An intelligent battery equalization scheme based on fuzzy logic control is presented to adaptively control the equalizing process of series-connected lithium-ion batteries. The proposed battery equalization scheme is a bidirectional dc-dc converter with energy transferring capacitor that can be used to design the bidirectional nondissipative equalizer for a battery balancing system. Furthermore, it can be designed as a ripple-free converter for improving the input current distortion of the battery charge supply power system. A fuzzy-logic-controlled strategy is constructed with a set of membership functions to prescribe the cells equalizing behavior within a safe equalizing region for rapid cell voltage balancing. The simulation and experimental results show the advantage of the predicted equalizing performance of the lithium-ion battery stacks. The proposed fuzzy logic control battery equalization controller can abridge the equalization time about 32%. The proposed method maintains safe operation during the charge/discharge state in each lithium-ion cell of the battery strings.Index Terms-Battery charge equalization, battery management system (BMS), dc-dc power conversion, fuzzy logic, intelligent control, lithium-ion batteries.

No abstract

To solve learning problems with vast number of inputs, this paper proposes a novel learning structure merging a number of small fuzzy neural networks (FNNs) into a hierarchical learning structure called a merged-FNN. In this paper, the merged-FNN is proved to be a universal approximator. This computing approach uses a fusion of FNNs using B-spline membership functions (BMFs) with a reduced-form genetic algorithm (RGA). RGA is employed to tune all free parameters of the merged-FNN, including both the control points of the BMFs and the weights of the small FNNs. The merged-FNN can approximate a continuous nonlinear function to any desired degree of accuracy. For a practical application, a battery state-of-charge (BSOC) estimator, which is a twelve input, one output system, in a lithium-ion battery string is proposed to verify the effectiveness of the merged-FNN. From experimental results, the learning ability of the newly proposed merged-FNN with RGA is superior to that of the traditional neural networks with back-propagation learning. Index Terms-Battery state-of-charge (BSOC), battery string, B-spline membership functions (BMFs), fuzzy neural networks (FNNs), merged-FNN, reduced-form genetic algorithm (RGA).

No abstract

This study proposes Lithium-ion battery aging correction state-of-charge (SOC) estimation techniques. Although the battery is aging, the SOC error estimation system maintains the setting range using a low-cost 8 bit micro-controller. The proposed method can track and correct the open-circuit voltage against capacity in the battery management system by comparing the capacity error with the coulomb counting and look-up table methods. The experimental results verify that the SOC estimation error is still lower than 3.5% after 1000 cycles. The SOC estimation verification platform verifies the Sanyo UR18650 W lithium battery. After 300 accelerated aging cycle charge-discharge tests, the test results showed that the SOC prediction precision for an aged battery is as high as 2.67%. This BMS includes a microprocessor , temperature measuring unit, cell voltage measuring unit, current measuring unit, protection unit, communication unit, charge-discharge control unit and data storage unit. The proposed system uses a 1000-cycle battery life test using the Sanyo UR18650 W 1.5 Ah lithium-ion battery to determine the correlation between the battery freshness and OCV aging curve, as shown in Fig. 2. Fig. 2 shows that the Sanyo 18650 W cell is performed the OCV testing process after it goes through 200 cycling test. The cycling and OCV testing methods are shown below: Cycling test: † Charging condition at 25°C ± 1°C: † Constant current (CC) condition: 1.5A (1-crate) when cell voltage is <4.2 V. † Constant voltage (CV) condition: 4.2 V when cell voltage is >= 4.2 V. † Cutoff condition: charging current <30 mA (0.02-crate). † Rest condition: 20 min after charging. † Discharging condition: † CC condition: 1.5 A (1-crate). † Cutoff condition: cell voltage is < 2.8 V. † Rest condition: 20 min after charging. OCV testing condition at 25°C ± 1°C: † Charging condition: The same processes mentioned in the charging condition. † Discharging condition: † CC condition: 30 mA (0.02-crate). † Cutoff condition: cell voltage is < 2.8 V. † Sampling point: record voltage per 15 mAh (about 30 min) for 100 units. † Rest condition: 20 min after charging.