【作者】 刘阳春；

【导师】 赵化平； 张小超；

【作者基本信息】 中国农业机械化科学研究院 ， 机械设计及理论， 2012， 博士

【摘要】 本研究综合应用变量作业、机电一体化控制、传感器信号采集与处理、系统分析集成等技术思想和方法,设计可实现三种不同肥料变量施用的变量配肥施肥机械装置,在此基础上,研究开发变量配肥施肥作业控制系统。通过对系统进行静态标定试验、动态试验和田间试验,进行系统控制精度分析研究。研究土壤采样策略,对土壤进行合理采样并生成处方图,变量配肥施肥作业系统根据处方图进行实际田间作业,并与常规施肥作业结果进行了比较分析。主要研究内容包括：1.提出了氮、磷、钾三种主要肥料的自动配比技术方案,并开发了变量配肥施肥装置。设计安装了三个独立的肥料箱,三个肥箱的排肥轴能够独立控制,使得三个肥箱中不同的肥料能够按照不同配比施肥,达到变量配肥的目的。2.提出了基于称重反馈的伺服电机控制排肥机构方案,达到自动调节施肥量的目的。设计开发了变量配肥施肥控制系统以及机具前进速度检测模块。对伺服电机控制的排肥机构进行性能试验,试验结果表明：控制电机的转速对排肥机构的排肥性能有显著影响。对于尿素、磷酸二铵和硫酸钾等三种颗粒肥料而言,随着排肥轴转速加大,排肥量呈线性增加,说明改变电机转速是调节排肥量的有效措施。对伺服电机的控制精度进行了检测试验,试验结果表明：当肥箱中肥料称量值达到施肥量值时,变量配肥施肥控制器能控制电机自动停止运行,且施肥精度≥95%。3.开发基于X86架构的Atom主板为核心,集成存储、显示、输入输出等外设功能部件,扩展CAN总线通信接口的嵌入式农机机载作业控制终端设备。按照面向对象的设计思想,结合变量配肥施肥作业控制系统的需要,设计开发了变量配肥施肥作业控制软件。构建了远程数据管理平台,实现用户通过网络访问该数据管理平台,了解农田作业机械的作业情况和信息。4.通过分析影响采样密度的因素,结合试验地的地块情况,制定了一个适合本论文研制的变量配肥施肥装置的土壤采样策略,即按照10m*10m间隔进行栅格采样。对土壤养分数据进行常规分析,从均值来看,试验地块碱解氮含量超过了高量指标(150mg/kg),为1级肥力水平；有效磷为2级肥力水平(20-40mg/kg)；速效钾含量较低,为6级肥力水平(<30mg/kg).从变异系数看,碱解氮、有效磷和速效钾的变异系数都不高,属于中等变异强度,说明试验地块的养分含量相对稳定。采用Kriging方法对未测点参数值进行最优估值,然后采用ArcGIS进行绘图,获得了各养分含量在田间的空间变化图。根据黑龙江农垦对红星农场的玉米和大豆施肥的指导意见,结合土壤养分分布数据,得出配肥施肥指导处方,生成了变量配肥施肥处方图。5.对变量配肥施肥作业系统进行了田间作业试验,选择9个不同的施肥量等级和不同的N、P、K配肥比,进行了施肥试验。各次施肥误差都比较低,对每个肥箱按配比施肥的控制精度大于90%。对以上施肥试验还进行了重复性试验,每次施肥试验的平均误差都比较低,变异系数也不大,说明施肥试验的可重复性很好。在黑龙江红星农场进行了大豆和玉米的变量配肥施肥试验,按照配方施肥的方法种植玉米面积1.89hm2,种植大豆面积10.23hm2。在同一地号的田地上,按照常规种植模式,种植大豆面积10hm2,剩余面积种植玉米。精准农业测土配方施肥,磷酸二铵和硫酸钾的总体施肥量稍高于常规施肥量,但尿素的施用量基本节省一半。精准农业变量配肥施肥模式大豆总产量20930kg,单产2045.9kg/hm2；常规种植的大豆总产量19440kg,单产1997.9kg/hm2.精准农业玉米总产量24630kg,单产9735.2kg/hm2:常规种植的玉米总产量80770kg,单产8905.2kg/hm2.更多还原

