A battery state of charge estimation method and estimator

Abstract

The present application provides a battery state-of-charge estimation method and an estimator, the method comprising: determining at least two continuous state models in the battery time domain based on actual state data of the battery, the state models can be based on an input target time The current value corresponding to the output voltage value; the first processing result is obtained by processing the first voltage value based on the actual voltage value at the target time. The first voltage value is calculated by the first state model according to the current value at the target time. The processing result is used to adjust the second state model to improve the calculation accuracy of the voltage value by the second state model; adjust the second state model based on the first processing result, and obtain the effect of the second state model on the target The second voltage value outputted at the time; processing the first voltage value and the second voltage value to obtain the state of charge of the battery at the target time. The present application can estimate the state of charge of the battery in real time, efficiently and accurately.

Description

technical field [0001] The embodiments of the present application relate to the field of battery management, and in particular, to a battery state of charge estimation method and a battery state of charge estimator. Background technique [0002] In the field of battery management, the battery state estimator is a preliminary but very critical functional module, which is essential to ensure the safe operation of the battery and improve the efficiency of energy management. Among the battery states that need to be estimated, the state of charge (SOC) can reflect the remaining energy / capacity during battery use, thereby reducing user anxiety and ensuring that the battery operates in reliable conditions. Therefore, the online effective estimation of the state of charge is of great significance. [0003] At present, the estimation methods for the state of charge of power batteries mainly fall into the following categories: 1) Calculate the SOC directly from the known initial SOC ...

AI Technical Summary

Problems solved by technology

[0003] At present, there are mainly the following methods for estimating the state of charge of power batteries: 1) Using such as coulomb counting method or open circuit voltage method to directly calculate SOC from the known initial SOC state or open circuit voltage (Open circuit voltage, OCV)-SOC curve , this method is highly dependent on the setting of the initial value and is susceptible to noise interference, and cannot obtain reliable and consistent results under complex working conditions

2) Using machine learning techniques such as support vector machines, Gaussian process regression, etc. to establish data-driven models for SOC estimation, this method requires a large amount of experimental data for training and has poor performance under new working conditions

[0004] However, the state estimator needs to consume a certain amount of computing workload and device memory, and many general-purpose microcontroller units (MCU) on the market do not have complex and high-computing capabilities.

Advanced estimators such as Unscented Kalman filter (UKF) and Particle filter (PF) are difficult to achieve satisfactory battery SOC estimation results in these MCUs

Images