Dynamic Bayesian network-based disassembly sequencing optimization for electric vehicle battery

The sharply increasing end-of-life (EOF) battery volume in the global complex energy market has created significant challenges for its recycling and reuse, to reduce environmental pollution and resource waste, and efforts have been focused on the disassembly process considering the uncertainty of electric vehicle (EV) battery pack categories and quality. Compared with traditional disassembly, the EV battery disassembly process needs to consider more uncertainty factors for each EOF battery pack to represent its disassembly structure, which significantly reduces disassembly production efficiency. Even though sequence optimization methods for the disassembly process have been developed to solve these problems, there are still two important challenges that remain: uncertain disassembly structure representation and optimal disassembly sequence selection. To address these challenges, this paper proposes a dynamic disassembly Bayesian network approach based on an EV battery disassembly graph model. This method offers dynamic process optimization to manufacturers to deduce the optimal disassembly sequences using the forward–backward algorithm and the Viterbi decoding algorithm. To validate the proposed method, an EOF battery is used to demonstrate the disassembly sequence selection, which indicates the possibility of massive EV battery disassembly prediction.