抓斗挖泥船平挖智能PID控制策略与仿真研究

发布时间：2018-02-06 07:50

本文关键词： 抓斗挖泥船平挖模糊神经网络 PID控制器联合仿真　出处：《武汉理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着几年来,我国经济的高速发展,带来的贸易吞吐量也急剧增多,而现有港口航道均存在多年淤积,水道较浅,输水能力差等问题。因此,国家急需对这些航道进行大规模疏浚清理。然而与庞大的疏浚市场相对应的是,中国的疏浚设备落后,大型设备长期依赖进口。基于上述情况,自主研制新一代大型疏浚设备就显得具有十分重要的意义。现代挖泥船正逐步向着大型化、多功能化、高智能化的方向发展。本文的研究内容就是针对抓斗挖泥船的智能化控制。由于抓斗在水下作业,操作人员无法直观了解抓斗工作情况及挖掘效果,因此不能及时有效得调节抓斗的运动情况;同时,抓斗工作时,又会受到水流、波浪、水底地形等因素的影响,这些都将直接关系到最后的挖掘精度。因此,设计一种具有智能化控制算法的PID控制器来控制抓斗的运动,就显得十分必要了。本文主要的研究内容就是建立一套基于模糊神经网络的控制算法,将其与PID控制器相结合,并通过计算机仿真技术,验证抓斗在平挖作业过程中的挖掘效果。主要工作包括:(1)根据调节抓斗运动的PID控制器的工作原理,建立一个两输入三输出的模糊神经网络。在MATLAB中编写相关控制算法,并给定相关数据,通过网络的自学习功能,优化算法中的相关参数。(2)通过测量实验台所用抓斗的尺寸,建立抓斗的数学模型,并在Simulink中建立抓斗的控制系统模型。将控制器分别作用在上承梁和下承梁上,进行仿真,得出数据进行对比。(3)在Solidworks中画出抓斗的三维简化模型,并导入动力学分析软件ADAMS中。对相关部件进行关联,与Simulink中的控制系统模型建立关联,进行ADAMS和Simulink的联合仿真。得出相关运动曲线及受力情况,与试验数据进行对比,验证控制器挖掘效果。
[Abstract]:With the rapid development of our economy in recent years, the trade throughput has increased sharply, and the existing ports and waterways have many problems, such as sedimentation, shallow waterways, poor water transport capacity and so on. The country urgently needs to carry out large-scale dredging of these waterways. However, in contrast to the huge dredging market, China's dredging equipment is backward and large equipment has long relied on imports. It is of great significance to independently develop a new generation of large dredging equipment. The modern dredger is gradually becoming large and multi-functional. The research content of this paper is aimed at the intelligent control of grab dredger. Because the grab bucket is working under water, the operator can not understand the working condition of grab bucket and the effect of excavating directly. Therefore, the movement of grab can not be adjusted in time and effectively; At the same time, the grab work will be affected by the current, wave, bottom topography and other factors, which will directly affect the final excavation accuracy. It is necessary to design a PID controller with intelligent control algorithm to control the motion of grab. The main research content of this paper is to establish a set of control algorithm based on fuzzy neural network. It is combined with PID controller and is simulated by computer. The main work includes: 1) according to the working principle of PID controller which adjusts grab motion. A fuzzy neural network with two inputs and three outputs is established. The correlation control algorithm is written in MATLAB, and the relevant data is given, and the self-learning function of the network is adopted. The mathematical model of grab bucket is established by measuring the size of grab bucket used in the experiment bench. The control system model of grab bucket is established in Simulink. The controller is acted on the upper beam and the lower beam respectively, and the simulation is carried out. Draw the 3D simplified model of grab bucket in Solidworks, and import it into the dynamic analysis software ADAMS. The joint simulation of ADAMS and Simulink is carried out by establishing the correlation with the control system model in Simulink. The relative motion curve and force are obtained and compared with the experimental data. Verify the controller mining effect.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U615.351;TP273

【参考文献】