Simulation Study on Coverage Path Planning of Autonomous Tasks in Hilly Farmland Based on Energy Consumption Model

Abstract EN

The hilly farmland in China is characterized by small farmland areas and dense farmland distribution, and the working environment is three-dimensional topographic farmland, so the working conditions in the field are relatively complex.

In this working environment, the coverage path planning technique of a farmland autonomous task is harder than that of 2D farmland autonomous task.

Generally, the path planning problem of 2D farmland is to construct the path cost model to realize the planning of agricultural machinery driving route, while for the path planning problem of three-dimensional terrain farmland in the hilly region, this paper proposes a covering path planning scheme that meets the requirements of autonomous work.

Based on the energy consumption model, the scheme searches the optimal driving angle of agricultural machinery, prioritizes solutions to the problem of covering path planning within the scattered fields in the working area, and then searches through the genetic algorithm for the optimal order of traversing the paths of each field to complete the coverage path planning in the working area.

On the one hand, the scheme optimizes the planning route in the fields from the angle of optimal energy consumption; on the other hand, through the genetic algorithm, the fields are connected in an orderly manner, which solves the comprehensive problems brought by the unique agricultural environment and farming system in China’s hilly areas to the agricultural machinery operation.

The algorithm program is developed according to the research content, and a series of simulation experiments are carried out based on the program using actual farmland data and agricultural machinery parameters.

The results show that the planned path obtained at the cost of energy consumption has a total energy consumption of 4771897.17J, which is 17.4% less energy consumption than the optimal path found by the path cost search; the optimization effect is evident.