Towards Composing Efficient FPGA-Based Hardware Accelerators for Physics-Based Model Predictive Control Smart Sensor for HEV Battery Cell Management

In the era of climate change, and with rapid depletion fossil resources, efficient sustainable transportation systems, such as hybrid electric vehicles (HEVs), are becoming imperative. The cornerstone HEVs is battery technology, in order to extend useful life provide improved performance fuel technology. Model predictive control (MPC) an effective technique for management systems (BMS), which enables cell-level monitoring controlling pack, reducing safety margin operation, while maintaining requirements. Furthermore, utilizing physics-based model (PBM) allows system operate on chemical physical process that root cause degradation. non-linear extended Kalman filter (EKF) serves state observer monitor properties/processes each cell, since these internal processes battery, making them physically unobservable. amalgamation aforementioned techniques, i.e., MPC, PBM, EKF, can cell/pack, especially lithium-ion batteries. real-world scenarios, smart applications often require portable BMS, compelling BMS be executed highly constrained embedded devices. Also, inherent adaptive (PB) MPC uniquely suited systems/applications. However, high computational complexity PB-MPC, comprising PB-EKF, prevents it from being realized resource-constrained devices, infeasible BMS. this research work, we introduce novel, unique, FPGA-based hardware accelerator PB-MPC sensor (comprising PB-EKF) specifically by addressing PBM. Our proposed achieved 58 times speedup compared its software counterpart, maintain a small footprint required systems. This us manage more cells single chip processor-based solutions.