Dynamic identification of influential nodes in complex networks is of great significance for practical applications. In real-world scenarios, resources are often limited, making it necessary to evaluate nodes by iteratively assessing the remaining network after removing certain nodes. Therefore, a dynamic identification method for important nodes in complex networks is more suitable for real-world applications. This paper proposes a method that combines both local and global characteristics. For the global features, we introduce an improved k-shell method that integrates the fusion degree, enhancing the resolution of node rankings. For the local features, we introduce the Solton factor and the improved network constraint coefficient (INCC) to enhance the algorithm's understanding of the relationship between neighboring nodes. Through a comparison with existing methods, we find that the proposed KPDN-INCC method complements the KPDN method and accurately identifies important nodes, thus facilitating rapid network disintegration. The experiments on artificial networks further validate the effectiveness of the proposed method in identifying important nodes in small-world networks with a random parameter less than 0.4.