【作者】 秦云；

【导师】 赵德安；

【作者基本信息】 江苏大学 ， 农业电气化与自动化， 2012， 博士

【摘要】 我国是一个农业大国,社会经济对粮食生产中的收获质量、效率有很高要求。目前,联合收割机在我国农业生产中已逐步普及,有效提高了工作效率、减少了谷物损失。但跨区机收作业中,联合收割机特性会因工况的差异而变化。如不能及时调整工作状态,将导致作业效率下降,故障率上升,机械损耗加剧,严重时可能损坏机器。联合收割机特性复杂,人工控制无法获得良好的控制效果,因此须设计收割机总体负荷自动控制装置,根据实际工况自动控制其收割速度,保证联合收割机总体负荷处于最佳状态。本课题来源于十一五国家科技支撑计划项口“多功能农业装备与设施研制”(2006BAD11A03),以设计制作联合收割机负荷控制系统为目标,对收割机特性、控制算法、控制器软硬件设计等相关领域进行研究,设计了负荷控制算法和装置,并进行大量试验。试验结果表明所设计的控制算法及控制器能根据联合收割机的不同工况及时调整控制参数,在保证高作业质量的前提下大幅度提高作业效率。主要内容围绕以下工作展开：1)为了研究联合收割机的特性,检验各类算法的控制效果,建立了联合收割机整机计算机仿真模型。选择具有代表性的碧浪4LZ2.0型履带式全喂入横置轴流滚筒联合收割机为研究对象,分析其重要部件如柴油发动机、HST、V型传动带、割台、作物输送器、脱粒滚筒、清选筛、输粮搅龙等的工作原理和特性,建立各部件的动力学模型和作物运动模型。根据收割机结构有机整合各子系统模型,构成联合收割机整机数学模型。在Simulink软件环境下设计了收割机计算机仿真模型,对收割机的各种稳态特性、扰动瞬态响应进行研究。结果表明,联合收割机的内部参数漂移和外部扰动对其工作状态有明显的影响,而且会随作业时间、作业对象的变化而变化,难以精确获取。总之,联合收割机系统存在着严重的非线性和时变性,不仅模型阶次高,而且是典型的大惯性、纯延迟系统。2)为加快研究进度,降低研发成本,克服实车试验所受的各种制约,研究、设计了联合收割机半实物仿真模型,用以模拟联合收割机的各种工况,实现对负荷控制器实物的试验测试。通过仿真适配器接收控制器实物发出的控制信号,模型计算机根据数学模型获得各环节的响应,再由仿真适配器输出,模拟传感器检测信号作为控制器反馈。利用相同信号控制联合收割机实车和半实物仿真模型进行对比试验,试验表明半实物仿真模型具有与实车相似的特性,可以在很大程度上模拟收割机的作业过程,实现对负荷控制器实物的实时仿真。3)针对收割机高阶非线性、参数漂移、特性复杂的问题,设计了基于RBF神经网络的自适应负荷控制器。利用RBF网络在线辨识收割机特性参数,以此为基础构成逆模型控制器对收割机进行控制。多种工况条件下的计算机仿真试验表明,控制器对系统参数漂移具有一定适应能力,在不同工况条件下获得较好的控制效果。4)为了解决收割机大惯性、存在纯延时环节等问题,同时基于对其外部扰动的研究,设计了负荷-车速串级控制器。分别利用单神经元PID技术和直接广义预测控制技术设计了车速控制器和负荷控制器。计算机仿真试验表明,引入车速副回路有效提高了系统工作频率,强化了对系统内部参数、外部扰动的适应能力,获得良好的控制效果。5)基于上述研究内容,设计了嵌入式联合收割机负荷控制器,并利用半实物仿真模型和4LZ2.0型联合收割机实车进行大量试验,试验结果表明控制器能够达到设计要求,可实现收割机总体负荷的自动控制,获得较好的作业质量和作业效率。更多还原

【Abstract】 Agriculture occupies an important position in China. The social told the grain production a higher request in the high harvest quality and efficiency. With the popularity of combine, the efficiency is improved and the grain loss is reduced. But in the process of trans-regional harvest, the combine’s characteristics will change because of the different conditions. If operators can not timely adjust the combine’s state, working efficiency will drop, failure rate will rise and the mechanical consumption will aggravate. In severe cases, the combine will be damaged. Combine is a complex system. Artificial control can’t get a good control effect. Therefore, overall load control device must be designed to control the harvest speed automatically and keep the best state of combine.This research originates from the project in the National Science & Technology Pillar Program during the Eleventh Five-year Plan Period(2006BAD11A03). The final purpose of this paper is designing a combine load control system. For this reason, we extensive research the combine’s characteristics, control algorithm, and the design of controller’s software and hardware, etc. And we carried out a lot of test. Test shows that the control algorithm and the controller can adjust its control parameters according to conditions. The working efficiency is improved greatly based on the premise of keeping high working quality. Main contents focus in the work followed.1) In order to study the the combine’s characteristics, inspect the control effect of algorithm, a computer simulation model of combines is designed. In the study, Bilang 4LZ2.0 combine, a crawler-type full-feeding transverse axial flow threshing cylinder harvester is choiced because of its representativeness. Each subsystem’s working principle and characteristics are studied, such as the diesel engine, HST, transmission belt, header, crops conveyor, threshing cylinder, selected sieve, screw conveyor, etc. Their dynamical models and crop movement models are set up respectively. Furtherly, according to the combine’s whole structure, all subsystem models are integrated judiciously, and the overall model of the combine is constituted. Based on this, a computer simulation model of combines is designed using Simulink. Kinds of steady characteristics and disturbance transient response of the combine are studied. Research shows that combined harvesters have serious nonlinear, and its model order is high, there are large inertia and pure delay links. Many parameters of the combine will change slowly. These parameters can’t be identified accurately.2) In order to verify the effectiveness of the load control system, we have to carry out a lot of test under kinds of conditions. For the reason of speeding up the research progress and reducing the cost of research and development, a hardware-in-the-loop simulation model of combines is designed to simulate conditions in the test of load controller object. In the hardware-in-the-loop simulation, the emulation adapter receives control signals from the real controller. The mathematical model in the computer calculates the response of the combine. Then the emulation adapter outputs the response to the controller as feedback. A contrast test between the real combine and the hardware-in-the-loop simulation model is carryed out. A control signal is input to them. Test shows that the characteristics of the hardware-in-the-loop simulation model and the real combine are similar. The model can largely simulate the combine.3) To solve the problem of higher order, nonlinear, parameter drift, a load adaptive controller based on RBF neural network is designed. The characteristic parameters of combines are identified online by RBF network. An inverse model controller is constituted based on these parameters. The computer simulation test under different conditions showed that the controller can adapt to the parameter drift in a certain degree, and get a better control effect.4) In order to solve the large inertia, pure delay, and based on the research of external disturbances, a load-speed cascade controller is designed. A single neurons PID speed controller and a direct generalized predictive control load controller are designed to control the speed and the load respectively. The computer simulation test showed that the system frequency is improved by the speed controller. The controller can adapt the parameter drift and the disturbances, and get a very good control effect.5) An embedded combine load controller is designed based on the research above. The real-time simulation and field test on the 4LZ2.0 combine is carried out. The test showed that the controller can achieve the designed goals. It can control the overall load automatically, and achieve good working quality and working efficiency.更多还原