联合收获机谷物破碎率、含杂率监测方法及系统研究

发布时间：2018-06-10 13:54

本文选题：联合收获机 + 破碎率　；参考：《江苏大学》2017年硕士论文

【摘要】：联合收获机是我国收获水稻和小麦的主要农业机械设备,近几年得到飞速的发展。但是国内联合收获机智能化程度相对较低,普遍缺乏工作参数与作业性能监测装置,作业效率依赖机手的熟练程度,且操纵强度大,堵塞故障频发。在联合收获机作业过程中发生谷物破碎率和含杂率超标的情况时,由于缺少在线监测方法、驾驶人员经验不足等原因不能及时调整相关工作参数,一方面会给农民带来直接的经济损失,另一方面,农民用破碎率较高的谷物育种时,发芽率较低,破碎严重的会影响下季的产量。针对上述情况,本文着重研究联合收获机谷物破碎率与含杂率监测方法,并研制监测系统对联合收获机粮箱中谷物的破碎率与含杂率进行监测,监测结果通过联合收获机驾驶室内的显示屏显示,为驾驶人员合理设置联合收获机相关工作参数提供依据。本文主要研究内容如下:(1)研究谷物破碎率与含杂率的监测方法。根据联合收获机的工作流程,分析破碎率含杂率产生原因。并提出利用图像处理技术监测联合收获机谷物破碎率与含杂率的方法,通过理论分析得出破碎率与含杂率的计算模型。(2)对谷物破碎率与含杂率监测系统进行总体设计,包括系统硬件设计及系统软件设计。系统硬件设计包括谷物在线采集装置的设计、采集装置照明系统的设计、图像采集模块的选择及系统处理器模块的选择。其中,谷物在线采集装置选择安装在联合收获机升运器出口处,并通过微型电磁铁控制其底板的闭合,保证谷物采集装置能够周期性地采集并释放谷物。(3)研究出谷物破碎率与含杂率监测算法。着重分析谷物图像分割算法,提出了利用K-Means算法对破碎谷物与杂余进行粗提取,再通过分水岭算法进行细分割;提取出杂余与谷物的特征值,作为BP神经网络模型输入参数,识别出杂余、破碎谷物与完好谷物;同时针对设计的算法,分别在Windows平台与ARM平台进行了软件设计,并给出了具体实现步骤及部分代码。(4)在联合收获机试验台上进行了台架实验,验证了系统装置以及算法的可行性。同时,对比不同图像传感器的性价比以及采集的图像质量,分别采用摄像头与工业相机在台架试验中采集图像,得到两种情况下破碎率与含杂率的识别率。台架试验表明,采用工业相机采集谷物图像时,破碎率与含杂率的平均识别率分别为88.96%、88.71%;采用摄像头采集谷物图像时,破碎率与含杂率的平均识别率分别为68.76%、66.6%。(5)将设计的算法移植到嵌入式平台,并将基于嵌入式平台的的谷物破碎率与含杂率监测系统安装在联合收获机上,进行田间在线监测试验。试验表明,联合收获机谷物破碎率、含杂率监测系统的谷物破碎率平均识别率为86.63%;忽略毫米级的微小杂余后,谷物含杂率的平均识别率为85.62%。谷物破碎率与含杂率的监测结果能够通过联合收获机驾驶室内的显示屏显示,为驾驶人员及时调整联合收获机工作参数提供依据,满足了联合收获机监测系统设计的预期要求,为实现联合收获机自动化控制提供了技术支撑。
[Abstract]:Combined harvester is the main agricultural machinery equipment for harvesting rice and wheat in China, and it has developed rapidly in recent years. However, the intelligent degree of combined harvesters in China is relatively low, and there is a general lack of working parameters and performance monitoring devices. The operation efficiency depends on the proficiency of the machine hand, and the operation intensity is large and the failure frequency is frequent. When the grain breakage rate and the impurity content of the harvester are exceeding the standard, the lack of on-line monitoring methods and the lack of drivers' experience can not adjust the relevant working parameters in time. On the one hand, it will bring direct economic losses to the farmers. On the other hand, the farmers have a low germination rate when the grain breeding is high in broken rate. Serious breakage affects the production of the next season. In view of the above conditions, this paper focuses on the monitoring methods of the grain breakage rate and impurity content of the combined harvester, and develops the monitoring system to monitor the crushing rate and impurity of the grain in the combined harvester, and the monitoring results are shown by the display of the combined harvester driver's driver's display screen. The main research contents of this paper are as follows: (1) the monitoring methods of grain breaking rate and impurity ratio are studied. According to the working flow of the combined harvester, the reasons of the rate of breakage are analyzed and the image processing technology is put forward to monitor the grain breaking rate and impurity rate of the combined harvester. The calculation model of crushing rate and impurity rate is obtained through theoretical analysis. (2) overall design of grain crushing rate and impurity content monitoring system, including system hardware design and system software design. The system hardware design includes the design of grain on-line collection device, the design of the lighting system of the collection device, and the selection of the image acquisition module. Selection and selection of the system processor module. Among them, the grain collection device is chosen to be installed at the outlet of the hoisting machine of the combined harvester and controls the closure of the floor by the micro electromagnet to ensure that the grain collection device can collect and release grain periodically. (3) the monitoring algorithm of grain breakage rate and impurity rate is studied. In the image segmentation algorithm, the K-Means algorithm is used to extract the broken grain and miscellaneous residual, and then the segmentation is carried out by the watershed algorithm, and the characteristic values of the surplus and grain are extracted. As the input parameters of the BP neural network model, the miscellaneous and broken grain and the perfect grain are identified. At the same time, the design algorithm is in the Windows level, respectively. The software design of the platform and ARM platform is carried out, and the concrete implementation steps and part of the code are given. (4) the bench test is carried out on the joint harvester test platform, and the feasibility of the system device and the algorithm is verified. At the same time, the camera and the industrial camera are used to compare the cost performance of different image sensors and the quality of the collected images. In the bench test, the fragmentation rate and the rate of identification were obtained in two cases. The bench test showed that the average recognition rate of crushing rate and heterozygosity was 88.96% and 88.71%, respectively, when the industrial camera was used to collect grain images, and the average recognition rate of crushing rate and heterozygosity was 68.76% and 66. respectively when using the camera to collect grain images. 6%. (5) transplanted the design algorithm to the embedded platform, and installed the grain breakage rate and heterozygosity monitoring system based on the embedded platform on the joint harvester to carry out field on-line monitoring test. The experiment showed that the average recognition rate of grain breakage rate of the combined harvester was 86.63%, and that of the miscellaneous rate monitoring system was negligible. The average recognition rate of grain heterozygosity is 85.62%. grain breaking rate and heterozygosity, which can be shown by the monitor of the combined harvester in the driver's driving room. It provides the basis for the driver to adjust the working parameters of the combined harvester in time, and satisfies the expected requirements of the design of the combined harvester monitoring system. Now the automatic control of combine harvester provides technical support.
【学位授予单位】：江苏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S225.3;TP391.41

【参考文献】