Estimation method of electric vehicle battery remaining power based on particle filter algorithm

As electric vehicles gain popularity, accurate power estimation can optimize vehicle energy management, extend battery life, and enhance the user’s driving experience. This paper suggests an approach for estimating the State of Charge using a particle filter, integrated with a second-order equivalent model to enhance both the accuracy and robustness of the estimation. By simulating the dynamic behavior of the battery, the particle filter produces numerous particles to depict the probability distribution of the State of Charge through nonlinear equations for state transition and observation. By constantly updating particle weights and combining them with real-time measurement data, this approach can successfully address the shortcomings of conventional estimation techniques in the presence of nonlinearity, noise, and uncertainty. The experimental findings indicate that the discrepancy between the state of charge estimated by the particle filter and the actual value has been notably decreased, which verifies the effectiveness and feasibility of this method in the EV battery control system.