A real-time estimator for model parameters and state of charge of lead acid batteries in photovoltaic applications

The intermittent nature of photovoltaic energy source has revealed concerns about the stability of the power electric system. For that, a massive use of storage elements becomes needed. Batteries are considered as one of the most important technologies for energy storage. In order to achieve the needs of safety, durability and reliability for the battery management system (BMS), data monitoring and diagnosis system has been discussed in this work. In fact, several methods have been presented with the intention of estimating the internal parameters of an AGM lead acid battery model such as the Recursive least square algorithm (RLS) with variable forgetting factor, a novel Adaptive Joint Extended Kalman Filter (AJEKF), and the Unscented Kalman Filter (UKF). The state of charge (SOC) of electrochemical storage device has been considered as one of the most important state which needs to be monitored in order to optimize the performance and to extend the lifetime of the system. SOC estimation is a challenging task hindered by considerable changes of the battery characteristic throughout its lifetime due to its nonlinear behavior and aging factor. In this work, online state of the charge estimation has been evaluated through the Extended Kalman Filter, Unscented Kalman Filter (UKF), and Cubature Kalman Filter (CKF). An Open circuit voltage experimental test (OCV) has been established in order to validate the operation of the proposed battery model under real work conditions. This paper proposes an exhaustive review for the online identification of battery parameters and its state of charge, where the tendency of estimation techniques has been oriented toward a mixture of different techniques that clearly raised the challenge of establishing a relationship between the accuracy and the robustness of these methods, without forgetting their low complexity of implementation.