Designing a battery state estimator considering the challenges for real-world applications

Given the extensive use of batteries in numerous applications, such as portable small devices, electric vehicles, and energy storage in the power industry, novel methods must be developed to supervise the battery state. This is necessary to ensure optimal energy management in battery-reliant processes. This research introduces a comprehensive approach that harnesses data from the battery manufacturer and current and voltage measurements from a pulsed full discharge test to design estimators and observers for determining: the battery backup period, the state of charge (SOC), the state of health (SOH), and the battery capacity decay rate. The innovative method was designed considering attributes that allow it to be embedded in real-time platforms. These attributes include using general specifications provided by the manufacturer and requiring no extensive measurements and minimal computational resources to provide an accurate estimation and fully viable implementation into any battery management system. The method was validated using voltage and current measurements obtained during an 87-day test designed to emulate battery aging conditions similar to those in real-time applications.