Lithium-ion battery life evaluation method based on fuzzy nonlinear accelerated degradation process

In engineering practice, restricted by time and expense of the test, we generally choose a small sample as the test object of accelerated degradation test. The size of the sample is so small that the parameters gotten from solving model will not precise enough, which will result in that the rationality of life evaluation results is not guaranteed. We think that this kind of uncertainty belongs to cognitive uncertainty. Through analyzing the mechanism of lithium-ion batteries, we select Wiener process, the most commonly used in degradation model, as the degradation model, and Arrhenius model, in which temperature is considered as the sensitive stress, as acceleration model. Fuzzy theory is used to fuzzify the activation energy of lithium-ion battery to quantify the cognitive uncertainty caused by sample size. Thereby we build a new accelerated degradation model in which the nonlinear degeneration, stochastic uncertainty and cognitive uncertainty of lithium-ion acceleration degradation process are all taken into account. Then we give out the statistical analysis method and evaluation process of fuzzy reliability and life assessment results. Finally, we use accelerate degradation simulated data of lithium-ion battery to illustrate the effectiveness of this method, and analyze the influence of cognitive uncertainty on evaluation results.