Diagnosing potential faults is of great importance to ensure reliability of battery management systems. This is because a current or voltage sensor fault often results in an inaccurate state-of-charge estimate. A temperature sensor fault will cause abnormal thermal management. A battery internal resistance (BIR) fault can lead to an increase in energy and power losses, capacity fading, and further degradation of health. In addition, frequent data transmission to fault diagnosis unit will cause a great waste of communication resources. To this end, a combined model-based and data-driven fault diagnosis scheme for lithium-ion batteries is proposed in this article. First, a model-based fault estimation method with sliding mode observer is developed to estimate the voltage, current, and temperature sensor faults. By integrating an adaptive event-triggered mechanism, the communication resource costs are alleviated. Second, a data-driven gap metric approach is presented to detect the BIR fault. By combining the model-based and data-driven strategies, the fault diagnosis logic is put forward to isolate the BIR fault. Finally, several experiments of the single cell and the battery pack are conducted to verify the effectiveness and superiority of the developed method over the existing results.