SOC estimation method based on fractional order extended Kalman filtering algorithm

Abstract

The invention discloses an SOC estimation method based on fractional order extended Kalman filtering. The method comprises the following steps: establishing a lithium battery fractional order second-order equivalent circuit model; determining a function relationship between each parameter of the circuit and the SOC, and establishing a state-space equation of the lithium battery; firstly, initializing parameters, and adopting a self-adaptive genetic algorithm to carry out parameter identification on fractional order model parameters; and after the battery fractional order model is identified, carrying out SOC estimation by adopting a fractional order extended Kalman filtering algorithm. According to the method, parameter identification is performed on the fractional order model through the adaptive genetic algorithm, the lithium battery SOC is estimated in combination with the fractional order extended Kalman filtering algorithm, the problems that the integer order model is not accurate enough and the working condition characteristics of the battery cannot be well described are solved, the fractional order extended Kalman filtering algorithm is combined, information of past data is utilized, and the precision and robustness of SOC estimation of the lithium battery are improved.

Description

technical field [0001] The invention relates to the field of state of charge estimation of lithium batteries, in particular to an SOC estimation method based on a fractional order extended Kalman filter algorithm. Background technique [0002] With the introduction of the concept of clean and green energy, electric vehicles have developed rapidly, and lithium batteries are the core of energy for electric vehicles, and research on lithium batteries has become a current hot spot. [0003] At present, lithium battery SOC estimation mainly includes traditional methods based on battery characteristics such as the ampere-hour integration method, data-driven methods such as neural networks, methods based on battery models and observer techniques, and methods based on model and observer techniques are the most researched. Widely, the equivalent circuit model of the lithium battery is mainly used in combination with the Kalman filter technology to estimate the battery SOC. For the eq...

AI Technical Summary

Problems solved by technology

[0003] At present, lithium battery SOC estimation mainly includes traditional methods based on battery characteristics such as the ampere-hour integration method, data-driven methods such as neural networks, methods based on battery models and observer techniques, and methods based on model and observer techniques are the most researched. Widely, the equivalent circuit model of the lithium battery is mainly used in combination with the Kalman filter technology to estimate the battery SOC. For the equivalent circuit model of the lithium battery, the integer-order equivalent circuit model is the most studied; but the integer-order equivalent circuit model does not It fits the actual characteristics of lithium batteries, and the selection error of different models is large; with the application and development of fractional calculus theory, fractional order systems are more suitable for nonlinear systems with hysteresis effects such as lithium batteries

Images