Authors

Abstract

Precision agriculture requires accurate and efficient crop yield distribution information. However, both traditional field-based yield measurement methods and existing combine harvester yield monitoring systems face significant limitations. Traditional methods, such as direct weighing or sampling, are time-consuming and inefficient, and they only provide average yield values - insufficient for large-scale farming needs. Meanwhile, current monitoring systems often suffer from high measurement errors, low spatial resolution, and limited generalizability. For this reason, this study designs a new type of grain yield monitoring system, which corrects the photoelectric sensor data through the load cell data, realizes the calibration of the photoelectric sensor, avoids the influence of external factors, and improves the accuracy of measurement. Firstly, tests were carried out at three rotational speeds of 10 Hz, 20 Hz and 25 Hz of the motor inverter setting, respectively, to determine the positive proportionality coefficient between the photoelectric signal and the grain mass, and the overall error of the system was measured to be less than 6.44%. For the load cell, a model of the relationship between tilt angle and weighing accuracy was established and a compensation algorithm was proposed, the weighing error data in different directions and at different tilt angles were measured and analyzed, and a mathematical model between the corrected angle and the weighing error was established. Through the tilting experiment, the feasibility of the modified angle compensation model is verified, and the overall error after compensation is less than 0.25%, and the systematic error of measurement and production after the intervention of the feedback system is less than 0.74%. The experimental results demonstrate that the system significantly enhances the accuracy and stability of yield measurement. It holds substantial potential for widespread application, provides strong support for the advancement of precision agriculture, and is expected to drive agricultural production toward greater efficiency and sustainability.

Abstract in Chinese

精准农业需要准确且高效的作物产量分布信息，但传统的田间作物产量获取方法以及现有的联合收割机产量监测系统存在诸多问题。传统方法，如直接称重或抽样，不仅耗时且效率低下，而且只能获取平均产量，难以满足大面积种植的需求，而现有的系统测量误差大、空间分辨率低且通用性差。为此，本研究设计了一种新型的谷物产量监测系统，该系统通过称重传感器数据对光电传感器数据进行校正，实现了对光电传感器的校准，避免了外部因素的影响，提高了测量精度。首先，分别在电机变频器设定的10赫兹、20赫兹和25赫兹这三种转速下进行测试，以确定光电信号与谷物质量之间的正比例系数，经测量该系统的整体误差小于6.44%。对于称重传感器，建立了倾斜角度与称重精度之间关系的模型，并提出了一种补偿算法，测量并分析了不同方向和不同倾斜角度下的称重误差数据，建立了校正角度与称重误差之间的数学模型。通过倾斜实验，验证了修正后的角度补偿模型的可行性，补偿后的整体误差小于0.25%，反馈系统介入后测量和生产的系统误差小于0.74%。实验结果表明，该系统有效地提高了产量测量的准确性和稳定性，具有广泛的应用价值，为精准农业的发展提供了有力支持，有望推动农业生产朝着更高效、更可持续的方向发展。