In the real world, as the complexity of optimization problems continues to increase, there is an urgent need to research more efficient optimization methods. Current optimization algorithms excel in solving problems with a fixed number of dimensions. However, their efficiency in searching dynamic multi-dimensional spaces is unsatisfactory. In response to the challenge of cross-dimensional search in multi-dimensional spaces with varying numbers of dimensions, this study proposes a new optimization algorithm-Dynamic Dimension Wrapping (DDW) algorithm. Firstly, by utilizing the Dynamic Time Warping (DTW) algorithm and Euclidean distance, a mapping relationship between different time series across dimensions is established, thus creating a fitness function suitable for dimensionally dynamic multi-dimensional space. Additionally, DDW introduces a novel, more efficient cross-dimensional search mechanism for dynamic multidimensional spaces. Finally, through comparative tests with 31 optimization algorithms in dynamic multidimensional space search, the results demonstrate that DDW exhibits outstanding search efficiency and provides search results closest to the actual optimal solution.

翻译：在现实世界中，随着优化问题复杂度的持续增加，亟需研究更高效的优化方法。现有优化算法在求解维度数量固定的问题上表现优异，但在动态多维空间中的搜索效率却不尽如人意。针对维度数量变化的多维空间中跨维度搜索的挑战，本研究提出了一种新的优化算法——动态维度封装（DDW）算法。首先，通过利用动态时间规整（DTW）算法与欧氏距离，建立了跨维度不同时间序列之间的映射关系，从而构建出适用于维度动态变化的多维空间的适应度函数。此外，DDW为动态多维空间引入了一种新颖且更高效的跨维度搜索机制。最后，通过在动态多维空间搜索中与31种优化算法进行对比测试，结果表明DDW展现出卓越的搜索效率，并能提供最接近实际最优解的搜索结果。