Abstract:

Because of the complex working environment of special robots, a state of charge (SOC) estimation method with high precision and strong tracking ability is required for real-time state monitoring and safety control of lithium-ion batteries of special robots. SOC is one of the most important parameters in battery management systems. The working environment of special robots has strong nonlinear characteristics, considering that the commonly used ampere-hour integration method depends heavily on the accuracy of the initial SOC and accumulates errors in the later stages of estimation. Therefore, considering the working characteristics of special robots, a ternary lithium-ion battery is taken as the research object based on the Thevenin equivalent circuit model and experiments under various operating conditions. An improved extended Kalman filter (IEKF) algorithm is used to estimate the SOC of lithium-ion batteries at 10 ℃, 25 ℃, and 35 ℃. The simulation model is built in MATLAB/Simulink and combined with various working condition data for performance analysis. The experimental results show that using the IEKF algorithm to estimate the SOC value of ternary lithium-ion batteries has a good tracking and convergence effect, and the convergence time is within 80 s. At different temperatures, the maximum estimation errors of the HPPC and BBDST conditions are less than 2.235% and 3.004%, respectively, after convergence, which 9.067% and 4.654% less than those of the corresponding maximum estimation errors of the extended Kalman filter (EKF) algorithm. This study verifies that the IEKF algorithm has high accuracy in estimating the SOC of lithium-ion batteries, and it provides an experimental basis for effectively resolving the inaccurate estimation of the SOC values of lithium-ion batteries of special robots.

Key words: special robots, lithium-ion batteries, Thevenin equivalent circuit model, state of charge, improved extended Kalman filter algorithm, different temperatures