【Abstract】 In this study, many ideas and methods such as variable operations, electromechanical integration control, sensor signal acquisition and processing, systematical analysis and integration were used. The fertilizer machine allows for variable application of three different fertilizer was designed. On this basis, variable fertilizer operating control system was developed. By the system static calibration tests, dynamic tests and field trials, the system control accuracy was analyzed. After doing research of sampling strategy, the soil was sampled reasonably and then a prescription map was generated. Variable fertilizer operating system was used to do the actual field work according to the prescription map, and a comparative analysis of operating results was done with conventional fertilization. Main results are as follows:1. According to crop demand for three main fertilizer nitrogen, phosphorus, potassium, the variable fertilizer device which has three separate manure containers was designed. The three fertilize axises can be independently controlled, so the different fertilizer in three manure containers can be fertilized in accordance with the different ratio.2. A program about using the servo motor to control the fertilize axis based on the weighing of fertilizer was proposed. The servo motor control system was processed and trial-produced. A variable manure fertilization control system and equipment forward speed detection module were designed and developed. Performance test on the institutions of the servo motor control system was done. The results showed that:Controlling the motor speed has a significant effect of fertilize performance. For the three granular fertilizer as urea, diammonium phosphate and potassium sulfate, the fertilizing amount increased linearly with the fertilize axis speed increase. That means changing the motor speed is an effective measure to adjust the fertilizing amount. Experiments were done to test the servo motor control accuracy. The results showed that: When the fertilizer measured value reaches the preset fertilization value, the variable manure fertilization controller can control the motor automatically stops running, and fertilization accuracy≥95%.3. The embedded agricultural airborne operations control terminal equipment which taking the Atom motherboard based on X86architecture as the core, integrating storage, display, input and output peripheral features and extending the CAN bus communication interface was developed. In accordance with the design of object-oriented thinking, combined variable fertilizer operation control system needs, the variable fertilization control software was designed and developed. The remote data management platform based on Spring3.0MVC framework was developed. Through the network to access the data management platform, users can learn about the mechanical operation of agricultural operations and information.4. Through the analysis of the sampling density factors, combined with the condition of plots, an appropriate soil sampling strategies was developed in accordance with the10m*10m interval grid sampling. After doing routine analysis of soil nutrient data, from the mean we know that the test plots nitrogen content than high-volume index (150mg/kg), is one fertility level; available phosphorus fertility level is two (20-40mg/kg); potassium content is low, for the six fertility levels (<30mg/kg). See from the coefficient of variation, the coefficients of variation of nitrogen, available phosphorus and potassium are all not high, belong to the moderate degree. It indicates that the nutrient content of the test plots is relatively stable. The Kriging method was used to do optimal valuation of unmeasured point parameter values, and then the ArcGIS software was used to plot. A nutrient content map in the space of the field was getted. According to the corn and soy fertilization direction of hongxing farm given by heilongjiang agricultural reclamation, combining with the soil nutrient distribution data, fertilization guidance prescription was concluded, and a variable fertilizer fertilization prescription map was generated.5. Variable manure fertilization operating system was used to do the field operation test. The fertilization experiments selected nine different fertilizer levels and different N, P, K fertilization ratio. Various fertilization errors are relatively low, each fertilizer box control precision of the fertilization ratio is more than90%. Those fertilization experiments were done repeatability. The results showed that the average error of each fertilization test were relatively low, the coefficient of variation was not large. Variable fertilization experiments of soybeans and corn were done in Heilongjiang Hongxing Farm. Accordance with the method of variables fertilization we growed corn area of1.89hectares and planted soybean area of10.23hectares. Fields in the same place, in accordance with the conventional cropping patterns, we planted soybean area of10hectares and planted corn on the remaining area. The diammonium phosphate and potassium sulfate fertilizer amount according to the precision agriculture variable fertilization method were slightly higher than the conventional fertilization amount, but the application rate of urea basic saved half. The soybean total output was41,860pounds, the yield was2045.9kg/ha on the precision agriculture variable fertilization mode. The soybean total output was38,880pounds, the yield was1997.9kg/ha on the conventional fertilization mode. The corn total output was24,630pounds, the yield was9735.2kg/ha on the precision agriculture variable fertilization mode. The corn total output was80,770pounds, the yield was8905.2kg/ha on the conventional fertilization mode.更多还原