An improved operating strategy of genetic operator sorting genetic algorithm (GA) was used for the multi-objective optimization of crane's lifting path planning in three-dimensional space. First, space path planning of multiple degrees of freedom was transformed into that of planar path points planning by constructing mathematical models of crane's lifting scene and pose space. Second, the optimization targets, including the shortest lifting path, best security, minimum of movement changes, were determined. Then, suitable direction and step size of the insert operator and mutation operator were selected with the addition of a memory operator for multi-objective optimization. Experiments have demonstrated that this algorithm can take several factors into account and generate multiple paths to select at one time.