SOC estimation method and system based on BP neural network, terminal equipment and readable storage medium

Abstract

The invention discloses an SOC estimation method and system based on a BP neural network, terminal equipment and a readable storage medium. The estimation method comprises the steps: S1, setting a temperature gradient, and carrying out the discharge operation at each temperature so as to collect the sample data of a battery at each temperature, wherein the battery characteristic quantity in the sample data at least comprises a temperature, a temperature change rate and a residual electric quantity at the previous moment; and S2, training a BP neural network based on the sample data collected in the step S1 to obtain an SOC prediction model, wherein the battery characteristic quantity of the to-be-detected battery is input into the SOC prediction model to obtain an SOC value. According to the invention, the relevance between the low-temperature environment and the SOC and the relevance between the SOC and the temperature environment change are introduced into the model by utilizing the temperature change rate and the characteristic quantity of the temperature so that the accuracy of the constructed SOC prediction model is greatly improved, and the situation that the SOC generates a relatively large error due to the change of the actual residual electric quantity under the temperature change is greatly improved.

Description

technical field [0001] The invention belongs to the field of battery management systems (BMS), and in particular relates to an SOC estimation method, system, terminal equipment and readable storage medium based on a BP neural network. Background technique [0002] The BMS system has important functions such as real-time monitoring of battery status information, analysis of battery safety performance, optimization of battery energy control, and prolongation of battery life. It is an important system for ensuring battery safety and regulating battery performance. Among them, the core of battery state of charge monitoring lies in the estimation of SOC. The difficulty of SOC monitoring is that it is impossible to directly measure the state of charge of the battery accurately. The commonly used estimation methods include ampere-hour integral method, open circuit voltage method, Kalman filter method, Extended Kalman filter method and neural network method, etc. At the same time, ...

AI Technical Summary

Problems solved by technology

At the same time, due to the obvious attenuation of battery capacity at low temperatures, the SOC detection deviates from the actual situation, and there is currently a lack of a reliable technical means to determine the SOC value at low temperatures

In different temperature environments, the SOC error is large, resulting in a decrease in the reliability of the BMS system

Images