Precise algorithms for minimizing feedback in new energy product development

The rapid development of new energy technologies necessitates efficient project management methodologies. Design Structure Matrix (DSM) serves as a powerful tool to represent interdependencies among tasks, but excessive feedback loops can significantly delay development cycles. This study introduces two novel properties of the feedback minimization problem and proposes a Parallel Branch-and-Bound (PBB) algorithm to optimize the task sequence. Furthermore, we enhance the PBB algorithm by integrating hash functions, leading to a significant improvement in computational efficiency. Experimental results demonstrate that the hash-enhanced PBB algorithm can efficiently solve DSM instances with up to 40 interrelated activities within one hour, outperforming traditional methods. The proposed approach provides a precise and scalable solution for optimizing complex product development processes.