Lithium battery state-of-charge estimation based on theory of evidence with interval analysis

In this paper, we investigate a new method for lithium battery state-of-charge (SoC) estimation. Our proposed method is based on the well-known bounded-error state estimation with forward-backward propagation. However, instead of describing each variable error propagating in the algorithm by a single error bound (or a single interval), we describe it by families of error bounds. The technique that we use to model our knowledge about these variables is derived from the theory of evidence. To evaluate our proposed method, we use two different current sensors and two different voltage sensors to represent our knowledge about those variables with different levels of uncertainty. Experimental results show that root-mean-square errors of the proposed method are slightly better than those of a Kalman-filter-based approach. Also, this work studied a few interesting aspects of the estimation based on the theory of evidence, such as effects of assigning different masses (or degrees of belief) to error bounds.