Lithium-ion batteries are widely used in energy storage and power systems due to their high energy density, long cycle life, and stability. Accurate prediction of the state of health (SOH) of batteries is critical to ensuring their safe and reliable operation. However, the prediction task remains challenging due to various complex factors. This paper proposes a hybrid TCN–Transformer–BiLSTM prediction model for battery SOH estimation. The model is first validated using the NASA public dataset, followed by further verification with dynamic operating condition simulation experimental data. Health features correlated with SOH are identified through Pearson analysis, and comparisons are conducted with existing LSTM, GRU, and BiLSTM methods. Experimental results demonstrate that the proposed model achieves outstanding performance across multiple datasets, with root mean square error (RMSE) values consistently below 2% and even below 1% in specific cases. Furthermore, the model maintains high prediction accuracy even when trained with only 50% of the data.

Due to the complex linear relationships within a battery, the prediction of state of charge (SOC) has become a significant challenge. This paper designs a three-layer backpropagation (BP) neural network to predict the battery’s SOC. To enhance the prediction accuracy, based on the charge–discharge mechanism of lithium batteries, the battery’s voltage, current, and temperature are selected as input variables, with the state of charge as the output variable. Under three different operating conditions, the BP neural network can achieve certain predictive effects. However, due to the instability and large error fluctuations of the BP neural network during the SOC prediction process, the BP neural network is optimized. The Sparrow Search Algorithm (SSA) can enhance the model’s ability to search for optimal values. By improving the Sparrow Search Algorithm, the uniform distribution of the initial population can be enhanced. The prediction results after optimizing the BP network show that, compared to the unimproved version, the stability of the prediction results can be improved, achieving more accurate predictions of SOC values, which has strong application value.