Battery residual life prediction model construction method and battery residual life prediction method

Abstract

The embodiment of the invention relates to a battery residual life prediction model construction method and a battery residual life prediction method, and the method comprises the steps: obtaining the feature data of each historical use cycle of a sample battery and the corresponding battery capacity, and the feature data represents the feature data of the health state of the sample battery; inputting the feature data into the initial model to obtain an output result; training the initial model based on the relationship between the output result and the feature data of the next historical use cycle to obtain a feature data prediction model; inputting the feature data into another initial model to obtain an output result; training the initial model based on the relationship between the output result and the battery capacity to obtain a battery capacity prediction model; and cascading the output end of the feature data prediction model with the input end of the battery capacity prediction model to construct the battery residual life prediction model, so that the battery residual life prediction model can be simply, conveniently and quickly trained, and the model prediction accuracy is relatively high.

Description

technical field [0001] The embodiments of the present invention relate to the field of battery health monitoring, and in particular to a method for constructing a battery remaining life prediction model and a battery remaining life prediction method. Background technique [0002] Lithium-ion batteries have the advantages of high capacity, long cycle life, and light weight. Since their commercialization in the 1990s, they have been applied to all aspects of life. However, in addition to being widely used and having many advantages, lithium-ion batteries still have certain safety and reliability problems, which limit the development of lithium-ion batteries. For example, batteries may cause fire and explosion accidents; , its capacity will gradually decrease, and when the capacity drops to a certain level, the battery will not work normally. [0003] Battery life can be divided into storage life, service life and cycle life. The storage life refers to the time required for t...

AI Technical Summary

Problems solved by technology

Although the stochastic process method has a wide range of applications and can express the uncertainty of prediction results, it relies heavily on the initialization of hyperparameters, which has a large amount of calculation and low long-term prediction accuracy.

The filtering method uses probabilistic prediction, which can eliminate noise in the data and express uncertainty, but the initialization process is complicated, modeling is difficult, and the timeliness is relatively poor

In artificial intelligence methods, neural network model training requires a large number of sample data, the model is complex, and the autoregressive algorithm is simple to implement, but the long-term prediction accuracy is low

Images