Lithium-Ion Battery Capacity Estimation: A Method Based on Visual Cognition

Abstract

This study introduces visual cognition into Lithium-ion battery capacity estimation. The proposed method consists of four steps. First, the acquired charging current or discharge voltage data in each cycle are arranged to form a two-dimensional image. Second, the generated image is decomposed into multiple spatial-frequency channels with a set of orientation subbands by using nonsubsampled contourlet transform (NSCT). NSCT imitates the multichannel characteristic of the human visual system (HVS) that provides multiresolution, localization, directionality, and shift invariance. Third, several time-domain indicators of the NSCT coefficients are extracted to form an initial high-dimensional feature vector. Similarly, inspired by the HVS manifold sensing characteristic, the Laplacian eigenmap manifold learning method, which is considered to reveal the evolutionary law of battery performance degradation within a low-dimensional intrinsic manifold, is used to further obtain a low-dimensional feature vector. Finally, battery capacity degradation is estimated using the geodesic distance on the manifold between the initial and the most recent features. Verification experiments were conducted using data obtained under different operating and aging conditions. Results suggest that the proposed visual cognition approach provides a highly accurate means of estimating battery capacity and thus offers a promising method derived from the emerging field of cognitive computing.

Effect of Current Rate and Prior Cycling on the Coulombic Efficiency of a Lithium-Ion Battery

“…It is possible to expand their market by increasing their cycle life. In the past few years, substantial efforts have been accomplished for model development and to anticipate capacity fade in lithium ion batteries [1][2][3]. Notwithstanding, experimental data are necessary for the investigation of the capacity fading mechanisms and the aging processes of a battery system [1].…”

Remaining capacity estimation of lithium-ion batteries based on the constant voltage charging profile

“…To overcome these disadvantages of PSO, QPSO was developed by introducing the quantum mechanics into the convergence process of PSO. From the quantum mechanics perspective, the velocity and position of the particle cannot be determined simultaneously due to the famous uncertainty principle [ 23 ]. Therefore, each particle involved in QPSO is hypothesized in a quantum state and is characterized by a wave function rather than its velocity and position.…”

“…For instance, Li et al [ 22 ] extracted four characteristic parameters from charging voltage curves and constructed a particle filter (PF) model to estimate discharge capacity. Cheng et al [ 23 ] applied visual cognition technique to build the capacity degradation model based on several geometrical features extracted from the current and voltage curves. To effectively capture the nonlinearity relationship between the features and capacity, various machine learning approaches have been integrated with feature-based methods as well.…”

“…During everyday usage, batteries are seldom fully discharged to the 0% SOC level but are usually recharged from a partially discharged state to the 100% SOC level. This incomplete discharging process will affect the initial state (such as the initial voltage) and processed variables (such as the charging time) of the subsequent charging process, thereby restricting the extraction of external features depending on a deterministic and intact charging/discharging process, such as the time ratio of CC phase to CV phase [ 22 ], image information transformed from the entire discharging data [ 23 ], the CC charging duration or capacity [ 25 – 27 ], discharging cutoff voltage [ 27 ], voltage variation in the CC phase [ 22 , 24 , 28 ], and time interval between two predifined discharging voltage [ 29 ], etc. Unfortunately, to the best of our knowledge, little work has been performed to solve this problem.…”

Damage Diagnosis in 3D Structures Using a Novel Hybrid Multiobjective Optimization and FE Model Updating Framework

“…Computational intelligence has been implemented for solving complex prediction, diagnosis, and detection problems in various fields such as mechanical engineering [16][17][18], computer science [19,20], biomedical engineering [21,22], and electrical engineering [23]. As major disciplines of computational intelligence, swarm intelligence and EAs are powerful stochastic optimization techniques that have been implemented for solving various engineering problems in recent years.…”