Coupling estimation of SOC and SOH of backup power supply based on deep learning

Abstract:
Backup power supply is very important to ensure the normal operation of underground safety system. State of charge (SOC) and state of Health (SOH) are the key state quantities of backup power supply. At present, they are decoupled and estimated, and SOH depends on experience; However, the current SOC estimation method has low accuracy and does not consider SOH. There will be a huge difference in accuracy before and after the service life of backup power battery. In order to realize the coupling estimation of power SOC and SOH, this paper proposes a joint estimation method of SOC and SOH based on deep learning by analyzing the correlation between backup power SOC and SOH. Through the gated cycle unit cycle neural network and convolution neural network, the voltage, current and temperature of the backup power supply are used to realize the simultaneous estimation of SOC and SOH in the whole service life of the power supply, and the estimated value of SOH is taken into account in the SOC estimation to eliminate the negative impact of the aging factor of the power supply battery on the SOC estimation, so as to improve the estimation accuracy, Realize reliable and accurate self state perception of backup power supply.