Based on the analysis of relationship between the crossover scale and reachable subspace of crossover operator, it can be found that the crossover scale is dynamically adjusted to the population structure. In this paper,three control mechanisms—the well-phased control strategy, the random distribution strategy and the adaptationevolution strategy are built up to adjust the crossover scale. The simulation tests of the classical function show theseoptimization mechanisms are available and valuable control knowledge of crossover scale for multi-dimensionfunctions have been generated by the adaptation evolution strategy. Furthermore, this research suggests a newmethod for the operator and parameter optimization of evolution algorithm.