Optimal-Sensing-Based Recursive Estimation for Temperature Distribution of Pouch-Type Batteries

Real-time monitoring of the battery temperature distribution is essential in developing advanced thermal management systems. This work proposes a recursive temperature field estimation methodology for pouch-type batteries under optimal deployment of limited sensing. First, we utilize the time/space spectral expansion to mathematically derive a physics-based dynamic model considering uncertainty from the complex thermal process. Then we find the optimal sensor placement through the observability degree of the system quantified by the attenuation rate of estimation error. Finally, we perform model-based temperature field estimation, whose performance is boosted by favorable measurements from layout-optimized sensors. Experimental studies indicate the validity of the proposed method.