Advanced Real-Time State of Charge Estimation in a Hybrid Aircraft Using Dual Filter Interacting Multiple Model

Hybrid and electric aircraft offer a viable solution to reducing the environmental impact of conventional aviation, which aligns with the industry’s intention to reduce its carbon footprint while maintaining safety and operational standards. However, current limitations in battery production technology and the insufficient robustness and accuracy of existing algorithms in Battery Management Systems (BMS) for monitoring and estimating battery states, including State of Charge (SOC) and State of Health (SOH), pose significant challenges in the adoption of electrifying technologies in aviation applications. This paper proposes an approach for SOC estimation by combining advanced model-based concepts into the aviation environment that have not been implemented before. This method is applied to an aircraft’s Energy Storage System (ESS) by integrating merged sets of third-order Equivalent Circuit Models (ECM) representing battery dynamics at the module level. A novel dual-model filtering structure is introduced, in which a Dual Filter (DF) architecture enables concurrent estimation of internal resistance and SOC bias at the module level. The DF structure is deployed based on the Smooth Variable Structure Filter (SVSF), in which one filter provides the estimated internal resistance vector, while the other filter estimates the SOC bias. These filters are embedded within the Interacting Multiple Model (IMM) framework, which lets the algorithm utilize a mixture of several model-filter structures at each time step with respect to the discharge rate. Instead of using off-line and simulated data, the developed methods are implemented and tested in real-time in the laboratory and on a hybrid electric aircraft. This work enables safer and more efficient hybrid-electric aircraft operations by reducing SOC estimation error to below 1%.