Abstract: To reduce the cost of converting farm roads in hilly and mountainous areas under the "household mechanization" model and shorten the transformation time, a method is proposed to plan the layout of farm roads that connects all plots in a given area and has the shortest length. This method is based on set coverings and minimum Steiner trees. It divides the entire planning process into two stages. First, it uses a set covering problem model to find all intersection combinations that connect all plots. Then, it treats these intersection groups as target nodes to be connected and finds the corresponding minimum Steiner tree in an undirected graph formed by plot boundary information. By comparing the network lengths of different results, the layout with the shortest total length of farm roads is determined. Experimental verification results show that this method is feasible and can reduce the cost of converting farm roads.