Design and Experiment of Intelligent Spiral Ditching and Fertilizing Machine for Walnut Orchards in Xinjiang

Aiming at the problems that existing ditching and fertilizing machines in Xinjiang walnut orchards are generally large in size, single in function, and lack precise variable-rate fertilization based on the individual size of fruit trees, this study designed and developed a small intelligent spiral ditching and fertilizing machine integrating ditching, fertilization, soil-fertilizer mixing, and soil covering. The core goal is to meet the agronomic requirements of narrow-row orchards in arid and semi-arid regions, realizing targeted nutrient supply for the optimal fertilizer-absorbing root zone of each walnut tree. Equipped with an intelligent control system based on multi-sensor fusion (Light Detection and Ranging sensor and vision camera), the machine can real-time detect trunk diameter and tree-machine distance, and dynamically adjust the lateral offset of the ditching tool, ditching depth (300～350mm), and fertilization rate accordingly, solving the problem of blind fertilization in traditional equipment. The research integrated mechanical design, theoretical analysis, and advanced simulation. Based on kinematic analysis of the ditching tool (trajectory as a cycloid) and dynamic analysis of fertilizer particle interactions with soil and tools, the key components including the spiral ditching tool and screw conveyor auger were designed. To accurately simulate the interaction with local sandy soil (moisture content of 6% and density of 1.638g/cm³), a high-fidelity discrete element method simulation model was established using the Hertz-Mindlin (no slip) contact model, with material parameters calibrated via field samples and relevant literature. Single-factor simulations were first conducted to determine feasible operating ranges: machine forward speed of 0.1～0.3m/s and ditching tool rotational speed of 400～600r/min. Three types of spiral tools (E, F, G) with different structural parameters (helix angle, blade length, tooth thickness, and number of teeth) were designed for comparative analysis. Subsequently, a Box-Behnken experimental design was adopted for multi-objective optimization, targeting minimum ditching power consumption and minimum coefficient of variation of fertilization uniformity. Analysis of variance of the experimental data revealed significant influences of tool rotational speed, forward speed, tool type, and their interactions on operational performance. The optimal parameter combination was obtained: ditching tool rotational speed of 460r/min, machine forward speed of 0.12m/s, and type F spiral tool (helix angle 30°, radius 170mm, tooth thickness 4mm, blade length 20 mm, number of teeth 8). Discrete element method simulation predicted a ditching power consumption of 4.32kW and a coefficient of variation of fertilization uniformity of 4.85% under this combination. Field verification tests were carried out in a walnut orchard in Yecheng, Xinjiang, in mid-April 2025. With 30 measurement points and three repeated runs, the prototype achieved a ditching depth stability coefficient of 97.43%, a ditch bottom width consistency coefficient of 96.84%, a coefficient of variation of fertilization uniformity of 5.21%, and an actual ditching power consumption of 4.78 kW. All indicators fully met national agronomic standards. The relative errors between simulation and field tests were 6.91% for power consumption and 9.62% for uniformity coefficient of variation, verifying the reliability of the optimization model. Comparative tests with traditional machines showed that the intelligent variable-rate system increased fertilizer use efficiency from 73.43% to 87.49%, a significant improvement of 14.06%, attributed to concentrating fertilizer in effective root zones and reducing inter-row waste. This study successfully developed an intelligent spiral ditching and fertilizing machine suitable for Xinjiang walnut orchards. The integration of discrete element method simulation, response surface methodology-guided multi-objective parameter optimization, and multi-sensor fusion control achieved significant improvements in operational stability, fertilization uniformity, and fertilizer use efficiency compared with traditional equipment, providing a feasible technical solution for the mechanization and intelligentization of fertilization in arid and semi-arid orchard systems. This study provides a referenceable solution for promoting the mechanization, precision of orchard management and resource-efficient fertilization practices in arid regions.