【作者】 张峰；

【导师】 冯静安； 王玉刚；

【作者基本信息】 石河子大学 ， 机械（专业学位）， 2024， 硕士

【摘要】 全球农业正逐步向现代化转型,精准农业更是成为该领域发展趋势。在我国,新疆棉花种植规模庞大且机械化水平较高,运用机载测产手段获取棉花产量数据及其产量分布图,对于精准农业战略实施尤为关键。然而,国产采棉机测产技术起步较晚,机载测产系统开发尚未成熟。针对该问题,本文开展了基于微波检测技术的机采籽棉产量估测系统研究,搭建了测产系统试验平台,为国产采棉机测产系统开发提供了参考。论文主要研究内容如下:(1)分析机采籽棉产量估测系统的工作原理,主要包括籽棉速度估测原理与籽棉产量估测原理。随后概述系统组成,完成系统总体方案设计。(2)完成了机采籽棉产量估测系统硬件设计。该部分主要包括系统硬件功能需求分析与硬件结构设计。在硬件结构设计中,首先通过对微波传感器工作模式与谐振方式的分析,选取平面微带天线作为微波天线模块;随后制定信号调理电路设计规范,绘制相应的信号调理电路图,经仿真无误后连接微波传感器硬件电路。最后根据硬件功能需求,选取适用于本系统的数据采集模块与电源模块,从而完成了硬件系统搭建。(3)完成了机采籽棉产量估测系统软件设计。首先对系统软件功能需求进行分析,确定系统软件工作流程。其次通过对比基于离散傅里叶变换(DFT)与时域抽取基2FFT(DIT-FFT)的籽棉速度估测算法的优劣,选取了计算复杂度较低的DIT-FFT算法用于籽棉速度估测。然后对经典功率谱估计算法与现代功率谱估计算法进行研究。考虑到窗函数对经典功率谱估计算法的重要性,对四种窗函数的时域与频域特性进行仿真分析,最终选取了Hanning窗作为经典功率谱估计算法的窗函数。随后在MATLAB中对上述研究的功率谱估计算法进行仿真分析,选取了加Hanning窗的Welch方法(Hanning-Welch)作为本系统的产量估测算法。最后建立籽棉微波损耗模型来确定籽棉含水率对系统精度的影响,以便对测量信号进行补偿。(4)设计籽棉速度估测试验与籽棉产量估测试验,验证了机采籽棉产量估测系统的可行性。在试验开始前,通过微波损耗模型证明所测籽棉含水率对系统精度影响可忽略不计。在速度估测试验中,通过测量传感器不同安装角度下单朵籽棉的瞬时速度,确定了微波传感器的最佳安装角度为45度。在此基础上设计籽棉下落位置试验与大量籽棉下落速度试验,得到影响籽棉运动状态的因素主要包括管壁摩擦碰撞、空气阻力以及籽棉之间的相互作用力,其中籽棉之间的相互作用力对籽棉运动速度影响最大。在产量估测试验中,采用最小二乘法将多组籽棉质量与其对应的功率谱密度进行拟合,在两种籽棉输送条件下,分别建立自由落体籽棉质量模型和风力输送籽棉质量模型,并设计试验验证两种模型的精度。最后针对系统存在的不足之处,从理论层面提出系统改进方案。更多还原

【Abstract】 Global agriculture is gradually transforming to modernization,and precision agriculture has become the development trend in this field.In China,Xinjiang cotton planting scale and high level of mechanization,the use of airborne yield measurement means to obtain cotton yield data and its yield distribution map,especially for the implementation of precision agriculture strategy is critical.However,the domestic cotton picker yield measurement technology started late,and the development of on-board yield measurement system is not yet mature.To address this problem,this thesis carries out the research on the machine-picked seed cotton yield estimation system based on microwave detection technology,builds a yield measurement system test platform,and provides a reference for the development of yield measurement system for domestic cotton pickers.The main research content of the thesis is as follows:(1)Analyze the working principle of the machine picking seed cotton yield estimation system,mainly including the principle of seed cotton speed estimation and seed cotton yield estimation principle.Subsequently,outline the system components and complete the overall program design of the system.(2)The hardware design of machine-picked seed cotton yield estimation system was completed.This part mainly includes the system hardware function requirement analysis and hardware structure design.In the hardware structure design,firstly,through the analysis of the microwave sensor working mode and resonance mode,the planar microstrip antenna is selected as the microwave antenna module.Then,the signal conditioning circuit design specification is formulated,the corresponding signal conditioning circuit diagram is drawn,and the microwave sensor hardware circuit is connected after the simulation is correct.Finally,according to the hardware functional requirements,the data acquisition module and power supply module applicable to this system are selected,thus completing the hardware system construction.(3)The software design of the machine-picked seed cotton yield estimation system was completed.Firstly,the functional requirements of the system software were analyzed to determine the system software workflow.Secondly,by comparing the advantages and disadvantages of seed cotton velocity estimation algorithms based on discrete Fourier transform(DFT)and decimation-in-time fast Fourier transform(DIT-FFT),the DIT-FFT algorithm with lower computational complexity was selected for seed cotton velocity estimation.Then,the classical power spectrum estimation algorithm and the modern power spectrum estimation algorithm are studied.Considering the importance of the window function to the classical power spectrum estimation algorithm,the time-domain and frequency-domain characteristics of the four window functions were simulated and analyzed,and finally the Hanning window was selected as the window function of the classical power spectrum estimation algorithm.Subsequently,the power spectrum estimation algorithms studied above were simulated and analyzed in MATLAB,and the Welch method with the addition of Hanning window(Hanning-Welch)was selected as the yield estimation algorithm of this system.Finally,the seed cotton microwave loss model was developed to determine the effect of seed cotton moisture content on the accuracy of the system in order to compensate the measured signal.(4)The seed cotton velocity estimation test and seed cotton yield estimation test were designed to verify the feasibility of the machine harvested seed cotton yield estimation system.Before the test,the microwave loss model was used to demonstrate that the moisture content of the measured seed cotton has negligible effect on the accuracy of the system.In the velocity estimation test,the optimal installation angle of the microwave sensor was determined to be 45 degrees by measuring the instantaneous velocity of a single seed cotton under different installation angles of the sensor.On this basis,we designed the seed cotton falling position test and a large number of seed cotton falling velocity test,and obtained that the factors affecting the movement state of seed cotton mainly include the friction and collision of the tube wall,air resistance,and the interaction force between seed cotton,among which the interaction force between seed cotton has the greatest influence on the movement velocity of seed cotton.In the yield estimation test,the least squares method is used to fit multiple seed cotton masses to their corresponding power spectral densities,and the free-fall seed cotton mass model and the wind-transport seed cotton mass model are established under two seed cotton conveying conditions,and tests are designed to verify the accuracy of the two models.Finally,for the deficiencies of the system,the improvement program of the system is proposed from the theoretical level.更多还原