Ant Colony Particle Filter Based State of Energy Estimation Method for Lithium-ion Batteries

In orderto estimatethe state of energy (SOE), an improved algorithm based on particle filter (PF) is proposed in this paper. The ant colony adaptive particle filter (AAPF) combines the ant colony algorithm and adaptive algorithm. It employs an ant colony algorithm to optimize the resampling process of particle filer and adaptive algorithm to reduce the influence of noise from the system. Firstly, an equivalent circuit model is built, and thenits parameters are calculated by the least-square algorithm. Thirdly, the ant colony particle filter (APF) and AAPF are proposed to optimize the resampling of the PF and estimate SOE. Finally, the accuracies of the PF, the extended Kalman filter (EKF), the APF and the AAPF are testedby estimating SOE on driving cycle tests. Experimental results indicate that compared with the EKF and the PF, the errors of the APF decrease by 5% and 1.8% respectively. Combining the APF with the adaptive algorithm, the average error of the AAPF declines byabout 1 % on driving cycle tests.