State-of-health estimation of lithium-ion batteries based on QPSO-BPNN

Lithium battery state of health (SOH) estimation is crucial to ensure the safe and reliable operation of the battery. To enhance the accuracy of lithium battery SOH estimation, a model for estimating the state of health of lithium-ion batteries based on quantum particle swarm optimization (QPSO) optimized backpropagation neural network (BPNN) was proposed. Initially, the degradation mechanism of lithium-ion batteries is analyzed. Subsequently, BPNN is utilized to learn from the database training samples, quantum particle swarm optimization is employed to determine the optimal connection threshold and weight, and the key parameters of the model are optimized by QPSO to improve the estimation accuracy of the model. Finally, the SOH estimation result of the lithium battery is obtained. The NASA public dataset was utilized to validate the model. The results indicate that the proposed model exhibits higher accuracy compared to prediction models utilizing standard particle swarm optimization, dung beetle algorithm, and seagull algorithm optimization BPNN on various lithium-ion datasets. The average absolute error, root mean square error, and average relative percentage error are maintained within 0.00805, 0.01140, and 1.30850%, respectively, demonstrating the effective enhancement of the estimation accuracy of lithium-ion battery SOH.