (RUL), which reflects the capacit...

AI Technical Summary

Problems solved by technology

When the available capacity drops to 70% to 80% of its nominal capacity, the battery stability is seriously affected, impairing the basic charging and discharging performance, and even causing catastrophic accidents

Images