Table 1 Current status of domestic and international research in path planning for land leveling path planning.

Liu et al.

Taking the shortest shovel empty or full load time as the optimal evaluation benchmark, clustering farmland grids, and according to the designed shovel dig-high-fill-low method and local search strategy to get the global planning of levelling paths

Only the shortest empty and full load time of the grading shovel was used as an evaluation index, without considering the efficiency of earth volume transport, and the path planning time was long

Path planning for flatland operations is a multi-objective optimization problem

Kang et al.

A global path planning method was proposed, which could plan effective paths and reduce the ineffective operation time to some extent

The least ineffective operation status of empty and full load, the least steering operation and repeated travelling were taken as evaluation indexes, the combined load ratio of the levelling shovel was not taken into account, and the distance of the planned path was long

Path planning for flatland operations is a multi-objective optimization problem

Jin et al.

A three-dimensional path planning method based on improved ant colony algorithm for farmland levelling navigation was proposed for the purpose of reasonable earth transport and unloading with the shortest path in levelling operation

Only able to plan paths in a single pass, with low operational efficiency

Dynamic continuous planning of paths is required for efficient levelling

Ojima et al.

It was demonstrated that the dynamic programming algorithm could be applied to the land levelling problem and dynamically adjusted the planning path according to the load of the levelling shovel

The arithmetic power increased abruptly with more nodes

Fields need to be zoned first to reduce planning time

Jeon et al.

Genetic algorithm was used to determine the optimal sequence of internal trajectories in a paddy field by minimizing the turning distance at the head of the field and traversing the entire field

It was easy to fall into local optimal solution, a flaw of the genetic algorithm itself

The selection of path planning algorithm needs to consider high global search capability and robustness

Jing et al.

A tri-objective (travel distance, steering angle, earthwork) problem was modelled and an improved (model decomposition and further mutation ant colony) multi-objective evolutionary algorithm was used to optimize the grading path

Failure to include grading shovel load characteristics as one of the optimization objectives to increase the number of repetitive round trips

Path planning for flatland operations is a multi-objective optimization problem