Optimizing Sampling Points and Path Planning for Soil Monitoring in Agricultural Land

Current soil mixing sampling surveys face several challenges. These challenges include significant variations in the size and shape of farmland plots, which complicate the sampling process. Additionally, the procedures for establishing mixing points are often cumbersome and non-standardized. Furthermore, when the number of mixing points is large and the plot area is extensive, sampling efficiency is significantly reduced. In this study, an automated method for establishing the layout of sampling points and a threshold-based mixed path planning algorithm are proposed. These approaches are designed to automatically arrange sampling points within delineated sampling units according to various sampling methods. Additionally, real-time path planning is conducted on the basis of the initial position of the sampling vehicle. The experimental results indicate that the maximum relative error in estimating farmland area is 0.22%, which is less than 666.67 m2. The average deviation between the locations of the sampling points generated by the sampling point layout algorithm and those established via related software is 0.0336 m. Furthermore, the threshold-based hybrid algorithm consistently yields optimal or near-optimal paths while reducing the path planning time, with an average duration of 1.0783 s for path planning. This study provides reliable technical support for standardizing the sampling point layout and enhancing the soil sampling efficiency.